Annotation of imach/src/imach.c, revision 1.265
1.265 ! brouard 1: /* $Id: imach.c,v 1.264 2017/04/26 06:01:29 brouard Exp $
1.126 brouard 2: $State: Exp $
1.163 brouard 3: $Log: imach.c,v $
1.265 ! brouard 4: Revision 1.264 2017/04/26 06:01:29 brouard
! 5: Summary: Labels in graphs
! 6:
1.264 brouard 7: Revision 1.263 2017/04/24 15:23:15 brouard
8: Summary: to save
9:
1.263 brouard 10: Revision 1.262 2017/04/18 16:48:12 brouard
11: *** empty log message ***
12:
1.262 brouard 13: Revision 1.261 2017/04/05 10:14:09 brouard
14: Summary: Bug in E_ as well as in T_ fixed nres-1 vs k1-1
15:
1.261 brouard 16: Revision 1.260 2017/04/04 17:46:59 brouard
17: Summary: Gnuplot indexations fixed (humm)
18:
1.260 brouard 19: Revision 1.259 2017/04/04 13:01:16 brouard
20: Summary: Some errors to warnings only if date of death is unknown but status is death we could set to pi3
21:
1.259 brouard 22: Revision 1.258 2017/04/03 10:17:47 brouard
23: Summary: Version 0.99r12
24:
25: Some cleanings, conformed with updated documentation.
26:
1.258 brouard 27: Revision 1.257 2017/03/29 16:53:30 brouard
28: Summary: Temp
29:
1.257 brouard 30: Revision 1.256 2017/03/27 05:50:23 brouard
31: Summary: Temporary
32:
1.256 brouard 33: Revision 1.255 2017/03/08 16:02:28 brouard
34: Summary: IMaCh version 0.99r10 bugs in gnuplot fixed
35:
1.255 brouard 36: Revision 1.254 2017/03/08 07:13:00 brouard
37: Summary: Fixing data parameter line
38:
1.254 brouard 39: Revision 1.253 2016/12/15 11:59:41 brouard
40: Summary: 0.99 in progress
41:
1.253 brouard 42: Revision 1.252 2016/09/15 21:15:37 brouard
43: *** empty log message ***
44:
1.252 brouard 45: Revision 1.251 2016/09/15 15:01:13 brouard
46: Summary: not working
47:
1.251 brouard 48: Revision 1.250 2016/09/08 16:07:27 brouard
49: Summary: continue
50:
1.250 brouard 51: Revision 1.249 2016/09/07 17:14:18 brouard
52: Summary: Starting values from frequencies
53:
1.249 brouard 54: Revision 1.248 2016/09/07 14:10:18 brouard
55: *** empty log message ***
56:
1.248 brouard 57: Revision 1.247 2016/09/02 11:11:21 brouard
58: *** empty log message ***
59:
1.247 brouard 60: Revision 1.246 2016/09/02 08:49:22 brouard
61: *** empty log message ***
62:
1.246 brouard 63: Revision 1.245 2016/09/02 07:25:01 brouard
64: *** empty log message ***
65:
1.245 brouard 66: Revision 1.244 2016/09/02 07:17:34 brouard
67: *** empty log message ***
68:
1.244 brouard 69: Revision 1.243 2016/09/02 06:45:35 brouard
70: *** empty log message ***
71:
1.243 brouard 72: Revision 1.242 2016/08/30 15:01:20 brouard
73: Summary: Fixing a lots
74:
1.242 brouard 75: Revision 1.241 2016/08/29 17:17:25 brouard
76: Summary: gnuplot problem in Back projection to fix
77:
1.241 brouard 78: Revision 1.240 2016/08/29 07:53:18 brouard
79: Summary: Better
80:
1.240 brouard 81: Revision 1.239 2016/08/26 15:51:03 brouard
82: Summary: Improvement in Powell output in order to copy and paste
83:
84: Author:
85:
1.239 brouard 86: Revision 1.238 2016/08/26 14:23:35 brouard
87: Summary: Starting tests of 0.99
88:
1.238 brouard 89: Revision 1.237 2016/08/26 09:20:19 brouard
90: Summary: to valgrind
91:
1.237 brouard 92: Revision 1.236 2016/08/25 10:50:18 brouard
93: *** empty log message ***
94:
1.236 brouard 95: Revision 1.235 2016/08/25 06:59:23 brouard
96: *** empty log message ***
97:
1.235 brouard 98: Revision 1.234 2016/08/23 16:51:20 brouard
99: *** empty log message ***
100:
1.234 brouard 101: Revision 1.233 2016/08/23 07:40:50 brouard
102: Summary: not working
103:
1.233 brouard 104: Revision 1.232 2016/08/22 14:20:21 brouard
105: Summary: not working
106:
1.232 brouard 107: Revision 1.231 2016/08/22 07:17:15 brouard
108: Summary: not working
109:
1.231 brouard 110: Revision 1.230 2016/08/22 06:55:53 brouard
111: Summary: Not working
112:
1.230 brouard 113: Revision 1.229 2016/07/23 09:45:53 brouard
114: Summary: Completing for func too
115:
1.229 brouard 116: Revision 1.228 2016/07/22 17:45:30 brouard
117: Summary: Fixing some arrays, still debugging
118:
1.227 brouard 119: Revision 1.226 2016/07/12 18:42:34 brouard
120: Summary: temp
121:
1.226 brouard 122: Revision 1.225 2016/07/12 08:40:03 brouard
123: Summary: saving but not running
124:
1.225 brouard 125: Revision 1.224 2016/07/01 13:16:01 brouard
126: Summary: Fixes
127:
1.224 brouard 128: Revision 1.223 2016/02/19 09:23:35 brouard
129: Summary: temporary
130:
1.223 brouard 131: Revision 1.222 2016/02/17 08:14:50 brouard
132: Summary: Probably last 0.98 stable version 0.98r6
133:
1.222 brouard 134: Revision 1.221 2016/02/15 23:35:36 brouard
135: Summary: minor bug
136:
1.220 brouard 137: Revision 1.219 2016/02/15 00:48:12 brouard
138: *** empty log message ***
139:
1.219 brouard 140: Revision 1.218 2016/02/12 11:29:23 brouard
141: Summary: 0.99 Back projections
142:
1.218 brouard 143: Revision 1.217 2015/12/23 17:18:31 brouard
144: Summary: Experimental backcast
145:
1.217 brouard 146: Revision 1.216 2015/12/18 17:32:11 brouard
147: Summary: 0.98r4 Warning and status=-2
148:
149: Version 0.98r4 is now:
150: - displaying an error when status is -1, date of interview unknown and date of death known;
151: - permitting a status -2 when the vital status is unknown at a known date of right truncation.
152: Older changes concerning s=-2, dating from 2005 have been supersed.
153:
1.216 brouard 154: Revision 1.215 2015/12/16 08:52:24 brouard
155: Summary: 0.98r4 working
156:
1.215 brouard 157: Revision 1.214 2015/12/16 06:57:54 brouard
158: Summary: temporary not working
159:
1.214 brouard 160: Revision 1.213 2015/12/11 18:22:17 brouard
161: Summary: 0.98r4
162:
1.213 brouard 163: Revision 1.212 2015/11/21 12:47:24 brouard
164: Summary: minor typo
165:
1.212 brouard 166: Revision 1.211 2015/11/21 12:41:11 brouard
167: Summary: 0.98r3 with some graph of projected cross-sectional
168:
169: Author: Nicolas Brouard
170:
1.211 brouard 171: Revision 1.210 2015/11/18 17:41:20 brouard
1.252 brouard 172: Summary: Start working on projected prevalences Revision 1.209 2015/11/17 22:12:03 brouard
1.210 brouard 173: Summary: Adding ftolpl parameter
174: Author: N Brouard
175:
176: We had difficulties to get smoothed confidence intervals. It was due
177: to the period prevalence which wasn't computed accurately. The inner
178: parameter ftolpl is now an outer parameter of the .imach parameter
179: file after estepm. If ftolpl is small 1.e-4 and estepm too,
180: computation are long.
181:
1.209 brouard 182: Revision 1.208 2015/11/17 14:31:57 brouard
183: Summary: temporary
184:
1.208 brouard 185: Revision 1.207 2015/10/27 17:36:57 brouard
186: *** empty log message ***
187:
1.207 brouard 188: Revision 1.206 2015/10/24 07:14:11 brouard
189: *** empty log message ***
190:
1.206 brouard 191: Revision 1.205 2015/10/23 15:50:53 brouard
192: Summary: 0.98r3 some clarification for graphs on likelihood contributions
193:
1.205 brouard 194: Revision 1.204 2015/10/01 16:20:26 brouard
195: Summary: Some new graphs of contribution to likelihood
196:
1.204 brouard 197: Revision 1.203 2015/09/30 17:45:14 brouard
198: Summary: looking at better estimation of the hessian
199:
200: Also a better criteria for convergence to the period prevalence And
201: therefore adding the number of years needed to converge. (The
202: prevalence in any alive state shold sum to one
203:
1.203 brouard 204: Revision 1.202 2015/09/22 19:45:16 brouard
205: Summary: Adding some overall graph on contribution to likelihood. Might change
206:
1.202 brouard 207: Revision 1.201 2015/09/15 17:34:58 brouard
208: Summary: 0.98r0
209:
210: - Some new graphs like suvival functions
211: - Some bugs fixed like model=1+age+V2.
212:
1.201 brouard 213: Revision 1.200 2015/09/09 16:53:55 brouard
214: Summary: Big bug thanks to Flavia
215:
216: Even model=1+age+V2. did not work anymore
217:
1.200 brouard 218: Revision 1.199 2015/09/07 14:09:23 brouard
219: Summary: 0.98q6 changing default small png format for graph to vectorized svg.
220:
1.199 brouard 221: Revision 1.198 2015/09/03 07:14:39 brouard
222: Summary: 0.98q5 Flavia
223:
1.198 brouard 224: Revision 1.197 2015/09/01 18:24:39 brouard
225: *** empty log message ***
226:
1.197 brouard 227: Revision 1.196 2015/08/18 23:17:52 brouard
228: Summary: 0.98q5
229:
1.196 brouard 230: Revision 1.195 2015/08/18 16:28:39 brouard
231: Summary: Adding a hack for testing purpose
232:
233: After reading the title, ftol and model lines, if the comment line has
234: a q, starting with #q, the answer at the end of the run is quit. It
235: permits to run test files in batch with ctest. The former workaround was
236: $ echo q | imach foo.imach
237:
1.195 brouard 238: Revision 1.194 2015/08/18 13:32:00 brouard
239: Summary: Adding error when the covariance matrix doesn't contain the exact number of lines required by the model line.
240:
1.194 brouard 241: Revision 1.193 2015/08/04 07:17:42 brouard
242: Summary: 0.98q4
243:
1.193 brouard 244: Revision 1.192 2015/07/16 16:49:02 brouard
245: Summary: Fixing some outputs
246:
1.192 brouard 247: Revision 1.191 2015/07/14 10:00:33 brouard
248: Summary: Some fixes
249:
1.191 brouard 250: Revision 1.190 2015/05/05 08:51:13 brouard
251: Summary: Adding digits in output parameters (7 digits instead of 6)
252:
253: Fix 1+age+.
254:
1.190 brouard 255: Revision 1.189 2015/04/30 14:45:16 brouard
256: Summary: 0.98q2
257:
1.189 brouard 258: Revision 1.188 2015/04/30 08:27:53 brouard
259: *** empty log message ***
260:
1.188 brouard 261: Revision 1.187 2015/04/29 09:11:15 brouard
262: *** empty log message ***
263:
1.187 brouard 264: Revision 1.186 2015/04/23 12:01:52 brouard
265: Summary: V1*age is working now, version 0.98q1
266:
267: Some codes had been disabled in order to simplify and Vn*age was
268: working in the optimization phase, ie, giving correct MLE parameters,
269: but, as usual, outputs were not correct and program core dumped.
270:
1.186 brouard 271: Revision 1.185 2015/03/11 13:26:42 brouard
272: Summary: Inclusion of compile and links command line for Intel Compiler
273:
1.185 brouard 274: Revision 1.184 2015/03/11 11:52:39 brouard
275: Summary: Back from Windows 8. Intel Compiler
276:
1.184 brouard 277: Revision 1.183 2015/03/10 20:34:32 brouard
278: Summary: 0.98q0, trying with directest, mnbrak fixed
279:
280: We use directest instead of original Powell test; probably no
281: incidence on the results, but better justifications;
282: We fixed Numerical Recipes mnbrak routine which was wrong and gave
283: wrong results.
284:
1.183 brouard 285: Revision 1.182 2015/02/12 08:19:57 brouard
286: Summary: Trying to keep directest which seems simpler and more general
287: Author: Nicolas Brouard
288:
1.182 brouard 289: Revision 1.181 2015/02/11 23:22:24 brouard
290: Summary: Comments on Powell added
291:
292: Author:
293:
1.181 brouard 294: Revision 1.180 2015/02/11 17:33:45 brouard
295: Summary: Finishing move from main to function (hpijx and prevalence_limit)
296:
1.180 brouard 297: Revision 1.179 2015/01/04 09:57:06 brouard
298: Summary: back to OS/X
299:
1.179 brouard 300: Revision 1.178 2015/01/04 09:35:48 brouard
301: *** empty log message ***
302:
1.178 brouard 303: Revision 1.177 2015/01/03 18:40:56 brouard
304: Summary: Still testing ilc32 on OSX
305:
1.177 brouard 306: Revision 1.176 2015/01/03 16:45:04 brouard
307: *** empty log message ***
308:
1.176 brouard 309: Revision 1.175 2015/01/03 16:33:42 brouard
310: *** empty log message ***
311:
1.175 brouard 312: Revision 1.174 2015/01/03 16:15:49 brouard
313: Summary: Still in cross-compilation
314:
1.174 brouard 315: Revision 1.173 2015/01/03 12:06:26 brouard
316: Summary: trying to detect cross-compilation
317:
1.173 brouard 318: Revision 1.172 2014/12/27 12:07:47 brouard
319: Summary: Back from Visual Studio and Intel, options for compiling for Windows XP
320:
1.172 brouard 321: Revision 1.171 2014/12/23 13:26:59 brouard
322: Summary: Back from Visual C
323:
324: Still problem with utsname.h on Windows
325:
1.171 brouard 326: Revision 1.170 2014/12/23 11:17:12 brouard
327: Summary: Cleaning some \%% back to %%
328:
329: The escape was mandatory for a specific compiler (which one?), but too many warnings.
330:
1.170 brouard 331: Revision 1.169 2014/12/22 23:08:31 brouard
332: Summary: 0.98p
333:
334: Outputs some informations on compiler used, OS etc. Testing on different platforms.
335:
1.169 brouard 336: Revision 1.168 2014/12/22 15:17:42 brouard
1.170 brouard 337: Summary: update
1.169 brouard 338:
1.168 brouard 339: Revision 1.167 2014/12/22 13:50:56 brouard
340: Summary: Testing uname and compiler version and if compiled 32 or 64
341:
342: Testing on Linux 64
343:
1.167 brouard 344: Revision 1.166 2014/12/22 11:40:47 brouard
345: *** empty log message ***
346:
1.166 brouard 347: Revision 1.165 2014/12/16 11:20:36 brouard
348: Summary: After compiling on Visual C
349:
350: * imach.c (Module): Merging 1.61 to 1.162
351:
1.165 brouard 352: Revision 1.164 2014/12/16 10:52:11 brouard
353: Summary: Merging with Visual C after suppressing some warnings for unused variables. Also fixing Saito's bug 0.98Xn
354:
355: * imach.c (Module): Merging 1.61 to 1.162
356:
1.164 brouard 357: Revision 1.163 2014/12/16 10:30:11 brouard
358: * imach.c (Module): Merging 1.61 to 1.162
359:
1.163 brouard 360: Revision 1.162 2014/09/25 11:43:39 brouard
361: Summary: temporary backup 0.99!
362:
1.162 brouard 363: Revision 1.1 2014/09/16 11:06:58 brouard
364: Summary: With some code (wrong) for nlopt
365:
366: Author:
367:
368: Revision 1.161 2014/09/15 20:41:41 brouard
369: Summary: Problem with macro SQR on Intel compiler
370:
1.161 brouard 371: Revision 1.160 2014/09/02 09:24:05 brouard
372: *** empty log message ***
373:
1.160 brouard 374: Revision 1.159 2014/09/01 10:34:10 brouard
375: Summary: WIN32
376: Author: Brouard
377:
1.159 brouard 378: Revision 1.158 2014/08/27 17:11:51 brouard
379: *** empty log message ***
380:
1.158 brouard 381: Revision 1.157 2014/08/27 16:26:55 brouard
382: Summary: Preparing windows Visual studio version
383: Author: Brouard
384:
385: In order to compile on Visual studio, time.h is now correct and time_t
386: and tm struct should be used. difftime should be used but sometimes I
387: just make the differences in raw time format (time(&now).
388: Trying to suppress #ifdef LINUX
389: Add xdg-open for __linux in order to open default browser.
390:
1.157 brouard 391: Revision 1.156 2014/08/25 20:10:10 brouard
392: *** empty log message ***
393:
1.156 brouard 394: Revision 1.155 2014/08/25 18:32:34 brouard
395: Summary: New compile, minor changes
396: Author: Brouard
397:
1.155 brouard 398: Revision 1.154 2014/06/20 17:32:08 brouard
399: Summary: Outputs now all graphs of convergence to period prevalence
400:
1.154 brouard 401: Revision 1.153 2014/06/20 16:45:46 brouard
402: Summary: If 3 live state, convergence to period prevalence on same graph
403: Author: Brouard
404:
1.153 brouard 405: Revision 1.152 2014/06/18 17:54:09 brouard
406: Summary: open browser, use gnuplot on same dir than imach if not found in the path
407:
1.152 brouard 408: Revision 1.151 2014/06/18 16:43:30 brouard
409: *** empty log message ***
410:
1.151 brouard 411: Revision 1.150 2014/06/18 16:42:35 brouard
412: Summary: If gnuplot is not in the path try on same directory than imach binary (OSX)
413: Author: brouard
414:
1.150 brouard 415: Revision 1.149 2014/06/18 15:51:14 brouard
416: Summary: Some fixes in parameter files errors
417: Author: Nicolas Brouard
418:
1.149 brouard 419: Revision 1.148 2014/06/17 17:38:48 brouard
420: Summary: Nothing new
421: Author: Brouard
422:
423: Just a new packaging for OS/X version 0.98nS
424:
1.148 brouard 425: Revision 1.147 2014/06/16 10:33:11 brouard
426: *** empty log message ***
427:
1.147 brouard 428: Revision 1.146 2014/06/16 10:20:28 brouard
429: Summary: Merge
430: Author: Brouard
431:
432: Merge, before building revised version.
433:
1.146 brouard 434: Revision 1.145 2014/06/10 21:23:15 brouard
435: Summary: Debugging with valgrind
436: Author: Nicolas Brouard
437:
438: Lot of changes in order to output the results with some covariates
439: After the Edimburgh REVES conference 2014, it seems mandatory to
440: improve the code.
441: No more memory valgrind error but a lot has to be done in order to
442: continue the work of splitting the code into subroutines.
443: Also, decodemodel has been improved. Tricode is still not
444: optimal. nbcode should be improved. Documentation has been added in
445: the source code.
446:
1.144 brouard 447: Revision 1.143 2014/01/26 09:45:38 brouard
448: Summary: Version 0.98nR (to be improved, but gives same optimization results as 0.98k. Nice, promising
449:
450: * imach.c (Module): Trying to merge old staffs together while being at Tokyo. Not tested...
451: (Module): Version 0.98nR Running ok, but output format still only works for three covariates.
452:
1.143 brouard 453: Revision 1.142 2014/01/26 03:57:36 brouard
454: Summary: gnuplot changed plot w l 1 has to be changed to plot w l lt 2
455:
456: * imach.c (Module): Trying to merge old staffs together while being at Tokyo. Not tested...
457:
1.142 brouard 458: Revision 1.141 2014/01/26 02:42:01 brouard
459: * imach.c (Module): Trying to merge old staffs together while being at Tokyo. Not tested...
460:
1.141 brouard 461: Revision 1.140 2011/09/02 10:37:54 brouard
462: Summary: times.h is ok with mingw32 now.
463:
1.140 brouard 464: Revision 1.139 2010/06/14 07:50:17 brouard
465: After the theft of my laptop, I probably lost some lines of codes which were not uploaded to the CVS tree.
466: I remember having already fixed agemin agemax which are pointers now but not cvs saved.
467:
1.139 brouard 468: Revision 1.138 2010/04/30 18:19:40 brouard
469: *** empty log message ***
470:
1.138 brouard 471: Revision 1.137 2010/04/29 18:11:38 brouard
472: (Module): Checking covariates for more complex models
473: than V1+V2. A lot of change to be done. Unstable.
474:
1.137 brouard 475: Revision 1.136 2010/04/26 20:30:53 brouard
476: (Module): merging some libgsl code. Fixing computation
477: of likelione (using inter/intrapolation if mle = 0) in order to
478: get same likelihood as if mle=1.
479: Some cleaning of code and comments added.
480:
1.136 brouard 481: Revision 1.135 2009/10/29 15:33:14 brouard
482: (Module): Now imach stops if date of birth, at least year of birth, is not given. Some cleaning of the code.
483:
1.135 brouard 484: Revision 1.134 2009/10/29 13:18:53 brouard
485: (Module): Now imach stops if date of birth, at least year of birth, is not given. Some cleaning of the code.
486:
1.134 brouard 487: Revision 1.133 2009/07/06 10:21:25 brouard
488: just nforces
489:
1.133 brouard 490: Revision 1.132 2009/07/06 08:22:05 brouard
491: Many tings
492:
1.132 brouard 493: Revision 1.131 2009/06/20 16:22:47 brouard
494: Some dimensions resccaled
495:
1.131 brouard 496: Revision 1.130 2009/05/26 06:44:34 brouard
497: (Module): Max Covariate is now set to 20 instead of 8. A
498: lot of cleaning with variables initialized to 0. Trying to make
499: V2+V3*age+V1+V4 strb=V3*age+V1+V4 working better.
500:
1.130 brouard 501: Revision 1.129 2007/08/31 13:49:27 lievre
502: Modification of the way of exiting when the covariate is not binary in order to see on the window the error message before exiting
503:
1.129 lievre 504: Revision 1.128 2006/06/30 13:02:05 brouard
505: (Module): Clarifications on computing e.j
506:
1.128 brouard 507: Revision 1.127 2006/04/28 18:11:50 brouard
508: (Module): Yes the sum of survivors was wrong since
509: imach-114 because nhstepm was no more computed in the age
510: loop. Now we define nhstepma in the age loop.
511: (Module): In order to speed up (in case of numerous covariates) we
512: compute health expectancies (without variances) in a first step
513: and then all the health expectancies with variances or standard
514: deviation (needs data from the Hessian matrices) which slows the
515: computation.
516: In the future we should be able to stop the program is only health
517: expectancies and graph are needed without standard deviations.
518:
1.127 brouard 519: Revision 1.126 2006/04/28 17:23:28 brouard
520: (Module): Yes the sum of survivors was wrong since
521: imach-114 because nhstepm was no more computed in the age
522: loop. Now we define nhstepma in the age loop.
523: Version 0.98h
524:
1.126 brouard 525: Revision 1.125 2006/04/04 15:20:31 lievre
526: Errors in calculation of health expectancies. Age was not initialized.
527: Forecasting file added.
528:
529: Revision 1.124 2006/03/22 17:13:53 lievre
530: Parameters are printed with %lf instead of %f (more numbers after the comma).
531: The log-likelihood is printed in the log file
532:
533: Revision 1.123 2006/03/20 10:52:43 brouard
534: * imach.c (Module): <title> changed, corresponds to .htm file
535: name. <head> headers where missing.
536:
537: * imach.c (Module): Weights can have a decimal point as for
538: English (a comma might work with a correct LC_NUMERIC environment,
539: otherwise the weight is truncated).
540: Modification of warning when the covariates values are not 0 or
541: 1.
542: Version 0.98g
543:
544: Revision 1.122 2006/03/20 09:45:41 brouard
545: (Module): Weights can have a decimal point as for
546: English (a comma might work with a correct LC_NUMERIC environment,
547: otherwise the weight is truncated).
548: Modification of warning when the covariates values are not 0 or
549: 1.
550: Version 0.98g
551:
552: Revision 1.121 2006/03/16 17:45:01 lievre
553: * imach.c (Module): Comments concerning covariates added
554:
555: * imach.c (Module): refinements in the computation of lli if
556: status=-2 in order to have more reliable computation if stepm is
557: not 1 month. Version 0.98f
558:
559: Revision 1.120 2006/03/16 15:10:38 lievre
560: (Module): refinements in the computation of lli if
561: status=-2 in order to have more reliable computation if stepm is
562: not 1 month. Version 0.98f
563:
564: Revision 1.119 2006/03/15 17:42:26 brouard
565: (Module): Bug if status = -2, the loglikelihood was
566: computed as likelihood omitting the logarithm. Version O.98e
567:
568: Revision 1.118 2006/03/14 18:20:07 brouard
569: (Module): varevsij Comments added explaining the second
570: table of variances if popbased=1 .
571: (Module): Covariances of eij, ekl added, graphs fixed, new html link.
572: (Module): Function pstamp added
573: (Module): Version 0.98d
574:
575: Revision 1.117 2006/03/14 17:16:22 brouard
576: (Module): varevsij Comments added explaining the second
577: table of variances if popbased=1 .
578: (Module): Covariances of eij, ekl added, graphs fixed, new html link.
579: (Module): Function pstamp added
580: (Module): Version 0.98d
581:
582: Revision 1.116 2006/03/06 10:29:27 brouard
583: (Module): Variance-covariance wrong links and
584: varian-covariance of ej. is needed (Saito).
585:
586: Revision 1.115 2006/02/27 12:17:45 brouard
587: (Module): One freematrix added in mlikeli! 0.98c
588:
589: Revision 1.114 2006/02/26 12:57:58 brouard
590: (Module): Some improvements in processing parameter
591: filename with strsep.
592:
593: Revision 1.113 2006/02/24 14:20:24 brouard
594: (Module): Memory leaks checks with valgrind and:
595: datafile was not closed, some imatrix were not freed and on matrix
596: allocation too.
597:
598: Revision 1.112 2006/01/30 09:55:26 brouard
599: (Module): Back to gnuplot.exe instead of wgnuplot.exe
600:
601: Revision 1.111 2006/01/25 20:38:18 brouard
602: (Module): Lots of cleaning and bugs added (Gompertz)
603: (Module): Comments can be added in data file. Missing date values
604: can be a simple dot '.'.
605:
606: Revision 1.110 2006/01/25 00:51:50 brouard
607: (Module): Lots of cleaning and bugs added (Gompertz)
608:
609: Revision 1.109 2006/01/24 19:37:15 brouard
610: (Module): Comments (lines starting with a #) are allowed in data.
611:
612: Revision 1.108 2006/01/19 18:05:42 lievre
613: Gnuplot problem appeared...
614: To be fixed
615:
616: Revision 1.107 2006/01/19 16:20:37 brouard
617: Test existence of gnuplot in imach path
618:
619: Revision 1.106 2006/01/19 13:24:36 brouard
620: Some cleaning and links added in html output
621:
622: Revision 1.105 2006/01/05 20:23:19 lievre
623: *** empty log message ***
624:
625: Revision 1.104 2005/09/30 16:11:43 lievre
626: (Module): sump fixed, loop imx fixed, and simplifications.
627: (Module): If the status is missing at the last wave but we know
628: that the person is alive, then we can code his/her status as -2
629: (instead of missing=-1 in earlier versions) and his/her
630: contributions to the likelihood is 1 - Prob of dying from last
631: health status (= 1-p13= p11+p12 in the easiest case of somebody in
632: the healthy state at last known wave). Version is 0.98
633:
634: Revision 1.103 2005/09/30 15:54:49 lievre
635: (Module): sump fixed, loop imx fixed, and simplifications.
636:
637: Revision 1.102 2004/09/15 17:31:30 brouard
638: Add the possibility to read data file including tab characters.
639:
640: Revision 1.101 2004/09/15 10:38:38 brouard
641: Fix on curr_time
642:
643: Revision 1.100 2004/07/12 18:29:06 brouard
644: Add version for Mac OS X. Just define UNIX in Makefile
645:
646: Revision 1.99 2004/06/05 08:57:40 brouard
647: *** empty log message ***
648:
649: Revision 1.98 2004/05/16 15:05:56 brouard
650: New version 0.97 . First attempt to estimate force of mortality
651: directly from the data i.e. without the need of knowing the health
652: state at each age, but using a Gompertz model: log u =a + b*age .
653: This is the basic analysis of mortality and should be done before any
654: other analysis, in order to test if the mortality estimated from the
655: cross-longitudinal survey is different from the mortality estimated
656: from other sources like vital statistic data.
657:
658: The same imach parameter file can be used but the option for mle should be -3.
659:
1.133 brouard 660: Agnès, who wrote this part of the code, tried to keep most of the
1.126 brouard 661: former routines in order to include the new code within the former code.
662:
663: The output is very simple: only an estimate of the intercept and of
664: the slope with 95% confident intervals.
665:
666: Current limitations:
667: A) Even if you enter covariates, i.e. with the
668: model= V1+V2 equation for example, the programm does only estimate a unique global model without covariates.
669: B) There is no computation of Life Expectancy nor Life Table.
670:
671: Revision 1.97 2004/02/20 13:25:42 lievre
672: Version 0.96d. Population forecasting command line is (temporarily)
673: suppressed.
674:
675: Revision 1.96 2003/07/15 15:38:55 brouard
676: * imach.c (Repository): Errors in subdirf, 2, 3 while printing tmpout is
677: rewritten within the same printf. Workaround: many printfs.
678:
679: Revision 1.95 2003/07/08 07:54:34 brouard
680: * imach.c (Repository):
681: (Repository): Using imachwizard code to output a more meaningful covariance
682: matrix (cov(a12,c31) instead of numbers.
683:
684: Revision 1.94 2003/06/27 13:00:02 brouard
685: Just cleaning
686:
687: Revision 1.93 2003/06/25 16:33:55 brouard
688: (Module): On windows (cygwin) function asctime_r doesn't
689: exist so I changed back to asctime which exists.
690: (Module): Version 0.96b
691:
692: Revision 1.92 2003/06/25 16:30:45 brouard
693: (Module): On windows (cygwin) function asctime_r doesn't
694: exist so I changed back to asctime which exists.
695:
696: Revision 1.91 2003/06/25 15:30:29 brouard
697: * imach.c (Repository): Duplicated warning errors corrected.
698: (Repository): Elapsed time after each iteration is now output. It
699: helps to forecast when convergence will be reached. Elapsed time
700: is stamped in powell. We created a new html file for the graphs
701: concerning matrix of covariance. It has extension -cov.htm.
702:
703: Revision 1.90 2003/06/24 12:34:15 brouard
704: (Module): Some bugs corrected for windows. Also, when
705: mle=-1 a template is output in file "or"mypar.txt with the design
706: of the covariance matrix to be input.
707:
708: Revision 1.89 2003/06/24 12:30:52 brouard
709: (Module): Some bugs corrected for windows. Also, when
710: mle=-1 a template is output in file "or"mypar.txt with the design
711: of the covariance matrix to be input.
712:
713: Revision 1.88 2003/06/23 17:54:56 brouard
714: * imach.c (Repository): Create a sub-directory where all the secondary files are. Only imach, htm, gp and r(imach) are on the main directory. Correct time and other things.
715:
716: Revision 1.87 2003/06/18 12:26:01 brouard
717: Version 0.96
718:
719: Revision 1.86 2003/06/17 20:04:08 brouard
720: (Module): Change position of html and gnuplot routines and added
721: routine fileappend.
722:
723: Revision 1.85 2003/06/17 13:12:43 brouard
724: * imach.c (Repository): Check when date of death was earlier that
725: current date of interview. It may happen when the death was just
726: prior to the death. In this case, dh was negative and likelihood
727: was wrong (infinity). We still send an "Error" but patch by
728: assuming that the date of death was just one stepm after the
729: interview.
730: (Repository): Because some people have very long ID (first column)
731: we changed int to long in num[] and we added a new lvector for
732: memory allocation. But we also truncated to 8 characters (left
733: truncation)
734: (Repository): No more line truncation errors.
735:
736: Revision 1.84 2003/06/13 21:44:43 brouard
737: * imach.c (Repository): Replace "freqsummary" at a correct
738: place. It differs from routine "prevalence" which may be called
739: many times. Probs is memory consuming and must be used with
740: parcimony.
741: Version 0.95a3 (should output exactly the same maximization than 0.8a2)
742:
743: Revision 1.83 2003/06/10 13:39:11 lievre
744: *** empty log message ***
745:
746: Revision 1.82 2003/06/05 15:57:20 brouard
747: Add log in imach.c and fullversion number is now printed.
748:
749: */
750: /*
751: Interpolated Markov Chain
752:
753: Short summary of the programme:
754:
1.227 brouard 755: This program computes Healthy Life Expectancies or State-specific
756: (if states aren't health statuses) Expectancies from
757: cross-longitudinal data. Cross-longitudinal data consist in:
758:
759: -1- a first survey ("cross") where individuals from different ages
760: are interviewed on their health status or degree of disability (in
761: the case of a health survey which is our main interest)
762:
763: -2- at least a second wave of interviews ("longitudinal") which
764: measure each change (if any) in individual health status. Health
765: expectancies are computed from the time spent in each health state
766: according to a model. More health states you consider, more time is
767: necessary to reach the Maximum Likelihood of the parameters involved
768: in the model. The simplest model is the multinomial logistic model
769: where pij is the probability to be observed in state j at the second
770: wave conditional to be observed in state i at the first
771: wave. Therefore the model is: log(pij/pii)= aij + bij*age+ cij*sex +
772: etc , where 'age' is age and 'sex' is a covariate. If you want to
773: have a more complex model than "constant and age", you should modify
774: the program where the markup *Covariates have to be included here
775: again* invites you to do it. More covariates you add, slower the
1.126 brouard 776: convergence.
777:
778: The advantage of this computer programme, compared to a simple
779: multinomial logistic model, is clear when the delay between waves is not
780: identical for each individual. Also, if a individual missed an
781: intermediate interview, the information is lost, but taken into
782: account using an interpolation or extrapolation.
783:
784: hPijx is the probability to be observed in state i at age x+h
785: conditional to the observed state i at age x. The delay 'h' can be
786: split into an exact number (nh*stepm) of unobserved intermediate
787: states. This elementary transition (by month, quarter,
788: semester or year) is modelled as a multinomial logistic. The hPx
789: matrix is simply the matrix product of nh*stepm elementary matrices
790: and the contribution of each individual to the likelihood is simply
791: hPijx.
792:
793: Also this programme outputs the covariance matrix of the parameters but also
1.218 brouard 794: of the life expectancies. It also computes the period (stable) prevalence.
795:
796: Back prevalence and projections:
1.227 brouard 797:
798: - back_prevalence_limit(double *p, double **bprlim, double ageminpar,
799: double agemaxpar, double ftolpl, int *ncvyearp, double
800: dateprev1,double dateprev2, int firstpass, int lastpass, int
801: mobilavproj)
802:
803: Computes the back prevalence limit for any combination of
804: covariate values k at any age between ageminpar and agemaxpar and
805: returns it in **bprlim. In the loops,
806:
807: - **bprevalim(**bprlim, ***mobaverage, nlstate, *p, age, **oldm,
808: **savm, **dnewm, **doldm, **dsavm, ftolpl, ncvyearp, k);
809:
810: - hBijx Back Probability to be in state i at age x-h being in j at x
1.218 brouard 811: Computes for any combination of covariates k and any age between bage and fage
812: p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
813: oldm=oldms;savm=savms;
1.227 brouard 814:
815: - hbxij(p3mat,nhstepm,agedeb,hstepm,p,nlstate,stepm,oldm,savm, k);
1.218 brouard 816: Computes the transition matrix starting at age 'age' over
817: 'nhstepm*hstepm*stepm' months (i.e. until
818: age (in years) age+nhstepm*hstepm*stepm/12) by multiplying
1.227 brouard 819: nhstepm*hstepm matrices.
820:
821: Returns p3mat[i][j][h] after calling
822: p3mat[i][j][h]=matprod2(newm,
823: bmij(pmmij,cov,ncovmodel,x,nlstate,prevacurrent, dnewm, doldm,
824: dsavm,ij),\ 1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath,
825: oldm);
1.226 brouard 826:
827: Important routines
828:
829: - func (or funcone), computes logit (pij) distinguishing
830: o fixed variables (single or product dummies or quantitative);
831: o varying variables by:
832: (1) wave (single, product dummies, quantitative),
833: (2) by age (can be month) age (done), age*age (done), age*Vn where Vn can be:
834: % fixed dummy (treated) or quantitative (not done because time-consuming);
835: % varying dummy (not done) or quantitative (not done);
836: - Tricode which tests the modality of dummy variables (in order to warn with wrong or empty modalities)
837: and returns the number of efficient covariates cptcoveff and modalities nbcode[Tvar[k]][1]= 0 and nbcode[Tvar[k]][2]= 1 usually.
838: - printinghtml which outputs results like life expectancy in and from a state for a combination of modalities of dummy variables
839: o There are 2*cptcoveff combinations of (0,1) for cptcoveff variables. Outputting only combinations with people, éliminating 1 1 if
840: race White (0 0), Black vs White (1 0), Hispanic (0 1) and 1 1 being meaningless.
1.218 brouard 841:
1.226 brouard 842:
843:
1.133 brouard 844: Authors: Nicolas Brouard (brouard@ined.fr) and Agnès Lièvre (lievre@ined.fr).
845: Institut national d'études démographiques, Paris.
1.126 brouard 846: This software have been partly granted by Euro-REVES, a concerted action
847: from the European Union.
848: It is copyrighted identically to a GNU software product, ie programme and
849: software can be distributed freely for non commercial use. Latest version
850: can be accessed at http://euroreves.ined.fr/imach .
851:
852: Help to debug: LD_PRELOAD=/usr/local/lib/libnjamd.so ./imach foo.imach
853: or better on gdb : set env LD_PRELOAD=/usr/local/lib/libnjamd.so
854:
855: **********************************************************************/
856: /*
857: main
858: read parameterfile
859: read datafile
860: concatwav
861: freqsummary
862: if (mle >= 1)
863: mlikeli
864: print results files
865: if mle==1
866: computes hessian
867: read end of parameter file: agemin, agemax, bage, fage, estepm
868: begin-prev-date,...
869: open gnuplot file
870: open html file
1.145 brouard 871: period (stable) prevalence | pl_nom 1-1 2-2 etc by covariate
872: for age prevalim() | #****** V1=0 V2=1 V3=1 V4=0 ******
873: | 65 1 0 2 1 3 1 4 0 0.96326 0.03674
874: freexexit2 possible for memory heap.
875:
876: h Pij x | pij_nom ficrestpij
877: # Cov Agex agex+h hpijx with i,j= 1-1 1-2 1-3 2-1 2-2 2-3
878: 1 85 85 1.00000 0.00000 0.00000 0.00000 1.00000 0.00000
879: 1 85 86 0.68299 0.22291 0.09410 0.71093 0.00000 0.28907
880:
881: 1 65 99 0.00364 0.00322 0.99314 0.00350 0.00310 0.99340
882: 1 65 100 0.00214 0.00204 0.99581 0.00206 0.00196 0.99597
883: variance of p one-step probabilities varprob | prob_nom ficresprob #One-step probabilities and stand. devi in ()
884: Standard deviation of one-step probabilities | probcor_nom ficresprobcor #One-step probabilities and correlation matrix
885: Matrix of variance covariance of one-step probabilities | probcov_nom ficresprobcov #One-step probabilities and covariance matrix
886:
1.126 brouard 887: forecasting if prevfcast==1 prevforecast call prevalence()
888: health expectancies
889: Variance-covariance of DFLE
890: prevalence()
891: movingaverage()
892: varevsij()
893: if popbased==1 varevsij(,popbased)
894: total life expectancies
895: Variance of period (stable) prevalence
896: end
897: */
898:
1.187 brouard 899: /* #define DEBUG */
900: /* #define DEBUGBRENT */
1.203 brouard 901: /* #define DEBUGLINMIN */
902: /* #define DEBUGHESS */
903: #define DEBUGHESSIJ
1.224 brouard 904: /* #define LINMINORIGINAL /\* Don't use loop on scale in linmin (accepting nan) *\/ */
1.165 brouard 905: #define POWELL /* Instead of NLOPT */
1.224 brouard 906: #define POWELLNOF3INFF1TEST /* Skip test */
1.186 brouard 907: /* #define POWELLORIGINAL /\* Don't use Directest to decide new direction but original Powell test *\/ */
908: /* #define MNBRAKORIGINAL /\* Don't use mnbrak fix *\/ */
1.126 brouard 909:
910: #include <math.h>
911: #include <stdio.h>
912: #include <stdlib.h>
913: #include <string.h>
1.226 brouard 914: #include <ctype.h>
1.159 brouard 915:
916: #ifdef _WIN32
917: #include <io.h>
1.172 brouard 918: #include <windows.h>
919: #include <tchar.h>
1.159 brouard 920: #else
1.126 brouard 921: #include <unistd.h>
1.159 brouard 922: #endif
1.126 brouard 923:
924: #include <limits.h>
925: #include <sys/types.h>
1.171 brouard 926:
927: #if defined(__GNUC__)
928: #include <sys/utsname.h> /* Doesn't work on Windows */
929: #endif
930:
1.126 brouard 931: #include <sys/stat.h>
932: #include <errno.h>
1.159 brouard 933: /* extern int errno; */
1.126 brouard 934:
1.157 brouard 935: /* #ifdef LINUX */
936: /* #include <time.h> */
937: /* #include "timeval.h" */
938: /* #else */
939: /* #include <sys/time.h> */
940: /* #endif */
941:
1.126 brouard 942: #include <time.h>
943:
1.136 brouard 944: #ifdef GSL
945: #include <gsl/gsl_errno.h>
946: #include <gsl/gsl_multimin.h>
947: #endif
948:
1.167 brouard 949:
1.162 brouard 950: #ifdef NLOPT
951: #include <nlopt.h>
952: typedef struct {
953: double (* function)(double [] );
954: } myfunc_data ;
955: #endif
956:
1.126 brouard 957: /* #include <libintl.h> */
958: /* #define _(String) gettext (String) */
959:
1.251 brouard 960: #define MAXLINE 2048 /* Was 256 and 1024. Overflow with 312 with 2 states and 4 covariates. Should be ok */
1.126 brouard 961:
962: #define GNUPLOTPROGRAM "gnuplot"
963: /*#define GNUPLOTPROGRAM "..\\gp37mgw\\wgnuplot"*/
964: #define FILENAMELENGTH 132
965:
966: #define GLOCK_ERROR_NOPATH -1 /* empty path */
967: #define GLOCK_ERROR_GETCWD -2 /* cannot get cwd */
968:
1.144 brouard 969: #define MAXPARM 128 /**< Maximum number of parameters for the optimization */
970: #define NPARMAX 64 /**< (nlstate+ndeath-1)*nlstate*ncovmodel */
1.126 brouard 971:
972: #define NINTERVMAX 8
1.144 brouard 973: #define NLSTATEMAX 8 /**< Maximum number of live states (for func) */
974: #define NDEATHMAX 8 /**< Maximum number of dead states (for func) */
975: #define NCOVMAX 20 /**< Maximum number of covariates, including generated covariates V1*V2 */
1.197 brouard 976: #define codtabm(h,k) (1 & (h-1) >> (k-1))+1
1.211 brouard 977: /*#define decodtabm(h,k,cptcoveff)= (h <= (1<<cptcoveff)?(((h-1) >> (k-1)) & 1) +1 : -1)*/
978: #define decodtabm(h,k,cptcoveff) (((h-1) >> (k-1)) & 1) +1
1.126 brouard 979: #define MAXN 20000
1.144 brouard 980: #define YEARM 12. /**< Number of months per year */
1.218 brouard 981: /* #define AGESUP 130 */
982: #define AGESUP 150
983: #define AGEMARGE 25 /* Marge for agemin and agemax for(iage=agemin-AGEMARGE; iage <= agemax+3+AGEMARGE; iage++) */
1.126 brouard 984: #define AGEBASE 40
1.194 brouard 985: #define AGEOVERFLOW 1.e20
1.164 brouard 986: #define AGEGOMP 10 /**< Minimal age for Gompertz adjustment */
1.157 brouard 987: #ifdef _WIN32
988: #define DIRSEPARATOR '\\'
989: #define CHARSEPARATOR "\\"
990: #define ODIRSEPARATOR '/'
991: #else
1.126 brouard 992: #define DIRSEPARATOR '/'
993: #define CHARSEPARATOR "/"
994: #define ODIRSEPARATOR '\\'
995: #endif
996:
1.265 ! brouard 997: /* $Id: imach.c,v 1.264 2017/04/26 06:01:29 brouard Exp $ */
1.126 brouard 998: /* $State: Exp $ */
1.196 brouard 999: #include "version.h"
1000: char version[]=__IMACH_VERSION__;
1.224 brouard 1001: char copyright[]="February 2016,INED-EUROREVES-Institut de longevite-Japan Society for the Promotion of Science (Grant-in-Aid for Scientific Research 25293121), Intel Software 2015-2018";
1.265 ! brouard 1002: char fullversion[]="$Revision: 1.264 $ $Date: 2017/04/26 06:01:29 $";
1.126 brouard 1003: char strstart[80];
1004: char optionfilext[10], optionfilefiname[FILENAMELENGTH];
1.130 brouard 1005: int erreur=0, nberr=0, nbwarn=0; /* Error number, number of errors number of warnings */
1.187 brouard 1006: int nagesqr=0, nforce=0; /* nagesqr=1 if model is including age*age, number of forces */
1.145 brouard 1007: /* Number of covariates model=V2+V1+ V3*age+V2*V4 */
1008: int cptcovn=0; /**< cptcovn number of covariates added in the model (excepting constant and age and age*product) */
1009: int cptcovt=0; /**< cptcovt number of covariates added in the model (excepting constant and age) */
1.225 brouard 1010: int cptcovs=0; /**< cptcovs number of simple covariates in the model V2+V1 =2 */
1011: int cptcovsnq=0; /**< cptcovsnq number of simple covariates in the model but non quantitative V2+V1 =2 */
1.145 brouard 1012: int cptcovage=0; /**< Number of covariates with age: V3*age only =1 */
1013: int cptcovprodnoage=0; /**< Number of covariate products without age */
1014: int cptcoveff=0; /* Total number of covariates to vary for printing results */
1.233 brouard 1015: int ncovf=0; /* Total number of effective fixed covariates (dummy or quantitative) in the model */
1016: int ncovv=0; /* Total number of effective (wave) varying covariates (dummy or quantitative) in the model */
1.232 brouard 1017: int ncova=0; /* Total number of effective (wave and stepm) varying with age covariates (dummy of quantitative) in the model */
1.234 brouard 1018: int nsd=0; /**< Total number of single dummy variables (output) */
1019: int nsq=0; /**< Total number of single quantitative variables (output) */
1.232 brouard 1020: int ncoveff=0; /* Total number of effective fixed dummy covariates in the model */
1.225 brouard 1021: int nqfveff=0; /**< nqfveff Number of Quantitative Fixed Variables Effective */
1.224 brouard 1022: int ntveff=0; /**< ntveff number of effective time varying variables */
1023: int nqtveff=0; /**< ntqveff number of effective time varying quantitative variables */
1.145 brouard 1024: int cptcov=0; /* Working variable */
1.218 brouard 1025: int ncovcombmax=NCOVMAX; /* Maximum calculated number of covariate combination = pow(2, cptcoveff) */
1.126 brouard 1026: int npar=NPARMAX;
1027: int nlstate=2; /* Number of live states */
1028: int ndeath=1; /* Number of dead states */
1.130 brouard 1029: int ncovmodel=0, ncovcol=0; /* Total number of covariables including constant a12*1 +b12*x ncovmodel=2 */
1.223 brouard 1030: int nqv=0, ntv=0, nqtv=0; /* Total number of quantitative variables, time variable (dummy), quantitative and time variable */
1.126 brouard 1031: int popbased=0;
1032:
1033: int *wav; /* Number of waves for this individuual 0 is possible */
1.130 brouard 1034: int maxwav=0; /* Maxim number of waves */
1035: int jmin=0, jmax=0; /* min, max spacing between 2 waves */
1036: int ijmin=0, ijmax=0; /* Individuals having jmin and jmax */
1037: int gipmx=0, gsw=0; /* Global variables on the number of contributions
1.126 brouard 1038: to the likelihood and the sum of weights (done by funcone)*/
1.130 brouard 1039: int mle=1, weightopt=0;
1.126 brouard 1040: int **mw; /* mw[mi][i] is number of the mi wave for this individual */
1041: int **dh; /* dh[mi][i] is number of steps between mi,mi+1 for this individual */
1042: int **bh; /* bh[mi][i] is the bias (+ or -) for this individual if the delay between
1043: * wave mi and wave mi+1 is not an exact multiple of stepm. */
1.162 brouard 1044: int countcallfunc=0; /* Count the number of calls to func */
1.230 brouard 1045: int selected(int kvar); /* Is covariate kvar selected for printing results */
1046:
1.130 brouard 1047: double jmean=1; /* Mean space between 2 waves */
1.145 brouard 1048: double **matprod2(); /* test */
1.126 brouard 1049: double **oldm, **newm, **savm; /* Working pointers to matrices */
1050: double **oldms, **newms, **savms; /* Fixed working pointers to matrices */
1.218 brouard 1051: double **ddnewms, **ddoldms, **ddsavms; /* for freeing later */
1052:
1.136 brouard 1053: /*FILE *fic ; */ /* Used in readdata only */
1.217 brouard 1054: FILE *ficpar, *ficparo,*ficres, *ficresp, *ficresphtm, *ficresphtmfr, *ficrespl, *ficresplb,*ficrespij, *ficrespijb, *ficrest,*ficresf, *ficresfb,*ficrespop;
1.126 brouard 1055: FILE *ficlog, *ficrespow;
1.130 brouard 1056: int globpr=0; /* Global variable for printing or not */
1.126 brouard 1057: double fretone; /* Only one call to likelihood */
1.130 brouard 1058: long ipmx=0; /* Number of contributions */
1.126 brouard 1059: double sw; /* Sum of weights */
1060: char filerespow[FILENAMELENGTH];
1061: char fileresilk[FILENAMELENGTH]; /* File of individual contributions to the likelihood */
1062: FILE *ficresilk;
1063: FILE *ficgp,*ficresprob,*ficpop, *ficresprobcov, *ficresprobcor;
1064: FILE *ficresprobmorprev;
1065: FILE *fichtm, *fichtmcov; /* Html File */
1066: FILE *ficreseij;
1067: char filerese[FILENAMELENGTH];
1068: FILE *ficresstdeij;
1069: char fileresstde[FILENAMELENGTH];
1070: FILE *ficrescveij;
1071: char filerescve[FILENAMELENGTH];
1072: FILE *ficresvij;
1073: char fileresv[FILENAMELENGTH];
1074: FILE *ficresvpl;
1075: char fileresvpl[FILENAMELENGTH];
1076: char title[MAXLINE];
1.234 brouard 1077: char model[MAXLINE]; /**< The model line */
1.217 brouard 1078: char optionfile[FILENAMELENGTH], datafile[FILENAMELENGTH], filerespl[FILENAMELENGTH], fileresplb[FILENAMELENGTH];
1.126 brouard 1079: char plotcmd[FILENAMELENGTH], pplotcmd[FILENAMELENGTH];
1080: char tmpout[FILENAMELENGTH], tmpout2[FILENAMELENGTH];
1081: char command[FILENAMELENGTH];
1082: int outcmd=0;
1083:
1.217 brouard 1084: char fileres[FILENAMELENGTH], filerespij[FILENAMELENGTH], filerespijb[FILENAMELENGTH], filereso[FILENAMELENGTH], rfileres[FILENAMELENGTH];
1.202 brouard 1085: char fileresu[FILENAMELENGTH]; /* fileres without r in front */
1.126 brouard 1086: char filelog[FILENAMELENGTH]; /* Log file */
1087: char filerest[FILENAMELENGTH];
1088: char fileregp[FILENAMELENGTH];
1089: char popfile[FILENAMELENGTH];
1090:
1091: char optionfilegnuplot[FILENAMELENGTH], optionfilehtm[FILENAMELENGTH], optionfilehtmcov[FILENAMELENGTH] ;
1092:
1.157 brouard 1093: /* struct timeval start_time, end_time, curr_time, last_time, forecast_time; */
1094: /* struct timezone tzp; */
1095: /* extern int gettimeofday(); */
1096: struct tm tml, *gmtime(), *localtime();
1097:
1098: extern time_t time();
1099:
1100: struct tm start_time, end_time, curr_time, last_time, forecast_time;
1101: time_t rstart_time, rend_time, rcurr_time, rlast_time, rforecast_time; /* raw time */
1102: struct tm tm;
1103:
1.126 brouard 1104: char strcurr[80], strfor[80];
1105:
1106: char *endptr;
1107: long lval;
1108: double dval;
1109:
1110: #define NR_END 1
1111: #define FREE_ARG char*
1112: #define FTOL 1.0e-10
1113:
1114: #define NRANSI
1.240 brouard 1115: #define ITMAX 200
1116: #define ITPOWMAX 20 /* This is now multiplied by the number of parameters */
1.126 brouard 1117:
1118: #define TOL 2.0e-4
1119:
1120: #define CGOLD 0.3819660
1121: #define ZEPS 1.0e-10
1122: #define SHFT(a,b,c,d) (a)=(b);(b)=(c);(c)=(d);
1123:
1124: #define GOLD 1.618034
1125: #define GLIMIT 100.0
1126: #define TINY 1.0e-20
1127:
1128: static double maxarg1,maxarg2;
1129: #define FMAX(a,b) (maxarg1=(a),maxarg2=(b),(maxarg1)>(maxarg2)? (maxarg1):(maxarg2))
1130: #define FMIN(a,b) (maxarg1=(a),maxarg2=(b),(maxarg1)<(maxarg2)? (maxarg1):(maxarg2))
1131:
1132: #define SIGN(a,b) ((b)>0.0 ? fabs(a) : -fabs(a))
1133: #define rint(a) floor(a+0.5)
1.166 brouard 1134: /* http://www.thphys.uni-heidelberg.de/~robbers/cmbeasy/doc/html/myutils_8h-source.html */
1.183 brouard 1135: #define mytinydouble 1.0e-16
1.166 brouard 1136: /* #define DEQUAL(a,b) (fabs((a)-(b))<mytinydouble) */
1137: /* http://www.thphys.uni-heidelberg.de/~robbers/cmbeasy/doc/html/mynrutils_8h-source.html */
1138: /* static double dsqrarg; */
1139: /* #define DSQR(a) (DEQUAL((dsqrarg=(a)),0.0) ? 0.0 : dsqrarg*dsqrarg) */
1.126 brouard 1140: static double sqrarg;
1141: #define SQR(a) ((sqrarg=(a)) == 0.0 ? 0.0 :sqrarg*sqrarg)
1142: #define SWAP(a,b) {temp=(a);(a)=(b);(b)=temp;}
1143: int agegomp= AGEGOMP;
1144:
1145: int imx;
1146: int stepm=1;
1147: /* Stepm, step in month: minimum step interpolation*/
1148:
1149: int estepm;
1150: /* Estepm, step in month to interpolate survival function in order to approximate Life Expectancy*/
1151:
1152: int m,nb;
1153: long *num;
1.197 brouard 1154: int firstpass=0, lastpass=4,*cod, *cens;
1.192 brouard 1155: int *ncodemax; /* ncodemax[j]= Number of modalities of the j th
1156: covariate for which somebody answered excluding
1157: undefined. Usually 2: 0 and 1. */
1158: int *ncodemaxwundef; /* ncodemax[j]= Number of modalities of the j th
1159: covariate for which somebody answered including
1160: undefined. Usually 3: -1, 0 and 1. */
1.126 brouard 1161: double **agev,*moisnais, *annais, *moisdc, *andc,**mint, **anint;
1.218 brouard 1162: double **pmmij, ***probs; /* Global pointer */
1.219 brouard 1163: double ***mobaverage, ***mobaverages; /* New global variable */
1.126 brouard 1164: double *ageexmed,*agecens;
1165: double dateintmean=0;
1166:
1167: double *weight;
1168: int **s; /* Status */
1.141 brouard 1169: double *agedc;
1.145 brouard 1170: double **covar; /**< covar[j,i], value of jth covariate for individual i,
1.141 brouard 1171: * covar=matrix(0,NCOVMAX,1,n);
1.187 brouard 1172: * cov[Tage[kk]+2]=covar[Tvar[Tage[kk]]][i]*age; */
1.225 brouard 1173: double **coqvar; /* Fixed quantitative covariate iqv */
1174: double ***cotvar; /* Time varying covariate itv */
1175: double ***cotqvar; /* Time varying quantitative covariate itqv */
1.141 brouard 1176: double idx;
1177: int **nbcode, *Tvar; /**< model=V2 => Tvar[1]= 2 */
1.234 brouard 1178: /* V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1179: /*k 1 2 3 4 5 6 7 8 9 */
1180: /*Tvar[k]= 5 4 3 6 5 2 7 1 1 */
1181: /* Tndvar[k] 1 2 3 4 5 */
1182: /*TDvar 4 3 6 7 1 */ /* For outputs only; combination of dummies fixed or varying */
1183: /* Tns[k] 1 2 2 4 5 */ /* Number of single cova */
1184: /* TvarsD[k] 1 2 3 */ /* Number of single dummy cova */
1185: /* TvarsDind 2 3 9 */ /* position K of single dummy cova */
1186: /* TvarsQ[k] 1 2 */ /* Number of single quantitative cova */
1187: /* TvarsQind 1 6 */ /* position K of single quantitative cova */
1188: /* Tprod[i]=k 4 7 */
1189: /* Tage[i]=k 5 8 */
1190: /* */
1191: /* Type */
1192: /* V 1 2 3 4 5 */
1193: /* F F V V V */
1194: /* D Q D D Q */
1195: /* */
1196: int *TvarsD;
1197: int *TvarsDind;
1198: int *TvarsQ;
1199: int *TvarsQind;
1200:
1.235 brouard 1201: #define MAXRESULTLINES 10
1202: int nresult=0;
1.258 brouard 1203: int parameterline=0; /* # of the parameter (type) line */
1.235 brouard 1204: int TKresult[MAXRESULTLINES];
1.237 brouard 1205: int Tresult[MAXRESULTLINES][NCOVMAX];/* For dummy variable , value (output) */
1206: int Tinvresult[MAXRESULTLINES][NCOVMAX];/* For dummy variable , value (output) */
1.235 brouard 1207: int Tvresult[MAXRESULTLINES][NCOVMAX]; /* For dummy variable , variable # (output) */
1208: double Tqresult[MAXRESULTLINES][NCOVMAX]; /* For quantitative variable , value (output) */
1.237 brouard 1209: double Tqinvresult[MAXRESULTLINES][NCOVMAX]; /* For quantitative variable , value (output) */
1.235 brouard 1210: int Tvqresult[MAXRESULTLINES][NCOVMAX]; /* For quantitative variable , variable # (output) */
1211:
1.234 brouard 1212: /* int *TDvar; /\**< TDvar[1]=4, TDvarF[2]=3, TDvar[3]=6 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 *\/ */
1.232 brouard 1213: int *TvarF; /**< TvarF[1]=Tvar[6]=2, TvarF[2]=Tvar[7]=7, TvarF[3]=Tvar[9]=1 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1214: int *TvarFind; /**< TvarFind[1]=6, TvarFind[2]=7, Tvarind[3]=9 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1215: int *TvarV; /**< TvarV[1]=Tvar[1]=5, TvarV[2]=Tvar[2]=4 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1216: int *TvarVind; /**< TvarVind[1]=1, TvarVind[2]=2 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1217: int *TvarA; /**< TvarA[1]=Tvar[5]=5, TvarA[2]=Tvar[8]=1 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1218: int *TvarAind; /**< TvarindA[1]=5, TvarAind[2]=8 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1.231 brouard 1219: int *TvarFD; /**< TvarFD[1]=V1 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1220: int *TvarFDind; /* TvarFDind[1]=9 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1221: int *TvarFQ; /* TvarFQ[1]=V2 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple fixed quantitative variable */
1222: int *TvarFQind; /* TvarFQind[1]=6 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple fixed quantitative variable */
1223: int *TvarVD; /* TvarVD[1]=V5 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple fixed quantitative variable */
1224: int *TvarVDind; /* TvarVDind[1]=1 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple fixed quantitative variable */
1225: int *TvarVQ; /* TvarVQ[1]=V5 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple time varying quantitative variable */
1226: int *TvarVQind; /* TvarVQind[1]=1 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple time varying quantitative variable */
1227:
1.230 brouard 1228: int *Tvarsel; /**< Selected covariates for output */
1229: double *Tvalsel; /**< Selected modality value of covariate for output */
1.226 brouard 1230: int *Typevar; /**< 0 for simple covariate (dummy, quantitative, fixed or varying), 1 for age product, 2 for product */
1.227 brouard 1231: int *Fixed; /** Fixed[k] 0=fixed, 1 varying, 2 fixed with age product, 3 varying with age product */
1232: int *Dummy; /** Dummy[k] 0=dummy (0 1), 1 quantitative (single or product without age), 2 dummy with age product, 3 quant with age product */
1.238 brouard 1233: int *DummyV; /** Dummy[v] 0=dummy (0 1), 1 quantitative */
1234: int *FixedV; /** FixedV[v] 0 fixed, 1 varying */
1.197 brouard 1235: int *Tage;
1.227 brouard 1236: int anyvaryingduminmodel=0; /**< Any varying dummy in Model=1 yes, 0 no, to avoid a loop on waves in freq */
1.228 brouard 1237: int *Tmodelind; /** Tmodelind[Tvaraff[3]]=9 for V1 position,Tvaraff[1]@9={4, 3, 1, 0, 0, 0, 0, 0, 0}, model=V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1*/
1.230 brouard 1238: int *TmodelInvind; /** Tmodelind[Tvaraff[3]]=9 for V1 position,Tvaraff[1]@9={4, 3, 1, 0, 0, 0, 0, 0, 0}, model=V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1*/
1239: int *TmodelInvQind; /** Tmodelqind[1]=1 for V5(quantitative varying) position,Tvaraff[1]@9={4, 3, 1, 0, 0, 0, 0, 0, 0}, model=V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1.145 brouard 1240: int *Ndum; /** Freq of modality (tricode */
1.200 brouard 1241: /* int **codtab;*/ /**< codtab=imatrix(1,100,1,10); */
1.227 brouard 1242: int **Tvard;
1243: int *Tprod;/**< Gives the k position of the k1 product */
1.238 brouard 1244: /* Tprod[k1=1]=3(=V1*V4) for V2+V1+V1*V4+age*V3 */
1.227 brouard 1245: int *Tposprod; /**< Gives the k1 product from the k position */
1.238 brouard 1246: /* if V2+V1+V1*V4+age*V3+V3*V2 TProd[k1=2]=5 (V3*V2) */
1247: /* Tposprod[k]=k1 , Tposprod[3]=1, Tposprod[5(V3*V2)]=2 (2nd product without age) */
1.227 brouard 1248: int cptcovprod, *Tvaraff, *invalidvarcomb;
1.126 brouard 1249: double *lsurv, *lpop, *tpop;
1250:
1.231 brouard 1251: #define FD 1; /* Fixed dummy covariate */
1252: #define FQ 2; /* Fixed quantitative covariate */
1253: #define FP 3; /* Fixed product covariate */
1254: #define FPDD 7; /* Fixed product dummy*dummy covariate */
1255: #define FPDQ 8; /* Fixed product dummy*quantitative covariate */
1256: #define FPQQ 9; /* Fixed product quantitative*quantitative covariate */
1257: #define VD 10; /* Varying dummy covariate */
1258: #define VQ 11; /* Varying quantitative covariate */
1259: #define VP 12; /* Varying product covariate */
1260: #define VPDD 13; /* Varying product dummy*dummy covariate */
1261: #define VPDQ 14; /* Varying product dummy*quantitative covariate */
1262: #define VPQQ 15; /* Varying product quantitative*quantitative covariate */
1263: #define APFD 16; /* Age product * fixed dummy covariate */
1264: #define APFQ 17; /* Age product * fixed quantitative covariate */
1265: #define APVD 18; /* Age product * varying dummy covariate */
1266: #define APVQ 19; /* Age product * varying quantitative covariate */
1267:
1268: #define FTYPE 1; /* Fixed covariate */
1269: #define VTYPE 2; /* Varying covariate (loop in wave) */
1270: #define ATYPE 2; /* Age product covariate (loop in dh within wave)*/
1271:
1272: struct kmodel{
1273: int maintype; /* main type */
1274: int subtype; /* subtype */
1275: };
1276: struct kmodel modell[NCOVMAX];
1277:
1.143 brouard 1278: double ftol=FTOL; /**< Tolerance for computing Max Likelihood */
1279: double ftolhess; /**< Tolerance for computing hessian */
1.126 brouard 1280:
1281: /**************** split *************************/
1282: static int split( char *path, char *dirc, char *name, char *ext, char *finame )
1283: {
1284: /* From a file name with (full) path (either Unix or Windows) we extract the directory (dirc)
1285: the name of the file (name), its extension only (ext) and its first part of the name (finame)
1286: */
1287: char *ss; /* pointer */
1.186 brouard 1288: int l1=0, l2=0; /* length counters */
1.126 brouard 1289:
1290: l1 = strlen(path ); /* length of path */
1291: if ( l1 == 0 ) return( GLOCK_ERROR_NOPATH );
1292: ss= strrchr( path, DIRSEPARATOR ); /* find last / */
1293: if ( ss == NULL ) { /* no directory, so determine current directory */
1294: strcpy( name, path ); /* we got the fullname name because no directory */
1295: /*if(strrchr(path, ODIRSEPARATOR )==NULL)
1296: printf("Warning you should use %s as a separator\n",DIRSEPARATOR);*/
1297: /* get current working directory */
1298: /* extern char* getcwd ( char *buf , int len);*/
1.184 brouard 1299: #ifdef WIN32
1300: if (_getcwd( dirc, FILENAME_MAX ) == NULL ) {
1301: #else
1302: if (getcwd(dirc, FILENAME_MAX) == NULL) {
1303: #endif
1.126 brouard 1304: return( GLOCK_ERROR_GETCWD );
1305: }
1306: /* got dirc from getcwd*/
1307: printf(" DIRC = %s \n",dirc);
1.205 brouard 1308: } else { /* strip directory from path */
1.126 brouard 1309: ss++; /* after this, the filename */
1310: l2 = strlen( ss ); /* length of filename */
1311: if ( l2 == 0 ) return( GLOCK_ERROR_NOPATH );
1312: strcpy( name, ss ); /* save file name */
1313: strncpy( dirc, path, l1 - l2 ); /* now the directory */
1.186 brouard 1314: dirc[l1-l2] = '\0'; /* add zero */
1.126 brouard 1315: printf(" DIRC2 = %s \n",dirc);
1316: }
1317: /* We add a separator at the end of dirc if not exists */
1318: l1 = strlen( dirc ); /* length of directory */
1319: if( dirc[l1-1] != DIRSEPARATOR ){
1320: dirc[l1] = DIRSEPARATOR;
1321: dirc[l1+1] = 0;
1322: printf(" DIRC3 = %s \n",dirc);
1323: }
1324: ss = strrchr( name, '.' ); /* find last / */
1325: if (ss >0){
1326: ss++;
1327: strcpy(ext,ss); /* save extension */
1328: l1= strlen( name);
1329: l2= strlen(ss)+1;
1330: strncpy( finame, name, l1-l2);
1331: finame[l1-l2]= 0;
1332: }
1333:
1334: return( 0 ); /* we're done */
1335: }
1336:
1337:
1338: /******************************************/
1339:
1340: void replace_back_to_slash(char *s, char*t)
1341: {
1342: int i;
1343: int lg=0;
1344: i=0;
1345: lg=strlen(t);
1346: for(i=0; i<= lg; i++) {
1347: (s[i] = t[i]);
1348: if (t[i]== '\\') s[i]='/';
1349: }
1350: }
1351:
1.132 brouard 1352: char *trimbb(char *out, char *in)
1.137 brouard 1353: { /* Trim multiple blanks in line but keeps first blanks if line starts with blanks */
1.132 brouard 1354: char *s;
1355: s=out;
1356: while (*in != '\0'){
1.137 brouard 1357: while( *in == ' ' && *(in+1) == ' '){ /* && *(in+1) != '\0'){*/
1.132 brouard 1358: in++;
1359: }
1360: *out++ = *in++;
1361: }
1362: *out='\0';
1363: return s;
1364: }
1365:
1.187 brouard 1366: /* char *substrchaine(char *out, char *in, char *chain) */
1367: /* { */
1368: /* /\* Substract chain 'chain' from 'in', return and output 'out' *\/ */
1369: /* char *s, *t; */
1370: /* t=in;s=out; */
1371: /* while ((*in != *chain) && (*in != '\0')){ */
1372: /* *out++ = *in++; */
1373: /* } */
1374:
1375: /* /\* *in matches *chain *\/ */
1376: /* while ((*in++ == *chain++) && (*in != '\0')){ */
1377: /* printf("*in = %c, *out= %c *chain= %c \n", *in, *out, *chain); */
1378: /* } */
1379: /* in--; chain--; */
1380: /* while ( (*in != '\0')){ */
1381: /* printf("Bef *in = %c, *out= %c *chain= %c \n", *in, *out, *chain); */
1382: /* *out++ = *in++; */
1383: /* printf("Aft *in = %c, *out= %c *chain= %c \n", *in, *out, *chain); */
1384: /* } */
1385: /* *out='\0'; */
1386: /* out=s; */
1387: /* return out; */
1388: /* } */
1389: char *substrchaine(char *out, char *in, char *chain)
1390: {
1391: /* Substract chain 'chain' from 'in', return and output 'out' */
1392: /* in="V1+V1*age+age*age+V2", chain="age*age" */
1393:
1394: char *strloc;
1395:
1396: strcpy (out, in);
1397: strloc = strstr(out, chain); /* strloc points to out at age*age+V2 */
1398: printf("Bef strloc=%s chain=%s out=%s \n", strloc, chain, out);
1399: if(strloc != NULL){
1400: /* will affect out */ /* strloc+strlenc(chain)=+V2 */ /* Will also work in Unicode */
1401: memmove(strloc,strloc+strlen(chain), strlen(strloc+strlen(chain))+1);
1402: /* strcpy (strloc, strloc +strlen(chain));*/
1403: }
1404: printf("Aft strloc=%s chain=%s in=%s out=%s \n", strloc, chain, in, out);
1405: return out;
1406: }
1407:
1408:
1.145 brouard 1409: char *cutl(char *blocc, char *alocc, char *in, char occ)
1410: {
1.187 brouard 1411: /* cuts string in into blocc and alocc where blocc ends before FIRST occurence of char 'occ'
1.145 brouard 1412: and alocc starts after first occurence of char 'occ' : ex cutv(blocc,alocc,"abcdef2ghi2j",'2')
1.187 brouard 1413: gives blocc="abcdef" and alocc="ghi2j".
1.145 brouard 1414: If occ is not found blocc is null and alocc is equal to in. Returns blocc
1415: */
1.160 brouard 1416: char *s, *t;
1.145 brouard 1417: t=in;s=in;
1418: while ((*in != occ) && (*in != '\0')){
1419: *alocc++ = *in++;
1420: }
1421: if( *in == occ){
1422: *(alocc)='\0';
1423: s=++in;
1424: }
1425:
1426: if (s == t) {/* occ not found */
1427: *(alocc-(in-s))='\0';
1428: in=s;
1429: }
1430: while ( *in != '\0'){
1431: *blocc++ = *in++;
1432: }
1433:
1434: *blocc='\0';
1435: return t;
1436: }
1.137 brouard 1437: char *cutv(char *blocc, char *alocc, char *in, char occ)
1438: {
1.187 brouard 1439: /* cuts string in into blocc and alocc where blocc ends before LAST occurence of char 'occ'
1.137 brouard 1440: and alocc starts after last occurence of char 'occ' : ex cutv(blocc,alocc,"abcdef2ghi2j",'2')
1441: gives blocc="abcdef2ghi" and alocc="j".
1442: If occ is not found blocc is null and alocc is equal to in. Returns alocc
1443: */
1444: char *s, *t;
1445: t=in;s=in;
1446: while (*in != '\0'){
1447: while( *in == occ){
1448: *blocc++ = *in++;
1449: s=in;
1450: }
1451: *blocc++ = *in++;
1452: }
1453: if (s == t) /* occ not found */
1454: *(blocc-(in-s))='\0';
1455: else
1456: *(blocc-(in-s)-1)='\0';
1457: in=s;
1458: while ( *in != '\0'){
1459: *alocc++ = *in++;
1460: }
1461:
1462: *alocc='\0';
1463: return s;
1464: }
1465:
1.126 brouard 1466: int nbocc(char *s, char occ)
1467: {
1468: int i,j=0;
1469: int lg=20;
1470: i=0;
1471: lg=strlen(s);
1472: for(i=0; i<= lg; i++) {
1.234 brouard 1473: if (s[i] == occ ) j++;
1.126 brouard 1474: }
1475: return j;
1476: }
1477:
1.137 brouard 1478: /* void cutv(char *u,char *v, char*t, char occ) */
1479: /* { */
1480: /* /\* cuts string t into u and v where u ends before last occurence of char 'occ' */
1481: /* and v starts after last occurence of char 'occ' : ex cutv(u,v,"abcdef2ghi2j",'2') */
1482: /* gives u="abcdef2ghi" and v="j" *\/ */
1483: /* int i,lg,j,p=0; */
1484: /* i=0; */
1485: /* lg=strlen(t); */
1486: /* for(j=0; j<=lg-1; j++) { */
1487: /* if((t[j]!= occ) && (t[j+1]== occ)) p=j+1; */
1488: /* } */
1.126 brouard 1489:
1.137 brouard 1490: /* for(j=0; j<p; j++) { */
1491: /* (u[j] = t[j]); */
1492: /* } */
1493: /* u[p]='\0'; */
1.126 brouard 1494:
1.137 brouard 1495: /* for(j=0; j<= lg; j++) { */
1496: /* if (j>=(p+1))(v[j-p-1] = t[j]); */
1497: /* } */
1498: /* } */
1.126 brouard 1499:
1.160 brouard 1500: #ifdef _WIN32
1501: char * strsep(char **pp, const char *delim)
1502: {
1503: char *p, *q;
1504:
1505: if ((p = *pp) == NULL)
1506: return 0;
1507: if ((q = strpbrk (p, delim)) != NULL)
1508: {
1509: *pp = q + 1;
1510: *q = '\0';
1511: }
1512: else
1513: *pp = 0;
1514: return p;
1515: }
1516: #endif
1517:
1.126 brouard 1518: /********************** nrerror ********************/
1519:
1520: void nrerror(char error_text[])
1521: {
1522: fprintf(stderr,"ERREUR ...\n");
1523: fprintf(stderr,"%s\n",error_text);
1524: exit(EXIT_FAILURE);
1525: }
1526: /*********************** vector *******************/
1527: double *vector(int nl, int nh)
1528: {
1529: double *v;
1530: v=(double *) malloc((size_t)((nh-nl+1+NR_END)*sizeof(double)));
1531: if (!v) nrerror("allocation failure in vector");
1532: return v-nl+NR_END;
1533: }
1534:
1535: /************************ free vector ******************/
1536: void free_vector(double*v, int nl, int nh)
1537: {
1538: free((FREE_ARG)(v+nl-NR_END));
1539: }
1540:
1541: /************************ivector *******************************/
1542: int *ivector(long nl,long nh)
1543: {
1544: int *v;
1545: v=(int *) malloc((size_t)((nh-nl+1+NR_END)*sizeof(int)));
1546: if (!v) nrerror("allocation failure in ivector");
1547: return v-nl+NR_END;
1548: }
1549:
1550: /******************free ivector **************************/
1551: void free_ivector(int *v, long nl, long nh)
1552: {
1553: free((FREE_ARG)(v+nl-NR_END));
1554: }
1555:
1556: /************************lvector *******************************/
1557: long *lvector(long nl,long nh)
1558: {
1559: long *v;
1560: v=(long *) malloc((size_t)((nh-nl+1+NR_END)*sizeof(long)));
1561: if (!v) nrerror("allocation failure in ivector");
1562: return v-nl+NR_END;
1563: }
1564:
1565: /******************free lvector **************************/
1566: void free_lvector(long *v, long nl, long nh)
1567: {
1568: free((FREE_ARG)(v+nl-NR_END));
1569: }
1570:
1571: /******************* imatrix *******************************/
1572: int **imatrix(long nrl, long nrh, long ncl, long nch)
1573: /* allocate a int matrix with subscript range m[nrl..nrh][ncl..nch] */
1574: {
1575: long i, nrow=nrh-nrl+1,ncol=nch-ncl+1;
1576: int **m;
1577:
1578: /* allocate pointers to rows */
1579: m=(int **) malloc((size_t)((nrow+NR_END)*sizeof(int*)));
1580: if (!m) nrerror("allocation failure 1 in matrix()");
1581: m += NR_END;
1582: m -= nrl;
1583:
1584:
1585: /* allocate rows and set pointers to them */
1586: m[nrl]=(int *) malloc((size_t)((nrow*ncol+NR_END)*sizeof(int)));
1587: if (!m[nrl]) nrerror("allocation failure 2 in matrix()");
1588: m[nrl] += NR_END;
1589: m[nrl] -= ncl;
1590:
1591: for(i=nrl+1;i<=nrh;i++) m[i]=m[i-1]+ncol;
1592:
1593: /* return pointer to array of pointers to rows */
1594: return m;
1595: }
1596:
1597: /****************** free_imatrix *************************/
1598: void free_imatrix(m,nrl,nrh,ncl,nch)
1599: int **m;
1600: long nch,ncl,nrh,nrl;
1601: /* free an int matrix allocated by imatrix() */
1602: {
1603: free((FREE_ARG) (m[nrl]+ncl-NR_END));
1604: free((FREE_ARG) (m+nrl-NR_END));
1605: }
1606:
1607: /******************* matrix *******************************/
1608: double **matrix(long nrl, long nrh, long ncl, long nch)
1609: {
1610: long i, nrow=nrh-nrl+1, ncol=nch-ncl+1;
1611: double **m;
1612:
1613: m=(double **) malloc((size_t)((nrow+NR_END)*sizeof(double*)));
1614: if (!m) nrerror("allocation failure 1 in matrix()");
1615: m += NR_END;
1616: m -= nrl;
1617:
1618: m[nrl]=(double *) malloc((size_t)((nrow*ncol+NR_END)*sizeof(double)));
1619: if (!m[nrl]) nrerror("allocation failure 2 in matrix()");
1620: m[nrl] += NR_END;
1621: m[nrl] -= ncl;
1622:
1623: for (i=nrl+1; i<=nrh; i++) m[i]=m[i-1]+ncol;
1624: return m;
1.145 brouard 1625: /* print *(*(m+1)+70) or print m[1][70]; print m+1 or print &(m[1]) or &(m[1][0])
1626: m[i] = address of ith row of the table. &(m[i]) is its value which is another adress
1627: that of m[i][0]. In order to get the value p m[i][0] but it is unitialized.
1.126 brouard 1628: */
1629: }
1630:
1631: /*************************free matrix ************************/
1632: void free_matrix(double **m, long nrl, long nrh, long ncl, long nch)
1633: {
1634: free((FREE_ARG)(m[nrl]+ncl-NR_END));
1635: free((FREE_ARG)(m+nrl-NR_END));
1636: }
1637:
1638: /******************* ma3x *******************************/
1639: double ***ma3x(long nrl, long nrh, long ncl, long nch, long nll, long nlh)
1640: {
1641: long i, j, nrow=nrh-nrl+1, ncol=nch-ncl+1, nlay=nlh-nll+1;
1642: double ***m;
1643:
1644: m=(double ***) malloc((size_t)((nrow+NR_END)*sizeof(double*)));
1645: if (!m) nrerror("allocation failure 1 in matrix()");
1646: m += NR_END;
1647: m -= nrl;
1648:
1649: m[nrl]=(double **) malloc((size_t)((nrow*ncol+NR_END)*sizeof(double)));
1650: if (!m[nrl]) nrerror("allocation failure 2 in matrix()");
1651: m[nrl] += NR_END;
1652: m[nrl] -= ncl;
1653:
1654: for (i=nrl+1; i<=nrh; i++) m[i]=m[i-1]+ncol;
1655:
1656: m[nrl][ncl]=(double *) malloc((size_t)((nrow*ncol*nlay+NR_END)*sizeof(double)));
1657: if (!m[nrl][ncl]) nrerror("allocation failure 3 in matrix()");
1658: m[nrl][ncl] += NR_END;
1659: m[nrl][ncl] -= nll;
1660: for (j=ncl+1; j<=nch; j++)
1661: m[nrl][j]=m[nrl][j-1]+nlay;
1662:
1663: for (i=nrl+1; i<=nrh; i++) {
1664: m[i][ncl]=m[i-1l][ncl]+ncol*nlay;
1665: for (j=ncl+1; j<=nch; j++)
1666: m[i][j]=m[i][j-1]+nlay;
1667: }
1668: return m;
1669: /* gdb: p *(m+1) <=> p m[1] and p (m+1) <=> p (m+1) <=> p &(m[1])
1670: &(m[i][j][k]) <=> *((*(m+i) + j)+k)
1671: */
1672: }
1673:
1674: /*************************free ma3x ************************/
1675: void free_ma3x(double ***m, long nrl, long nrh, long ncl, long nch,long nll, long nlh)
1676: {
1677: free((FREE_ARG)(m[nrl][ncl]+ nll-NR_END));
1678: free((FREE_ARG)(m[nrl]+ncl-NR_END));
1679: free((FREE_ARG)(m+nrl-NR_END));
1680: }
1681:
1682: /*************** function subdirf ***********/
1683: char *subdirf(char fileres[])
1684: {
1685: /* Caution optionfilefiname is hidden */
1686: strcpy(tmpout,optionfilefiname);
1687: strcat(tmpout,"/"); /* Add to the right */
1688: strcat(tmpout,fileres);
1689: return tmpout;
1690: }
1691:
1692: /*************** function subdirf2 ***********/
1693: char *subdirf2(char fileres[], char *preop)
1694: {
1695:
1696: /* Caution optionfilefiname is hidden */
1697: strcpy(tmpout,optionfilefiname);
1698: strcat(tmpout,"/");
1699: strcat(tmpout,preop);
1700: strcat(tmpout,fileres);
1701: return tmpout;
1702: }
1703:
1704: /*************** function subdirf3 ***********/
1705: char *subdirf3(char fileres[], char *preop, char *preop2)
1706: {
1707:
1708: /* Caution optionfilefiname is hidden */
1709: strcpy(tmpout,optionfilefiname);
1710: strcat(tmpout,"/");
1711: strcat(tmpout,preop);
1712: strcat(tmpout,preop2);
1713: strcat(tmpout,fileres);
1714: return tmpout;
1715: }
1.213 brouard 1716:
1717: /*************** function subdirfext ***********/
1718: char *subdirfext(char fileres[], char *preop, char *postop)
1719: {
1720:
1721: strcpy(tmpout,preop);
1722: strcat(tmpout,fileres);
1723: strcat(tmpout,postop);
1724: return tmpout;
1725: }
1.126 brouard 1726:
1.213 brouard 1727: /*************** function subdirfext3 ***********/
1728: char *subdirfext3(char fileres[], char *preop, char *postop)
1729: {
1730:
1731: /* Caution optionfilefiname is hidden */
1732: strcpy(tmpout,optionfilefiname);
1733: strcat(tmpout,"/");
1734: strcat(tmpout,preop);
1735: strcat(tmpout,fileres);
1736: strcat(tmpout,postop);
1737: return tmpout;
1738: }
1739:
1.162 brouard 1740: char *asc_diff_time(long time_sec, char ascdiff[])
1741: {
1742: long sec_left, days, hours, minutes;
1743: days = (time_sec) / (60*60*24);
1744: sec_left = (time_sec) % (60*60*24);
1745: hours = (sec_left) / (60*60) ;
1746: sec_left = (sec_left) %(60*60);
1747: minutes = (sec_left) /60;
1748: sec_left = (sec_left) % (60);
1749: sprintf(ascdiff,"%ld day(s) %ld hour(s) %ld minute(s) %ld second(s)",days, hours, minutes, sec_left);
1750: return ascdiff;
1751: }
1752:
1.126 brouard 1753: /***************** f1dim *************************/
1754: extern int ncom;
1755: extern double *pcom,*xicom;
1756: extern double (*nrfunc)(double []);
1757:
1758: double f1dim(double x)
1759: {
1760: int j;
1761: double f;
1762: double *xt;
1763:
1764: xt=vector(1,ncom);
1765: for (j=1;j<=ncom;j++) xt[j]=pcom[j]+x*xicom[j];
1766: f=(*nrfunc)(xt);
1767: free_vector(xt,1,ncom);
1768: return f;
1769: }
1770:
1771: /*****************brent *************************/
1772: double brent(double ax, double bx, double cx, double (*f)(double), double tol, double *xmin)
1.187 brouard 1773: {
1774: /* Given a function f, and given a bracketing triplet of abscissas ax, bx, cx (such that bx is
1775: * between ax and cx, and f(bx) is less than both f(ax) and f(cx) ), this routine isolates
1776: * the minimum to a fractional precision of about tol using Brent’s method. The abscissa of
1777: * the minimum is returned as xmin, and the minimum function value is returned as brent , the
1778: * returned function value.
1779: */
1.126 brouard 1780: int iter;
1781: double a,b,d,etemp;
1.159 brouard 1782: double fu=0,fv,fw,fx;
1.164 brouard 1783: double ftemp=0.;
1.126 brouard 1784: double p,q,r,tol1,tol2,u,v,w,x,xm;
1785: double e=0.0;
1786:
1787: a=(ax < cx ? ax : cx);
1788: b=(ax > cx ? ax : cx);
1789: x=w=v=bx;
1790: fw=fv=fx=(*f)(x);
1791: for (iter=1;iter<=ITMAX;iter++) {
1792: xm=0.5*(a+b);
1793: tol2=2.0*(tol1=tol*fabs(x)+ZEPS);
1794: /* if (2.0*fabs(fp-(*fret)) <= ftol*(fabs(fp)+fabs(*fret)))*/
1795: printf(".");fflush(stdout);
1796: fprintf(ficlog,".");fflush(ficlog);
1.162 brouard 1797: #ifdef DEBUGBRENT
1.126 brouard 1798: printf("br %d,x=%.10e xm=%.10e b=%.10e a=%.10e tol=%.10e tol1=%.10e tol2=%.10e x-xm=%.10e fx=%.12e fu=%.12e,fw=%.12e,ftemp=%.12e,ftol=%.12e\n",iter,x,xm,b,a,tol,tol1,tol2,(x-xm),fx,fu,fw,ftemp,ftol);
1799: fprintf(ficlog,"br %d,x=%.10e xm=%.10e b=%.10e a=%.10e tol=%.10e tol1=%.10e tol2=%.10e x-xm=%.10e fx=%.12e fu=%.12e,fw=%.12e,ftemp=%.12e,ftol=%.12e\n",iter,x,xm,b,a,tol,tol1,tol2,(x-xm),fx,fu,fw,ftemp,ftol);
1800: /* if ((fabs(x-xm) <= (tol2-0.5*(b-a)))||(2.0*fabs(fu-ftemp) <= ftol*1.e-2*(fabs(fu)+fabs(ftemp)))) { */
1801: #endif
1802: if (fabs(x-xm) <= (tol2-0.5*(b-a))){
1803: *xmin=x;
1804: return fx;
1805: }
1806: ftemp=fu;
1807: if (fabs(e) > tol1) {
1808: r=(x-w)*(fx-fv);
1809: q=(x-v)*(fx-fw);
1810: p=(x-v)*q-(x-w)*r;
1811: q=2.0*(q-r);
1812: if (q > 0.0) p = -p;
1813: q=fabs(q);
1814: etemp=e;
1815: e=d;
1816: if (fabs(p) >= fabs(0.5*q*etemp) || p <= q*(a-x) || p >= q*(b-x))
1.224 brouard 1817: d=CGOLD*(e=(x >= xm ? a-x : b-x));
1.126 brouard 1818: else {
1.224 brouard 1819: d=p/q;
1820: u=x+d;
1821: if (u-a < tol2 || b-u < tol2)
1822: d=SIGN(tol1,xm-x);
1.126 brouard 1823: }
1824: } else {
1825: d=CGOLD*(e=(x >= xm ? a-x : b-x));
1826: }
1827: u=(fabs(d) >= tol1 ? x+d : x+SIGN(tol1,d));
1828: fu=(*f)(u);
1829: if (fu <= fx) {
1830: if (u >= x) a=x; else b=x;
1831: SHFT(v,w,x,u)
1.183 brouard 1832: SHFT(fv,fw,fx,fu)
1833: } else {
1834: if (u < x) a=u; else b=u;
1835: if (fu <= fw || w == x) {
1.224 brouard 1836: v=w;
1837: w=u;
1838: fv=fw;
1839: fw=fu;
1.183 brouard 1840: } else if (fu <= fv || v == x || v == w) {
1.224 brouard 1841: v=u;
1842: fv=fu;
1.183 brouard 1843: }
1844: }
1.126 brouard 1845: }
1846: nrerror("Too many iterations in brent");
1847: *xmin=x;
1848: return fx;
1849: }
1850:
1851: /****************** mnbrak ***********************/
1852:
1853: void mnbrak(double *ax, double *bx, double *cx, double *fa, double *fb, double *fc,
1854: double (*func)(double))
1.183 brouard 1855: { /* Given a function func , and given distinct initial points ax and bx , this routine searches in
1856: the downhill direction (defined by the function as evaluated at the initial points) and returns
1857: new points ax , bx , cx that bracket a minimum of the function. Also returned are the function
1858: values at the three points, fa, fb , and fc such that fa > fb and fb < fc.
1859: */
1.126 brouard 1860: double ulim,u,r,q, dum;
1861: double fu;
1.187 brouard 1862:
1863: double scale=10.;
1864: int iterscale=0;
1865:
1866: *fa=(*func)(*ax); /* xta[j]=pcom[j]+(*ax)*xicom[j]; fa=f(xta[j])*/
1867: *fb=(*func)(*bx); /* xtb[j]=pcom[j]+(*bx)*xicom[j]; fb=f(xtb[j]) */
1868:
1869:
1870: /* while(*fb != *fb){ /\* *ax should be ok, reducing distance to *ax *\/ */
1871: /* printf("Warning mnbrak *fb = %lf, *bx=%lf *ax=%lf *fa==%lf iter=%d\n",*fb, *bx, *ax, *fa, iterscale++); */
1872: /* *bx = *ax - (*ax - *bx)/scale; */
1873: /* *fb=(*func)(*bx); /\* xtb[j]=pcom[j]+(*bx)*xicom[j]; fb=f(xtb[j]) *\/ */
1874: /* } */
1875:
1.126 brouard 1876: if (*fb > *fa) {
1877: SHFT(dum,*ax,*bx,dum)
1.183 brouard 1878: SHFT(dum,*fb,*fa,dum)
1879: }
1.126 brouard 1880: *cx=(*bx)+GOLD*(*bx-*ax);
1881: *fc=(*func)(*cx);
1.183 brouard 1882: #ifdef DEBUG
1.224 brouard 1883: printf("mnbrak0 a=%lf *fa=%lf, b=%lf *fb=%lf, c=%lf *fc=%lf\n",*ax,*fa,*bx,*fb,*cx, *fc);
1884: fprintf(ficlog,"mnbrak0 a=%lf *fa=%lf, b=%lf *fb=%lf, c=%lf *fc=%lf\n",*ax,*fa,*bx,*fb,*cx, *fc);
1.183 brouard 1885: #endif
1.224 brouard 1886: while (*fb > *fc) { /* Declining a,b,c with fa> fb > fc. If fc=inf it exits and if flat fb=fc it exits too.*/
1.126 brouard 1887: r=(*bx-*ax)*(*fb-*fc);
1.224 brouard 1888: q=(*bx-*cx)*(*fb-*fa); /* What if fa=inf */
1.126 brouard 1889: u=(*bx)-((*bx-*cx)*q-(*bx-*ax)*r)/
1.183 brouard 1890: (2.0*SIGN(FMAX(fabs(q-r),TINY),q-r)); /* Minimum abscissa of a parabolic estimated from (a,fa), (b,fb) and (c,fc). */
1891: ulim=(*bx)+GLIMIT*(*cx-*bx); /* Maximum abscissa where function should be evaluated */
1892: if ((*bx-u)*(u-*cx) > 0.0) { /* if u_p is between b and c */
1.126 brouard 1893: fu=(*func)(u);
1.163 brouard 1894: #ifdef DEBUG
1895: /* f(x)=A(x-u)**2+f(u) */
1896: double A, fparabu;
1897: A= (*fb - *fa)/(*bx-*ax)/(*bx+*ax-2*u);
1898: fparabu= *fa - A*(*ax-u)*(*ax-u);
1.224 brouard 1899: printf("\nmnbrak (*ax=%.12f, *fa=%.12lf), (*bx=%.12f, *fb=%.12lf), (*cx=%.12f, *fc=%.12lf), (*u=%.12f, fu=%.12lf, fparabu=%.12f, q=%lf < %lf=r)\n",*ax,*fa,*bx,*fb,*cx,*fc,u,fu, fparabu,q,r);
1900: fprintf(ficlog,"\nmnbrak (*ax=%.12f, *fa=%.12lf), (*bx=%.12f, *fb=%.12lf), (*cx=%.12f, *fc=%.12lf), (*u=%.12f, fu=%.12lf, fparabu=%.12f, q=%lf < %lf=r)\n",*ax,*fa,*bx,*fb,*cx,*fc,u,fu, fparabu,q,r);
1.183 brouard 1901: /* And thus,it can be that fu > *fc even if fparabu < *fc */
1902: /* mnbrak (*ax=7.666299858533, *fa=299039.693133272231), (*bx=8.595447774979, *fb=298976.598289369489),
1903: (*cx=10.098840694817, *fc=298946.631474258087), (*u=9.852501168332, fu=298948.773013752128, fparabu=298945.434711494134) */
1904: /* In that case, there is no bracket in the output! Routine is wrong with many consequences.*/
1.163 brouard 1905: #endif
1.184 brouard 1906: #ifdef MNBRAKORIGINAL
1.183 brouard 1907: #else
1.191 brouard 1908: /* if (fu > *fc) { */
1909: /* #ifdef DEBUG */
1910: /* printf("mnbrak4 fu > fc \n"); */
1911: /* fprintf(ficlog, "mnbrak4 fu > fc\n"); */
1912: /* #endif */
1913: /* /\* SHFT(u,*cx,*cx,u) /\\* ie a=c, c=u and u=c; in that case, next SHFT(a,b,c,u) will give a=b=b, b=c=u, c=u=c and *\\/ *\/ */
1914: /* /\* SHFT(*fa,*fc,fu,*fc) /\\* (b, u, c) is a bracket while test fb > fc will be fu > fc will exit *\\/ *\/ */
1915: /* dum=u; /\* Shifting c and u *\/ */
1916: /* u = *cx; */
1917: /* *cx = dum; */
1918: /* dum = fu; */
1919: /* fu = *fc; */
1920: /* *fc =dum; */
1921: /* } else { /\* end *\/ */
1922: /* #ifdef DEBUG */
1923: /* printf("mnbrak3 fu < fc \n"); */
1924: /* fprintf(ficlog, "mnbrak3 fu < fc\n"); */
1925: /* #endif */
1926: /* dum=u; /\* Shifting c and u *\/ */
1927: /* u = *cx; */
1928: /* *cx = dum; */
1929: /* dum = fu; */
1930: /* fu = *fc; */
1931: /* *fc =dum; */
1932: /* } */
1.224 brouard 1933: #ifdef DEBUGMNBRAK
1934: double A, fparabu;
1935: A= (*fb - *fa)/(*bx-*ax)/(*bx+*ax-2*u);
1936: fparabu= *fa - A*(*ax-u)*(*ax-u);
1937: printf("\nmnbrak35 ax=%lf fa=%lf bx=%lf fb=%lf, u=%lf fp=%lf fu=%lf < or >= fc=%lf cx=%lf, q=%lf < %lf=r \n",*ax, *fa, *bx,*fb,u,fparabu,fu,*fc,*cx,q,r);
1938: fprintf(ficlog,"\nmnbrak35 ax=%lf fa=%lf bx=%lf fb=%lf, u=%lf fp=%lf fu=%lf < or >= fc=%lf cx=%lf, q=%lf < %lf=r \n",*ax, *fa, *bx,*fb,u,fparabu,fu,*fc,*cx,q,r);
1.183 brouard 1939: #endif
1.191 brouard 1940: dum=u; /* Shifting c and u */
1941: u = *cx;
1942: *cx = dum;
1943: dum = fu;
1944: fu = *fc;
1945: *fc =dum;
1.183 brouard 1946: #endif
1.162 brouard 1947: } else if ((*cx-u)*(u-ulim) > 0.0) { /* u is after c but before ulim */
1.183 brouard 1948: #ifdef DEBUG
1.224 brouard 1949: printf("\nmnbrak2 u=%lf after c=%lf but before ulim\n",u,*cx);
1950: fprintf(ficlog,"\nmnbrak2 u=%lf after c=%lf but before ulim\n",u,*cx);
1.183 brouard 1951: #endif
1.126 brouard 1952: fu=(*func)(u);
1953: if (fu < *fc) {
1.183 brouard 1954: #ifdef DEBUG
1.224 brouard 1955: printf("\nmnbrak2 u=%lf after c=%lf but before ulim=%lf AND fu=%lf < %lf=fc\n",u,*cx,ulim,fu, *fc);
1956: fprintf(ficlog,"\nmnbrak2 u=%lf after c=%lf but before ulim=%lf AND fu=%lf < %lf=fc\n",u,*cx,ulim,fu, *fc);
1957: #endif
1958: SHFT(*bx,*cx,u,*cx+GOLD*(*cx-*bx))
1959: SHFT(*fb,*fc,fu,(*func)(u))
1960: #ifdef DEBUG
1961: printf("\nmnbrak2 shift GOLD c=%lf",*cx+GOLD*(*cx-*bx));
1.183 brouard 1962: #endif
1963: }
1.162 brouard 1964: } else if ((u-ulim)*(ulim-*cx) >= 0.0) { /* u outside ulim (verifying that ulim is beyond c) */
1.183 brouard 1965: #ifdef DEBUG
1.224 brouard 1966: printf("\nmnbrak2 u=%lf outside ulim=%lf (verifying that ulim is beyond c=%lf)\n",u,ulim,*cx);
1967: fprintf(ficlog,"\nmnbrak2 u=%lf outside ulim=%lf (verifying that ulim is beyond c=%lf)\n",u,ulim,*cx);
1.183 brouard 1968: #endif
1.126 brouard 1969: u=ulim;
1970: fu=(*func)(u);
1.183 brouard 1971: } else { /* u could be left to b (if r > q parabola has a maximum) */
1972: #ifdef DEBUG
1.224 brouard 1973: printf("\nmnbrak2 u=%lf could be left to b=%lf (if r=%lf > q=%lf parabola has a maximum)\n",u,*bx,r,q);
1974: fprintf(ficlog,"\nmnbrak2 u=%lf could be left to b=%lf (if r=%lf > q=%lf parabola has a maximum)\n",u,*bx,r,q);
1.183 brouard 1975: #endif
1.126 brouard 1976: u=(*cx)+GOLD*(*cx-*bx);
1977: fu=(*func)(u);
1.224 brouard 1978: #ifdef DEBUG
1979: printf("\nmnbrak2 new u=%lf fu=%lf shifted gold left from c=%lf and b=%lf \n",u,fu,*cx,*bx);
1980: fprintf(ficlog,"\nmnbrak2 new u=%lf fu=%lf shifted gold left from c=%lf and b=%lf \n",u,fu,*cx,*bx);
1981: #endif
1.183 brouard 1982: } /* end tests */
1.126 brouard 1983: SHFT(*ax,*bx,*cx,u)
1.183 brouard 1984: SHFT(*fa,*fb,*fc,fu)
1985: #ifdef DEBUG
1.224 brouard 1986: printf("\nmnbrak2 shift (*ax=%.12f, *fa=%.12lf), (*bx=%.12f, *fb=%.12lf), (*cx=%.12f, *fc=%.12lf)\n",*ax,*fa,*bx,*fb,*cx,*fc);
1987: fprintf(ficlog, "\nmnbrak2 shift (*ax=%.12f, *fa=%.12lf), (*bx=%.12f, *fb=%.12lf), (*cx=%.12f, *fc=%.12lf)\n",*ax,*fa,*bx,*fb,*cx,*fc);
1.183 brouard 1988: #endif
1989: } /* end while; ie return (a, b, c, fa, fb, fc) such that a < b < c with f(a) > f(b) and fb < f(c) */
1.126 brouard 1990: }
1991:
1992: /*************** linmin ************************/
1.162 brouard 1993: /* Given an n -dimensional point p[1..n] and an n -dimensional direction xi[1..n] , moves and
1994: resets p to where the function func(p) takes on a minimum along the direction xi from p ,
1995: and replaces xi by the actual vector displacement that p was moved. Also returns as fret
1996: the value of func at the returned location p . This is actually all accomplished by calling the
1997: routines mnbrak and brent .*/
1.126 brouard 1998: int ncom;
1999: double *pcom,*xicom;
2000: double (*nrfunc)(double []);
2001:
1.224 brouard 2002: #ifdef LINMINORIGINAL
1.126 brouard 2003: void linmin(double p[], double xi[], int n, double *fret,double (*func)(double []))
1.224 brouard 2004: #else
2005: void linmin(double p[], double xi[], int n, double *fret,double (*func)(double []), int *flat)
2006: #endif
1.126 brouard 2007: {
2008: double brent(double ax, double bx, double cx,
2009: double (*f)(double), double tol, double *xmin);
2010: double f1dim(double x);
2011: void mnbrak(double *ax, double *bx, double *cx, double *fa, double *fb,
2012: double *fc, double (*func)(double));
2013: int j;
2014: double xx,xmin,bx,ax;
2015: double fx,fb,fa;
1.187 brouard 2016:
1.203 brouard 2017: #ifdef LINMINORIGINAL
2018: #else
2019: double scale=10., axs, xxs; /* Scale added for infinity */
2020: #endif
2021:
1.126 brouard 2022: ncom=n;
2023: pcom=vector(1,n);
2024: xicom=vector(1,n);
2025: nrfunc=func;
2026: for (j=1;j<=n;j++) {
2027: pcom[j]=p[j];
1.202 brouard 2028: xicom[j]=xi[j]; /* Former scale xi[j] of currrent direction i */
1.126 brouard 2029: }
1.187 brouard 2030:
1.203 brouard 2031: #ifdef LINMINORIGINAL
2032: xx=1.;
2033: #else
2034: axs=0.0;
2035: xxs=1.;
2036: do{
2037: xx= xxs;
2038: #endif
1.187 brouard 2039: ax=0.;
2040: mnbrak(&ax,&xx,&bx,&fa,&fx,&fb,f1dim); /* Outputs: xtx[j]=pcom[j]+(*xx)*xicom[j]; fx=f(xtx[j]) */
2041: /* brackets with inputs ax=0 and xx=1, but points, pcom=p, and directions values, xicom=xi, are sent via f1dim(x) */
2042: /* xt[x,j]=pcom[j]+x*xicom[j] f(ax) = f(xt(a,j=1,n)) = f(p(j) + 0 * xi(j)) and f(xx) = f(xt(x, j=1,n)) = f(p(j) + 1 * xi(j)) */
2043: /* Outputs: fa=f(p(j)) and fx=f(p(j) + xxs * xi(j) ) and f(bx)= f(p(j)+ bx* xi(j)) */
2044: /* Given input ax=axs and xx=xxs, xx might be too far from ax to get a finite f(xx) */
2045: /* Searches on line, outputs (ax, xx, bx) such that fx < min(fa and fb) */
2046: /* Find a bracket a,x,b in direction n=xi ie xicom, order may change. Scale is [0:xxs*xi[j]] et non plus [0:xi[j]]*/
1.203 brouard 2047: #ifdef LINMINORIGINAL
2048: #else
2049: if (fx != fx){
1.224 brouard 2050: xxs=xxs/scale; /* Trying a smaller xx, closer to initial ax=0 */
2051: printf("|");
2052: fprintf(ficlog,"|");
1.203 brouard 2053: #ifdef DEBUGLINMIN
1.224 brouard 2054: printf("\nLinmin NAN : input [axs=%lf:xxs=%lf], mnbrak outputs fx=%lf <(fb=%lf and fa=%lf) with xx=%lf in [ax=%lf:bx=%lf] \n", axs, xxs, fx,fb, fa, xx, ax, bx);
1.203 brouard 2055: #endif
2056: }
1.224 brouard 2057: }while(fx != fx && xxs > 1.e-5);
1.203 brouard 2058: #endif
2059:
1.191 brouard 2060: #ifdef DEBUGLINMIN
2061: printf("\nLinmin after mnbrak: ax=%12.7f xx=%12.7f bx=%12.7f fa=%12.2f fx=%12.2f fb=%12.2f\n", ax,xx,bx,fa,fx,fb);
1.202 brouard 2062: fprintf(ficlog,"\nLinmin after mnbrak: ax=%12.7f xx=%12.7f bx=%12.7f fa=%12.2f fx=%12.2f fb=%12.2f\n", ax,xx,bx,fa,fx,fb);
1.191 brouard 2063: #endif
1.224 brouard 2064: #ifdef LINMINORIGINAL
2065: #else
2066: if(fb == fx){ /* Flat function in the direction */
2067: xmin=xx;
2068: *flat=1;
2069: }else{
2070: *flat=0;
2071: #endif
2072: /*Flat mnbrak2 shift (*ax=0.000000000000, *fa=51626.272983130431), (*bx=-1.618034000000, *fb=51590.149499362531), (*cx=-4.236068025156, *fc=51590.149499362531) */
1.187 brouard 2073: *fret=brent(ax,xx,bx,f1dim,TOL,&xmin); /* Giving a bracketting triplet (ax, xx, bx), find a minimum, xmin, according to f1dim, *fret(xmin),*/
2074: /* fa = f(p[j] + ax * xi[j]), fx = f(p[j] + xx * xi[j]), fb = f(p[j] + bx * xi[j]) */
2075: /* fmin = f(p[j] + xmin * xi[j]) */
2076: /* P+lambda n in that direction (lambdamin), with TOL between abscisses */
2077: /* f1dim(xmin): for (j=1;j<=ncom;j++) xt[j]=pcom[j]+xmin*xicom[j]; */
1.126 brouard 2078: #ifdef DEBUG
1.224 brouard 2079: printf("retour brent from bracket (a=%lf fa=%lf, xx=%lf fx=%lf, b=%lf fb=%lf): fret=%lf xmin=%lf\n",ax,fa,xx,fx,bx,fb,*fret,xmin);
2080: fprintf(ficlog,"retour brent from bracket (a=%lf fa=%lf, xx=%lf fx=%lf, b=%lf fb=%lf): fret=%lf xmin=%lf\n",ax,fa,xx,fx,bx,fb,*fret,xmin);
2081: #endif
2082: #ifdef LINMINORIGINAL
2083: #else
2084: }
1.126 brouard 2085: #endif
1.191 brouard 2086: #ifdef DEBUGLINMIN
2087: printf("linmin end ");
1.202 brouard 2088: fprintf(ficlog,"linmin end ");
1.191 brouard 2089: #endif
1.126 brouard 2090: for (j=1;j<=n;j++) {
1.203 brouard 2091: #ifdef LINMINORIGINAL
2092: xi[j] *= xmin;
2093: #else
2094: #ifdef DEBUGLINMIN
2095: if(xxs <1.0)
2096: printf(" before xi[%d]=%12.8f", j,xi[j]);
2097: #endif
2098: xi[j] *= xmin*xxs; /* xi rescaled by xmin and number of loops: if xmin=-1.237 and xi=(1,0,...,0) xi=(-1.237,0,...,0) */
2099: #ifdef DEBUGLINMIN
2100: if(xxs <1.0)
2101: printf(" after xi[%d]=%12.8f, xmin=%12.8f, ax=%12.8f, xx=%12.8f, bx=%12.8f, xxs=%12.8f", j,xi[j], xmin, ax, xx, bx,xxs );
2102: #endif
2103: #endif
1.187 brouard 2104: p[j] += xi[j]; /* Parameters values are updated accordingly */
1.126 brouard 2105: }
1.191 brouard 2106: #ifdef DEBUGLINMIN
1.203 brouard 2107: printf("\n");
1.191 brouard 2108: printf("Comparing last *frec(xmin=%12.8f)=%12.8f from Brent and frec(0.)=%12.8f \n", xmin, *fret, (*func)(p));
1.202 brouard 2109: fprintf(ficlog,"Comparing last *frec(xmin=%12.8f)=%12.8f from Brent and frec(0.)=%12.8f \n", xmin, *fret, (*func)(p));
1.191 brouard 2110: for (j=1;j<=n;j++) {
1.202 brouard 2111: printf(" xi[%d]= %14.10f p[%d]= %12.7f",j,xi[j],j,p[j]);
2112: fprintf(ficlog," xi[%d]= %14.10f p[%d]= %12.7f",j,xi[j],j,p[j]);
2113: if(j % ncovmodel == 0){
1.191 brouard 2114: printf("\n");
1.202 brouard 2115: fprintf(ficlog,"\n");
2116: }
1.191 brouard 2117: }
1.203 brouard 2118: #else
1.191 brouard 2119: #endif
1.126 brouard 2120: free_vector(xicom,1,n);
2121: free_vector(pcom,1,n);
2122: }
2123:
2124:
2125: /*************** powell ************************/
1.162 brouard 2126: /*
2127: Minimization of a function func of n variables. Input consists of an initial starting point
2128: p[1..n] ; an initial matrix xi[1..n][1..n] , whose columns contain the initial set of di-
2129: rections (usually the n unit vectors); and ftol , the fractional tolerance in the function value
2130: such that failure to decrease by more than this amount on one iteration signals doneness. On
2131: output, p is set to the best point found, xi is the then-current direction set, fret is the returned
2132: function value at p , and iter is the number of iterations taken. The routine linmin is used.
2133: */
1.224 brouard 2134: #ifdef LINMINORIGINAL
2135: #else
2136: int *flatdir; /* Function is vanishing in that direction */
1.225 brouard 2137: int flat=0, flatd=0; /* Function is vanishing in that direction */
1.224 brouard 2138: #endif
1.126 brouard 2139: void powell(double p[], double **xi, int n, double ftol, int *iter, double *fret,
2140: double (*func)(double []))
2141: {
1.224 brouard 2142: #ifdef LINMINORIGINAL
2143: void linmin(double p[], double xi[], int n, double *fret,
1.126 brouard 2144: double (*func)(double []));
1.224 brouard 2145: #else
1.241 brouard 2146: void linmin(double p[], double xi[], int n, double *fret,
2147: double (*func)(double []),int *flat);
1.224 brouard 2148: #endif
1.239 brouard 2149: int i,ibig,j,jk,k;
1.126 brouard 2150: double del,t,*pt,*ptt,*xit;
1.181 brouard 2151: double directest;
1.126 brouard 2152: double fp,fptt;
2153: double *xits;
2154: int niterf, itmp;
1.224 brouard 2155: #ifdef LINMINORIGINAL
2156: #else
2157:
2158: flatdir=ivector(1,n);
2159: for (j=1;j<=n;j++) flatdir[j]=0;
2160: #endif
1.126 brouard 2161:
2162: pt=vector(1,n);
2163: ptt=vector(1,n);
2164: xit=vector(1,n);
2165: xits=vector(1,n);
2166: *fret=(*func)(p);
2167: for (j=1;j<=n;j++) pt[j]=p[j];
1.202 brouard 2168: rcurr_time = time(NULL);
1.126 brouard 2169: for (*iter=1;;++(*iter)) {
1.187 brouard 2170: fp=(*fret); /* From former iteration or initial value */
1.126 brouard 2171: ibig=0;
2172: del=0.0;
1.157 brouard 2173: rlast_time=rcurr_time;
2174: /* (void) gettimeofday(&curr_time,&tzp); */
2175: rcurr_time = time(NULL);
2176: curr_time = *localtime(&rcurr_time);
2177: printf("\nPowell iter=%d -2*LL=%.12f %ld sec. %ld sec.",*iter,*fret, rcurr_time-rlast_time, rcurr_time-rstart_time);fflush(stdout);
2178: fprintf(ficlog,"\nPowell iter=%d -2*LL=%.12f %ld sec. %ld sec.",*iter,*fret,rcurr_time-rlast_time, rcurr_time-rstart_time); fflush(ficlog);
2179: /* fprintf(ficrespow,"%d %.12f %ld",*iter,*fret,curr_time.tm_sec-start_time.tm_sec); */
1.192 brouard 2180: for (i=1;i<=n;i++) {
1.126 brouard 2181: fprintf(ficrespow," %.12lf", p[i]);
2182: }
1.239 brouard 2183: fprintf(ficrespow,"\n");fflush(ficrespow);
2184: printf("\n#model= 1 + age ");
2185: fprintf(ficlog,"\n#model= 1 + age ");
2186: if(nagesqr==1){
1.241 brouard 2187: printf(" + age*age ");
2188: fprintf(ficlog," + age*age ");
1.239 brouard 2189: }
2190: for(j=1;j <=ncovmodel-2;j++){
2191: if(Typevar[j]==0) {
2192: printf(" + V%d ",Tvar[j]);
2193: fprintf(ficlog," + V%d ",Tvar[j]);
2194: }else if(Typevar[j]==1) {
2195: printf(" + V%d*age ",Tvar[j]);
2196: fprintf(ficlog," + V%d*age ",Tvar[j]);
2197: }else if(Typevar[j]==2) {
2198: printf(" + V%d*V%d ",Tvard[Tposprod[j]][1],Tvard[Tposprod[j]][2]);
2199: fprintf(ficlog," + V%d*V%d ",Tvard[Tposprod[j]][1],Tvard[Tposprod[j]][2]);
2200: }
2201: }
1.126 brouard 2202: printf("\n");
1.239 brouard 2203: /* printf("12 47.0114589 0.0154322 33.2424412 0.3279905 2.3731903 */
2204: /* 13 -21.5392400 0.1118147 1.2680506 1.2973408 -1.0663662 */
1.126 brouard 2205: fprintf(ficlog,"\n");
1.239 brouard 2206: for(i=1,jk=1; i <=nlstate; i++){
2207: for(k=1; k <=(nlstate+ndeath); k++){
2208: if (k != i) {
2209: printf("%d%d ",i,k);
2210: fprintf(ficlog,"%d%d ",i,k);
2211: for(j=1; j <=ncovmodel; j++){
2212: printf("%12.7f ",p[jk]);
2213: fprintf(ficlog,"%12.7f ",p[jk]);
2214: jk++;
2215: }
2216: printf("\n");
2217: fprintf(ficlog,"\n");
2218: }
2219: }
2220: }
1.241 brouard 2221: if(*iter <=3 && *iter >1){
1.157 brouard 2222: tml = *localtime(&rcurr_time);
2223: strcpy(strcurr,asctime(&tml));
2224: rforecast_time=rcurr_time;
1.126 brouard 2225: itmp = strlen(strcurr);
2226: if(strcurr[itmp-1]=='\n') /* Windows outputs with a new line */
1.241 brouard 2227: strcurr[itmp-1]='\0';
1.162 brouard 2228: printf("\nConsidering the time needed for the last iteration #%d: %ld seconds,\n",*iter,rcurr_time-rlast_time);
1.157 brouard 2229: fprintf(ficlog,"\nConsidering the time needed for this last iteration #%d: %ld seconds,\n",*iter,rcurr_time-rlast_time);
1.126 brouard 2230: for(niterf=10;niterf<=30;niterf+=10){
1.241 brouard 2231: rforecast_time=rcurr_time+(niterf-*iter)*(rcurr_time-rlast_time);
2232: forecast_time = *localtime(&rforecast_time);
2233: strcpy(strfor,asctime(&forecast_time));
2234: itmp = strlen(strfor);
2235: if(strfor[itmp-1]=='\n')
2236: strfor[itmp-1]='\0';
2237: printf(" - if your program needs %d iterations to converge, convergence will be \n reached in %s i.e.\n on %s (current time is %s);\n",niterf, asc_diff_time(rforecast_time-rcurr_time,tmpout),strfor,strcurr);
2238: fprintf(ficlog," - if your program needs %d iterations to converge, convergence will be \n reached in %s i.e.\n on %s (current time is %s);\n",niterf, asc_diff_time(rforecast_time-rcurr_time,tmpout),strfor,strcurr);
1.126 brouard 2239: }
2240: }
1.187 brouard 2241: for (i=1;i<=n;i++) { /* For each direction i */
2242: for (j=1;j<=n;j++) xit[j]=xi[j][i]; /* Directions stored from previous iteration with previous scales */
1.126 brouard 2243: fptt=(*fret);
2244: #ifdef DEBUG
1.203 brouard 2245: printf("fret=%lf, %lf, %lf \n", *fret, *fret, *fret);
2246: fprintf(ficlog, "fret=%lf, %lf, %lf \n", *fret, *fret, *fret);
1.126 brouard 2247: #endif
1.203 brouard 2248: printf("%d",i);fflush(stdout); /* print direction (parameter) i */
1.126 brouard 2249: fprintf(ficlog,"%d",i);fflush(ficlog);
1.224 brouard 2250: #ifdef LINMINORIGINAL
1.188 brouard 2251: linmin(p,xit,n,fret,func); /* Point p[n]. xit[n] has been loaded for direction i as input.*/
1.224 brouard 2252: #else
2253: linmin(p,xit,n,fret,func,&flat); /* Point p[n]. xit[n] has been loaded for direction i as input.*/
2254: flatdir[i]=flat; /* Function is vanishing in that direction i */
2255: #endif
2256: /* Outputs are fret(new point p) p is updated and xit rescaled */
1.188 brouard 2257: if (fabs(fptt-(*fret)) > del) { /* We are keeping the max gain on each of the n directions */
1.224 brouard 2258: /* because that direction will be replaced unless the gain del is small */
2259: /* in comparison with the 'probable' gain, mu^2, with the last average direction. */
2260: /* Unless the n directions are conjugate some gain in the determinant may be obtained */
2261: /* with the new direction. */
2262: del=fabs(fptt-(*fret));
2263: ibig=i;
1.126 brouard 2264: }
2265: #ifdef DEBUG
2266: printf("%d %.12e",i,(*fret));
2267: fprintf(ficlog,"%d %.12e",i,(*fret));
2268: for (j=1;j<=n;j++) {
1.224 brouard 2269: xits[j]=FMAX(fabs(p[j]-pt[j]),1.e-5);
2270: printf(" x(%d)=%.12e",j,xit[j]);
2271: fprintf(ficlog," x(%d)=%.12e",j,xit[j]);
1.126 brouard 2272: }
2273: for(j=1;j<=n;j++) {
1.225 brouard 2274: printf(" p(%d)=%.12e",j,p[j]);
2275: fprintf(ficlog," p(%d)=%.12e",j,p[j]);
1.126 brouard 2276: }
2277: printf("\n");
2278: fprintf(ficlog,"\n");
2279: #endif
1.187 brouard 2280: } /* end loop on each direction i */
2281: /* Convergence test will use last linmin estimation (fret) and compare former iteration (fp) */
1.188 brouard 2282: /* But p and xit have been updated at the end of linmin, *fret corresponds to new p, xit */
1.187 brouard 2283: /* New value of last point Pn is not computed, P(n-1) */
1.224 brouard 2284: for(j=1;j<=n;j++) {
1.225 brouard 2285: if(flatdir[j] >0){
2286: printf(" p(%d)=%lf flat=%d ",j,p[j],flatdir[j]);
2287: fprintf(ficlog," p(%d)=%lf flat=%d ",j,p[j],flatdir[j]);
2288: }
2289: /* printf("\n"); */
2290: /* fprintf(ficlog,"\n"); */
2291: }
1.243 brouard 2292: /* if (2.0*fabs(fp-(*fret)) <= ftol*(fabs(fp)+fabs(*fret))) { /\* Did we reach enough precision? *\/ */
2293: if (2.0*fabs(fp-(*fret)) <= ftol) { /* Did we reach enough precision? */
1.188 brouard 2294: /* We could compare with a chi^2. chisquare(0.95,ddl=1)=3.84 */
2295: /* By adding age*age in a model, the new -2LL should be lower and the difference follows a */
2296: /* a chisquare statistics with 1 degree. To be significant at the 95% level, it should have */
2297: /* decreased of more than 3.84 */
2298: /* By adding age*age and V1*age the gain (-2LL) should be more than 5.99 (ddl=2) */
2299: /* By using V1+V2+V3, the gain should be 7.82, compared with basic 1+age. */
2300: /* By adding 10 parameters more the gain should be 18.31 */
1.224 brouard 2301:
1.188 brouard 2302: /* Starting the program with initial values given by a former maximization will simply change */
2303: /* the scales of the directions and the directions, because the are reset to canonical directions */
2304: /* Thus the first calls to linmin will give new points and better maximizations until fp-(*fret) is */
2305: /* under the tolerance value. If the tolerance is very small 1.e-9, it could last long. */
1.126 brouard 2306: #ifdef DEBUG
2307: int k[2],l;
2308: k[0]=1;
2309: k[1]=-1;
2310: printf("Max: %.12e",(*func)(p));
2311: fprintf(ficlog,"Max: %.12e",(*func)(p));
2312: for (j=1;j<=n;j++) {
2313: printf(" %.12e",p[j]);
2314: fprintf(ficlog," %.12e",p[j]);
2315: }
2316: printf("\n");
2317: fprintf(ficlog,"\n");
2318: for(l=0;l<=1;l++) {
2319: for (j=1;j<=n;j++) {
2320: ptt[j]=p[j]+(p[j]-pt[j])*k[l];
2321: printf("l=%d j=%d ptt=%.12e, xits=%.12e, p=%.12e, xit=%.12e", l,j,ptt[j],xits[j],p[j],xit[j]);
2322: fprintf(ficlog,"l=%d j=%d ptt=%.12e, xits=%.12e, p=%.12e, xit=%.12e", l,j,ptt[j],xits[j],p[j],xit[j]);
2323: }
2324: printf("func(ptt)=%.12e, deriv=%.12e\n",(*func)(ptt),(ptt[j]-p[j])/((*func)(ptt)-(*func)(p)));
2325: fprintf(ficlog,"func(ptt)=%.12e, deriv=%.12e\n",(*func)(ptt),(ptt[j]-p[j])/((*func)(ptt)-(*func)(p)));
2326: }
2327: #endif
2328:
1.224 brouard 2329: #ifdef LINMINORIGINAL
2330: #else
2331: free_ivector(flatdir,1,n);
2332: #endif
1.126 brouard 2333: free_vector(xit,1,n);
2334: free_vector(xits,1,n);
2335: free_vector(ptt,1,n);
2336: free_vector(pt,1,n);
2337: return;
1.192 brouard 2338: } /* enough precision */
1.240 brouard 2339: if (*iter == ITMAX*n) nrerror("powell exceeding maximum iterations.");
1.181 brouard 2340: for (j=1;j<=n;j++) { /* Computes the extrapolated point P_0 + 2 (P_n-P_0) */
1.126 brouard 2341: ptt[j]=2.0*p[j]-pt[j];
2342: xit[j]=p[j]-pt[j];
2343: pt[j]=p[j];
2344: }
1.181 brouard 2345: fptt=(*func)(ptt); /* f_3 */
1.224 brouard 2346: #ifdef NODIRECTIONCHANGEDUNTILNITER /* No change in drections until some iterations are done */
2347: if (*iter <=4) {
1.225 brouard 2348: #else
2349: #endif
1.224 brouard 2350: #ifdef POWELLNOF3INFF1TEST /* skips test F3 <F1 */
1.192 brouard 2351: #else
1.161 brouard 2352: if (fptt < fp) { /* If extrapolated point is better, decide if we keep that new direction or not */
1.192 brouard 2353: #endif
1.162 brouard 2354: /* (x1 f1=fp), (x2 f2=*fret), (x3 f3=fptt), (xm fm) */
1.161 brouard 2355: /* From x1 (P0) distance of x2 is at h and x3 is 2h */
1.162 brouard 2356: /* Let f"(x2) be the 2nd derivative equal everywhere. */
2357: /* Then the parabolic through (x1,f1), (x2,f2) and (x3,f3) */
2358: /* will reach at f3 = fm + h^2/2 f"m ; f" = (f1 -2f2 +f3 ) / h**2 */
1.224 brouard 2359: /* Conditional for using this new direction is that mu^2 = (f1-2f2+f3)^2 /2 < del or directest <0 */
2360: /* also lamda^2=(f1-f2)^2/mu² is a parasite solution of powell */
2361: /* For powell, inclusion of this average direction is only if t(del)<0 or del inbetween mu^2 and lambda^2 */
1.161 brouard 2362: /* t=2.0*(fp-2.0*(*fret)+fptt)*SQR(fp-(*fret)-del)-del*SQR(fp-fptt); */
1.224 brouard 2363: /* Even if f3 <f1, directest can be negative and t >0 */
2364: /* mu² and del² are equal when f3=f1 */
2365: /* f3 < f1 : mu² < del <= lambda^2 both test are equivalent */
2366: /* f3 < f1 : mu² < lambda^2 < del then directtest is negative and powell t is positive */
2367: /* f3 > f1 : lambda² < mu^2 < del then t is negative and directest >0 */
2368: /* f3 > f1 : lambda² < del < mu^2 then t is positive and directest >0 */
1.183 brouard 2369: #ifdef NRCORIGINAL
2370: t=2.0*(fp-2.0*(*fret)+fptt)*SQR(fp-(*fret)-del)- del*SQR(fp-fptt); /* Original Numerical Recipes in C*/
2371: #else
2372: t=2.0*(fp-2.0*(*fret)+fptt)*SQR(fp-(*fret)-del); /* Intel compiler doesn't work on one line; bug reported */
1.161 brouard 2373: t= t- del*SQR(fp-fptt);
1.183 brouard 2374: #endif
1.202 brouard 2375: directest = fp-2.0*(*fret)+fptt - 2.0 * del; /* If delta was big enough we change it for a new direction */
1.161 brouard 2376: #ifdef DEBUG
1.181 brouard 2377: printf("t1= %.12lf, t2= %.12lf, t=%.12lf directest=%.12lf\n", 2.0*(fp-2.0*(*fret)+fptt)*SQR(fp-(*fret)-del),del*SQR(fp-fptt),t,directest);
2378: fprintf(ficlog,"t1= %.12lf, t2= %.12lf, t=%.12lf directest=%.12lf\n", 2.0*(fp-2.0*(*fret)+fptt)*SQR(fp-(*fret)-del),del*SQR(fp-fptt),t,directest);
1.161 brouard 2379: printf("t3= %.12lf, t4= %.12lf, t3*= %.12lf, t4*= %.12lf\n",SQR(fp-(*fret)-del),SQR(fp-fptt),
2380: (fp-(*fret)-del)*(fp-(*fret)-del),(fp-fptt)*(fp-fptt));
2381: fprintf(ficlog,"t3= %.12lf, t4= %.12lf, t3*= %.12lf, t4*= %.12lf\n",SQR(fp-(*fret)-del),SQR(fp-fptt),
2382: (fp-(*fret)-del)*(fp-(*fret)-del),(fp-fptt)*(fp-fptt));
2383: printf("tt= %.12lf, t=%.12lf\n",2.0*(fp-2.0*(*fret)+fptt)*(fp-(*fret)-del)*(fp-(*fret)-del)-del*(fp-fptt)*(fp-fptt),t);
2384: fprintf(ficlog, "tt= %.12lf, t=%.12lf\n",2.0*(fp-2.0*(*fret)+fptt)*(fp-(*fret)-del)*(fp-(*fret)-del)-del*(fp-fptt)*(fp-fptt),t);
2385: #endif
1.183 brouard 2386: #ifdef POWELLORIGINAL
2387: if (t < 0.0) { /* Then we use it for new direction */
2388: #else
1.182 brouard 2389: if (directest*t < 0.0) { /* Contradiction between both tests */
1.224 brouard 2390: printf("directest= %.12lf (if <0 we include P0 Pn as new direction), t= %.12lf, f1= %.12lf,f2= %.12lf,f3= %.12lf, del= %.12lf\n",directest, t, fp,(*fret),fptt,del);
1.192 brouard 2391: printf("f1-2f2+f3= %.12lf, f1-f2-del= %.12lf, f1-f3= %.12lf\n",fp-2.0*(*fret)+fptt, fp -(*fret) -del, fp-fptt);
1.224 brouard 2392: fprintf(ficlog,"directest= %.12lf (if directest<0 or t<0 we include P0 Pn as new direction), t= %.12lf, f1= %.12lf,f2= %.12lf,f3= %.12lf, del= %.12lf\n",directest, t, fp,(*fret),fptt, del);
1.192 brouard 2393: fprintf(ficlog,"f1-2f2+f3= %.12lf, f1-f2-del= %.12lf, f1-f3= %.12lf\n",fp-2.0*(*fret)+fptt, fp -(*fret) -del, fp-fptt);
2394: }
1.181 brouard 2395: if (directest < 0.0) { /* Then we use it for new direction */
2396: #endif
1.191 brouard 2397: #ifdef DEBUGLINMIN
1.234 brouard 2398: printf("Before linmin in direction P%d-P0\n",n);
2399: for (j=1;j<=n;j++) {
2400: printf(" Before xit[%d]= %12.7f p[%d]= %12.7f",j,xit[j],j,p[j]);
2401: fprintf(ficlog," Before xit[%d]= %12.7f p[%d]= %12.7f",j,xit[j],j,p[j]);
2402: if(j % ncovmodel == 0){
2403: printf("\n");
2404: fprintf(ficlog,"\n");
2405: }
2406: }
1.224 brouard 2407: #endif
2408: #ifdef LINMINORIGINAL
1.234 brouard 2409: linmin(p,xit,n,fret,func); /* computes minimum on the extrapolated direction: changes p and rescales xit.*/
1.224 brouard 2410: #else
1.234 brouard 2411: linmin(p,xit,n,fret,func,&flat); /* computes minimum on the extrapolated direction: changes p and rescales xit.*/
2412: flatdir[i]=flat; /* Function is vanishing in that direction i */
1.191 brouard 2413: #endif
1.234 brouard 2414:
1.191 brouard 2415: #ifdef DEBUGLINMIN
1.234 brouard 2416: for (j=1;j<=n;j++) {
2417: printf("After xit[%d]= %12.7f p[%d]= %12.7f",j,xit[j],j,p[j]);
2418: fprintf(ficlog,"After xit[%d]= %12.7f p[%d]= %12.7f",j,xit[j],j,p[j]);
2419: if(j % ncovmodel == 0){
2420: printf("\n");
2421: fprintf(ficlog,"\n");
2422: }
2423: }
1.224 brouard 2424: #endif
1.234 brouard 2425: for (j=1;j<=n;j++) {
2426: xi[j][ibig]=xi[j][n]; /* Replace direction with biggest decrease by last direction n */
2427: xi[j][n]=xit[j]; /* and this nth direction by the by the average p_0 p_n */
2428: }
1.224 brouard 2429: #ifdef LINMINORIGINAL
2430: #else
1.234 brouard 2431: for (j=1, flatd=0;j<=n;j++) {
2432: if(flatdir[j]>0)
2433: flatd++;
2434: }
2435: if(flatd >0){
1.255 brouard 2436: printf("%d flat directions: ",flatd);
2437: fprintf(ficlog,"%d flat directions :",flatd);
1.234 brouard 2438: for (j=1;j<=n;j++) {
2439: if(flatdir[j]>0){
2440: printf("%d ",j);
2441: fprintf(ficlog,"%d ",j);
2442: }
2443: }
2444: printf("\n");
2445: fprintf(ficlog,"\n");
2446: }
1.191 brouard 2447: #endif
1.234 brouard 2448: printf("Gaining to use new average direction of P0 P%d instead of biggest increase direction %d :\n",n,ibig);
2449: fprintf(ficlog,"Gaining to use new average direction of P0 P%d instead of biggest increase direction %d :\n",n,ibig);
2450:
1.126 brouard 2451: #ifdef DEBUG
1.234 brouard 2452: printf("Direction changed last moved %d in place of ibig=%d, new last is the average:\n",n,ibig);
2453: fprintf(ficlog,"Direction changed last moved %d in place of ibig=%d, new last is the average:\n",n,ibig);
2454: for(j=1;j<=n;j++){
2455: printf(" %lf",xit[j]);
2456: fprintf(ficlog," %lf",xit[j]);
2457: }
2458: printf("\n");
2459: fprintf(ficlog,"\n");
1.126 brouard 2460: #endif
1.192 brouard 2461: } /* end of t or directest negative */
1.224 brouard 2462: #ifdef POWELLNOF3INFF1TEST
1.192 brouard 2463: #else
1.234 brouard 2464: } /* end if (fptt < fp) */
1.192 brouard 2465: #endif
1.225 brouard 2466: #ifdef NODIRECTIONCHANGEDUNTILNITER /* No change in drections until some iterations are done */
1.234 brouard 2467: } /*NODIRECTIONCHANGEDUNTILNITER No change in drections until some iterations are done */
1.225 brouard 2468: #else
1.224 brouard 2469: #endif
1.234 brouard 2470: } /* loop iteration */
1.126 brouard 2471: }
1.234 brouard 2472:
1.126 brouard 2473: /**** Prevalence limit (stable or period prevalence) ****************/
1.234 brouard 2474:
1.235 brouard 2475: double **prevalim(double **prlim, int nlstate, double x[], double age, double **oldm, double **savm, double ftolpl, int *ncvyear, int ij, int nres)
1.234 brouard 2476: {
1.235 brouard 2477: /* Computes the prevalence limit in each live state at age x and for covariate combination ij
2478: (and selected quantitative values in nres)
2479: by left multiplying the unit
1.234 brouard 2480: matrix by transitions matrix until convergence is reached with precision ftolpl */
1.206 brouard 2481: /* Wx= Wx-1 Px-1= Wx-2 Px-2 Px-1 = Wx-n Px-n ... Px-2 Px-1 I */
2482: /* Wx is row vector: population in state 1, population in state 2, population dead */
2483: /* or prevalence in state 1, prevalence in state 2, 0 */
2484: /* newm is the matrix after multiplications, its rows are identical at a factor */
2485: /* Initial matrix pimij */
2486: /* {0.85204250825084937, 0.13044499163996345, 0.017512500109187184, */
2487: /* 0.090851990222114765, 0.88271245433047185, 0.026435555447413338, */
2488: /* 0, 0 , 1} */
2489: /*
2490: * and after some iteration: */
2491: /* {0.45504275246439968, 0.42731458730878791, 0.11764266022681241, */
2492: /* 0.45201005341706885, 0.42865420071559901, 0.11933574586733192, */
2493: /* 0, 0 , 1} */
2494: /* And prevalence by suppressing the deaths are close to identical rows in prlim: */
2495: /* {0.51571254859325999, 0.4842874514067399, */
2496: /* 0.51326036147820708, 0.48673963852179264} */
2497: /* If we start from prlim again, prlim tends to a constant matrix */
1.234 brouard 2498:
1.126 brouard 2499: int i, ii,j,k;
1.209 brouard 2500: double *min, *max, *meandiff, maxmax,sumnew=0.;
1.145 brouard 2501: /* double **matprod2(); */ /* test */
1.218 brouard 2502: double **out, cov[NCOVMAX+1], **pmij(); /* **pmmij is a global variable feeded with oldms etc */
1.126 brouard 2503: double **newm;
1.209 brouard 2504: double agefin, delaymax=200. ; /* 100 Max number of years to converge */
1.203 brouard 2505: int ncvloop=0;
1.169 brouard 2506:
1.209 brouard 2507: min=vector(1,nlstate);
2508: max=vector(1,nlstate);
2509: meandiff=vector(1,nlstate);
2510:
1.218 brouard 2511: /* Starting with matrix unity */
1.126 brouard 2512: for (ii=1;ii<=nlstate+ndeath;ii++)
2513: for (j=1;j<=nlstate+ndeath;j++){
2514: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
2515: }
1.169 brouard 2516:
2517: cov[1]=1.;
2518:
2519: /* Even if hstepm = 1, at least one multiplication by the unit matrix */
1.202 brouard 2520: /* Start at agefin= age, computes the matrix of passage and loops decreasing agefin until convergence is reached */
1.126 brouard 2521: for(agefin=age-stepm/YEARM; agefin>=age-delaymax; agefin=agefin-stepm/YEARM){
1.202 brouard 2522: ncvloop++;
1.126 brouard 2523: newm=savm;
2524: /* Covariates have to be included here again */
1.138 brouard 2525: cov[2]=agefin;
1.187 brouard 2526: if(nagesqr==1)
2527: cov[3]= agefin*agefin;;
1.234 brouard 2528: for (k=1; k<=nsd;k++) { /* For single dummy covariates only */
2529: /* Here comes the value of the covariate 'ij' after renumbering k with single dummy covariates */
2530: cov[2+nagesqr+TvarsDind[k]]=nbcode[TvarsD[k]][codtabm(ij,k)];
1.235 brouard 2531: /* printf("prevalim Dummy combi=%d k=%d TvarsD[%d]=V%d TvarsDind[%d]=%d nbcode=%d cov=%lf codtabm(%d,Tvar[%d])=%d \n",ij,k, k, TvarsD[k],k,TvarsDind[k],nbcode[TvarsD[k]][codtabm(ij,k)],cov[2+nagesqr+TvarsDind[k]], ij, k, codtabm(ij,k)); */
1.234 brouard 2532: }
2533: for (k=1; k<=nsq;k++) { /* For single varying covariates only */
2534: /* Here comes the value of quantitative after renumbering k with single quantitative covariates */
1.235 brouard 2535: cov[2+nagesqr+TvarsQind[k]]=Tqresult[nres][k];
2536: /* printf("prevalim Quantitative k=%d TvarsQind[%d]=%d, TvarsQ[%d]=V%d,Tqresult[%d][%d]=%f\n",k,k,TvarsQind[k],k,TvarsQ[k],nres,k,Tqresult[nres][k]); */
1.138 brouard 2537: }
1.237 brouard 2538: for (k=1; k<=cptcovage;k++){ /* For product with age */
1.234 brouard 2539: if(Dummy[Tvar[Tage[k]]]){
2540: cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2];
2541: } else{
1.235 brouard 2542: cov[2+nagesqr+Tage[k]]=Tqresult[nres][k];
1.234 brouard 2543: }
1.235 brouard 2544: /* printf("prevalim Age combi=%d k=%d Tage[%d]=V%d Tqresult[%d][%d]=%f\n",ij,k,k,Tage[k],nres,k,Tqresult[nres][k]); */
1.234 brouard 2545: }
1.237 brouard 2546: for (k=1; k<=cptcovprod;k++){ /* For product without age */
1.235 brouard 2547: /* printf("prevalim Prod ij=%d k=%d Tprod[%d]=%d Tvard[%d][1]=V%d, Tvard[%d][2]=V%d\n",ij,k,k,Tprod[k], k,Tvard[k][1], k,Tvard[k][2]); */
1.237 brouard 2548: if(Dummy[Tvard[k][1]==0]){
2549: if(Dummy[Tvard[k][2]==0]){
2550: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * nbcode[Tvard[k][2]][codtabm(ij,k)];
2551: }else{
2552: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * Tqresult[nres][k];
2553: }
2554: }else{
2555: if(Dummy[Tvard[k][2]==0]){
2556: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][2]][codtabm(ij,k)] * Tqinvresult[nres][Tvard[k][1]];
2557: }else{
2558: cov[2+nagesqr+Tprod[k]]=Tqinvresult[nres][Tvard[k][1]]* Tqinvresult[nres][Tvard[k][2]];
2559: }
2560: }
1.234 brouard 2561: }
1.138 brouard 2562: /*printf("ij=%d cptcovprod=%d tvar=%d ", ij, cptcovprod, Tvar[1]);*/
2563: /*printf("ij=%d cov[3]=%lf cov[4]=%lf \n",ij, cov[3],cov[4]);*/
2564: /*printf("ij=%d cov[3]=%lf \n",ij, cov[3]);*/
1.145 brouard 2565: /* savm=pmij(pmmij,cov,ncovmodel,x,nlstate); */
2566: /* out=matprod2(newm, pmij(pmmij,cov,ncovmodel,x,nlstate),1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, oldm); /\* Bug Valgrind *\/ */
1.218 brouard 2567: /* age and covariate values of ij are in 'cov' */
1.142 brouard 2568: out=matprod2(newm, pmij(pmmij,cov,ncovmodel,x,nlstate),1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, oldm); /* Bug Valgrind */
1.138 brouard 2569:
1.126 brouard 2570: savm=oldm;
2571: oldm=newm;
1.209 brouard 2572:
2573: for(j=1; j<=nlstate; j++){
2574: max[j]=0.;
2575: min[j]=1.;
2576: }
2577: for(i=1;i<=nlstate;i++){
2578: sumnew=0;
2579: for(k=1; k<=ndeath; k++) sumnew+=newm[i][nlstate+k];
2580: for(j=1; j<=nlstate; j++){
2581: prlim[i][j]= newm[i][j]/(1-sumnew);
2582: max[j]=FMAX(max[j],prlim[i][j]);
2583: min[j]=FMIN(min[j],prlim[i][j]);
2584: }
2585: }
2586:
1.126 brouard 2587: maxmax=0.;
1.209 brouard 2588: for(j=1; j<=nlstate; j++){
2589: meandiff[j]=(max[j]-min[j])/(max[j]+min[j])*2.; /* mean difference for each column */
2590: maxmax=FMAX(maxmax,meandiff[j]);
2591: /* printf(" age= %d meandiff[%d]=%f, agefin=%d max[%d]=%f min[%d]=%f maxmax=%f\n", (int)age, j, meandiff[j],(int)agefin, j, max[j], j, min[j],maxmax); */
1.169 brouard 2592: } /* j loop */
1.203 brouard 2593: *ncvyear= (int)age- (int)agefin;
1.208 brouard 2594: /* printf("maxmax=%lf maxmin=%lf ncvloop=%d, age=%d, agefin=%d ncvyear=%d \n", maxmax, maxmin, ncvloop, (int)age, (int)agefin, *ncvyear); */
1.126 brouard 2595: if(maxmax < ftolpl){
1.209 brouard 2596: /* printf("maxmax=%lf ncvloop=%ld, age=%d, agefin=%d ncvyear=%d \n", maxmax, ncvloop, (int)age, (int)agefin, *ncvyear); */
2597: free_vector(min,1,nlstate);
2598: free_vector(max,1,nlstate);
2599: free_vector(meandiff,1,nlstate);
1.126 brouard 2600: return prlim;
2601: }
1.169 brouard 2602: } /* age loop */
1.208 brouard 2603: /* After some age loop it doesn't converge */
1.209 brouard 2604: printf("Warning: the stable prevalence at age %d did not converge with the required precision (%g > ftolpl=%g) within %.0f years. Try to lower 'ftolpl'. \n\
1.208 brouard 2605: Earliest age to start was %d-%d=%d, ncvloop=%d, ncvyear=%d\n", (int)age, maxmax, ftolpl, delaymax, (int)age, (int)delaymax, (int)agefin, ncvloop, *ncvyear);
1.209 brouard 2606: /* Try to lower 'ftol', for example from 1.e-8 to 6.e-9.\n", ftolpl, (int)age, (int)delaymax, (int)agefin, ncvloop, (int)age-(int)agefin); */
2607: free_vector(min,1,nlstate);
2608: free_vector(max,1,nlstate);
2609: free_vector(meandiff,1,nlstate);
1.208 brouard 2610:
1.169 brouard 2611: return prlim; /* should not reach here */
1.126 brouard 2612: }
2613:
1.217 brouard 2614:
2615: /**** Back Prevalence limit (stable or period prevalence) ****************/
2616:
1.218 brouard 2617: /* double **bprevalim(double **bprlim, double ***prevacurrent, int nlstate, double x[], double age, double ageminpar, double agemaxpar, double **oldm, double **savm, double **dnewm, double **doldm, double **dsavm, double ftolpl, int *ncvyear, int ij) */
2618: /* double **bprevalim(double **bprlim, double ***prevacurrent, int nlstate, double x[], double age, double **oldm, double **savm, double **dnewm, double **doldm, double **dsavm, double ftolpl, int *ncvyear, int ij) */
1.242 brouard 2619: double **bprevalim(double **bprlim, double ***prevacurrent, int nlstate, double x[], double age, double ftolpl, int *ncvyear, int ij, int nres)
1.217 brouard 2620: {
1.264 brouard 2621: /* Computes the prevalence limit in each live state at age x and for covariate combination ij (<=2**cptcoveff) by left multiplying the unit
1.217 brouard 2622: matrix by transitions matrix until convergence is reached with precision ftolpl */
2623: /* Wx= Wx-1 Px-1= Wx-2 Px-2 Px-1 = Wx-n Px-n ... Px-2 Px-1 I */
2624: /* Wx is row vector: population in state 1, population in state 2, population dead */
2625: /* or prevalence in state 1, prevalence in state 2, 0 */
2626: /* newm is the matrix after multiplications, its rows are identical at a factor */
2627: /* Initial matrix pimij */
2628: /* {0.85204250825084937, 0.13044499163996345, 0.017512500109187184, */
2629: /* 0.090851990222114765, 0.88271245433047185, 0.026435555447413338, */
2630: /* 0, 0 , 1} */
2631: /*
2632: * and after some iteration: */
2633: /* {0.45504275246439968, 0.42731458730878791, 0.11764266022681241, */
2634: /* 0.45201005341706885, 0.42865420071559901, 0.11933574586733192, */
2635: /* 0, 0 , 1} */
2636: /* And prevalence by suppressing the deaths are close to identical rows in prlim: */
2637: /* {0.51571254859325999, 0.4842874514067399, */
2638: /* 0.51326036147820708, 0.48673963852179264} */
2639: /* If we start from prlim again, prlim tends to a constant matrix */
2640:
2641: int i, ii,j,k;
1.247 brouard 2642: int first=0;
1.217 brouard 2643: double *min, *max, *meandiff, maxmax,sumnew=0.;
2644: /* double **matprod2(); */ /* test */
2645: double **out, cov[NCOVMAX+1], **bmij();
2646: double **newm;
1.218 brouard 2647: double **dnewm, **doldm, **dsavm; /* for use */
2648: double **oldm, **savm; /* for use */
2649:
1.217 brouard 2650: double agefin, delaymax=200. ; /* 100 Max number of years to converge */
2651: int ncvloop=0;
2652:
2653: min=vector(1,nlstate);
2654: max=vector(1,nlstate);
2655: meandiff=vector(1,nlstate);
2656:
1.218 brouard 2657: dnewm=ddnewms; doldm=ddoldms; dsavm=ddsavms;
2658: oldm=oldms; savm=savms;
2659:
2660: /* Starting with matrix unity */
2661: for (ii=1;ii<=nlstate+ndeath;ii++)
2662: for (j=1;j<=nlstate+ndeath;j++){
1.217 brouard 2663: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
2664: }
2665:
2666: cov[1]=1.;
2667:
2668: /* Even if hstepm = 1, at least one multiplication by the unit matrix */
2669: /* Start at agefin= age, computes the matrix of passage and loops decreasing agefin until convergence is reached */
1.218 brouard 2670: /* for(agefin=age+stepm/YEARM; agefin<=age+delaymax; agefin=agefin+stepm/YEARM){ /\* A changer en age *\/ */
2671: for(agefin=age; agefin<AGESUP; agefin=agefin+stepm/YEARM){ /* A changer en age */
1.217 brouard 2672: ncvloop++;
1.218 brouard 2673: newm=savm; /* oldm should be kept from previous iteration or unity at start */
2674: /* newm points to the allocated table savm passed by the function it can be written, savm could be reallocated */
1.217 brouard 2675: /* Covariates have to be included here again */
2676: cov[2]=agefin;
2677: if(nagesqr==1)
2678: cov[3]= agefin*agefin;;
1.242 brouard 2679: for (k=1; k<=nsd;k++) { /* For single dummy covariates only */
2680: /* Here comes the value of the covariate 'ij' after renumbering k with single dummy covariates */
2681: cov[2+nagesqr+TvarsDind[k]]=nbcode[TvarsD[k]][codtabm(ij,k)];
1.264 brouard 2682: /* printf("bprevalim Dummy agefin=%.0f combi=%d k=%d TvarsD[%d]=V%d TvarsDind[%d]=%d nbcode=%d cov[%d]=%lf codtabm(%d,Tvar[%d])=%d \n",agefin,ij,k, k, TvarsD[k],k,TvarsDind[k],nbcode[TvarsD[k]][codtabm(ij,k)],2+nagesqr+TvarsDind[k],cov[2+nagesqr+TvarsDind[k]], ij, k, codtabm(ij,k)); */
1.242 brouard 2683: }
2684: /* for (k=1; k<=cptcovn;k++) { */
2685: /* /\* cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(ij,Tvar[k])]; *\/ */
2686: /* cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(ij,k)]; */
2687: /* /\* printf("prevalim ij=%d k=%d Tvar[%d]=%d nbcode=%d cov=%lf codtabm(%d,Tvar[%d])=%d \n",ij,k, k, Tvar[k],nbcode[Tvar[k]][codtabm(ij,Tvar[k])],cov[2+k], ij, k, codtabm(ij,Tvar[k])]); *\/ */
2688: /* } */
2689: for (k=1; k<=nsq;k++) { /* For single varying covariates only */
2690: /* Here comes the value of quantitative after renumbering k with single quantitative covariates */
2691: cov[2+nagesqr+TvarsQind[k]]=Tqresult[nres][k];
2692: /* printf("prevalim Quantitative k=%d TvarsQind[%d]=%d, TvarsQ[%d]=V%d,Tqresult[%d][%d]=%f\n",k,k,TvarsQind[k],k,TvarsQ[k],nres,k,Tqresult[nres][k]); */
2693: }
2694: /* for (k=1; k<=cptcovage;k++) cov[2+nagesqr+Tage[k]]=nbcode[Tvar[k]][codtabm(ij,k)]*cov[2]; */
2695: /* for (k=1; k<=cptcovprod;k++) /\* Useless *\/ */
2696: /* /\* cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,Tvard[k][1])] * nbcode[Tvard[k][2]][codtabm(ij,Tvard[k][2])]; *\/ */
2697: /* cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * nbcode[Tvard[k][2]][codtabm(ij,k)]; */
2698: for (k=1; k<=cptcovage;k++){ /* For product with age */
2699: if(Dummy[Tvar[Tage[k]]]){
2700: cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2];
2701: } else{
2702: cov[2+nagesqr+Tage[k]]=Tqresult[nres][k];
2703: }
2704: /* printf("prevalim Age combi=%d k=%d Tage[%d]=V%d Tqresult[%d][%d]=%f\n",ij,k,k,Tage[k],nres,k,Tqresult[nres][k]); */
2705: }
2706: for (k=1; k<=cptcovprod;k++){ /* For product without age */
2707: /* printf("prevalim Prod ij=%d k=%d Tprod[%d]=%d Tvard[%d][1]=V%d, Tvard[%d][2]=V%d\n",ij,k,k,Tprod[k], k,Tvard[k][1], k,Tvard[k][2]); */
2708: if(Dummy[Tvard[k][1]==0]){
2709: if(Dummy[Tvard[k][2]==0]){
2710: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * nbcode[Tvard[k][2]][codtabm(ij,k)];
2711: }else{
2712: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * Tqresult[nres][k];
2713: }
2714: }else{
2715: if(Dummy[Tvard[k][2]==0]){
2716: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][2]][codtabm(ij,k)] * Tqinvresult[nres][Tvard[k][1]];
2717: }else{
2718: cov[2+nagesqr+Tprod[k]]=Tqinvresult[nres][Tvard[k][1]]* Tqinvresult[nres][Tvard[k][2]];
2719: }
2720: }
1.217 brouard 2721: }
2722:
2723: /*printf("ij=%d cptcovprod=%d tvar=%d ", ij, cptcovprod, Tvar[1]);*/
2724: /*printf("ij=%d cov[3]=%lf cov[4]=%lf \n",ij, cov[3],cov[4]);*/
2725: /*printf("ij=%d cov[3]=%lf \n",ij, cov[3]);*/
2726: /* savm=pmij(pmmij,cov,ncovmodel,x,nlstate); */
2727: /* out=matprod2(newm, pmij(pmmij,cov,ncovmodel,x,nlstate),1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, oldm); /\* Bug Valgrind *\/ */
1.218 brouard 2728: /* ij should be linked to the correct index of cov */
2729: /* age and covariate values ij are in 'cov', but we need to pass
2730: * ij for the observed prevalence at age and status and covariate
2731: * number: prevacurrent[(int)agefin][ii][ij]
2732: */
2733: /* out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, bmij(pmmij,cov,ncovmodel,x,nlstate,prevacurrent, ageminpar, agemaxpar, dnewm, doldm, dsavm,ij)); /\* Bug Valgrind *\/ */
2734: /* out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, bmij(pmmij,cov,ncovmodel,x,nlstate,prevacurrent, dnewm, doldm, dsavm,ij)); /\* Bug Valgrind *\/ */
2735: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, bmij(pmmij,cov,ncovmodel,x,nlstate,prevacurrent,ij)); /* Bug Valgrind */
1.217 brouard 2736: savm=oldm;
2737: oldm=newm;
2738: for(j=1; j<=nlstate; j++){
2739: max[j]=0.;
2740: min[j]=1.;
2741: }
2742: for(j=1; j<=nlstate; j++){
2743: for(i=1;i<=nlstate;i++){
1.234 brouard 2744: /* bprlim[i][j]= newm[i][j]/(1-sumnew); */
2745: bprlim[i][j]= newm[i][j];
2746: max[i]=FMAX(max[i],bprlim[i][j]); /* Max in line */
2747: min[i]=FMIN(min[i],bprlim[i][j]);
1.217 brouard 2748: }
2749: }
1.218 brouard 2750:
1.217 brouard 2751: maxmax=0.;
2752: for(i=1; i<=nlstate; i++){
2753: meandiff[i]=(max[i]-min[i])/(max[i]+min[i])*2.; /* mean difference for each column */
2754: maxmax=FMAX(maxmax,meandiff[i]);
2755: /* printf("Back age= %d meandiff[%d]=%f, agefin=%d max[%d]=%f min[%d]=%f maxmax=%f\n", (int)age, i, meandiff[i],(int)agefin, i, max[i], i, min[i],maxmax); */
2756: } /* j loop */
2757: *ncvyear= -( (int)age- (int)agefin);
1.218 brouard 2758: /* printf("Back maxmax=%lf ncvloop=%d, age=%d, agefin=%d ncvyear=%d \n", maxmax, ncvloop, (int)age, (int)agefin, *ncvyear);*/
1.217 brouard 2759: if(maxmax < ftolpl){
1.220 brouard 2760: /* printf("OK Back maxmax=%lf ncvloop=%d, age=%d, agefin=%d ncvyear=%d \n", maxmax, ncvloop, (int)age, (int)agefin, *ncvyear); */
1.217 brouard 2761: free_vector(min,1,nlstate);
2762: free_vector(max,1,nlstate);
2763: free_vector(meandiff,1,nlstate);
2764: return bprlim;
2765: }
2766: } /* age loop */
2767: /* After some age loop it doesn't converge */
1.247 brouard 2768: if(first){
2769: first=1;
2770: printf("Warning: the back stable prevalence at age %d did not converge with the required precision (%g > ftolpl=%g) within %.0f years. Try to lower 'ftolpl'. Others in log file only...\n\
2771: Oldest age to start was %d-%d=%d, ncvloop=%d, ncvyear=%d\n", (int)age, maxmax, ftolpl, delaymax, (int)age, (int)delaymax, (int)agefin, ncvloop, *ncvyear);
2772: }
2773: fprintf(ficlog,"Warning: the back stable prevalence at age %d did not converge with the required precision (%g > ftolpl=%g) within %.0f years. Try to lower 'ftolpl'. \n\
1.217 brouard 2774: Oldest age to start was %d-%d=%d, ncvloop=%d, ncvyear=%d\n", (int)age, maxmax, ftolpl, delaymax, (int)age, (int)delaymax, (int)agefin, ncvloop, *ncvyear);
2775: /* Try to lower 'ftol', for example from 1.e-8 to 6.e-9.\n", ftolpl, (int)age, (int)delaymax, (int)agefin, ncvloop, (int)age-(int)agefin); */
2776: free_vector(min,1,nlstate);
2777: free_vector(max,1,nlstate);
2778: free_vector(meandiff,1,nlstate);
2779:
2780: return bprlim; /* should not reach here */
2781: }
2782:
1.126 brouard 2783: /*************** transition probabilities ***************/
2784:
2785: double **pmij(double **ps, double *cov, int ncovmodel, double *x, int nlstate )
2786: {
1.138 brouard 2787: /* According to parameters values stored in x and the covariate's values stored in cov,
2788: computes the probability to be observed in state j being in state i by appying the
2789: model to the ncovmodel covariates (including constant and age).
2790: lnpijopii=ln(pij/pii)= aij+bij*age+cij*v1+dij*v2+... = sum_nc=1^ncovmodel xij(nc)*cov[nc]
2791: and, according on how parameters are entered, the position of the coefficient xij(nc) of the
2792: ncth covariate in the global vector x is given by the formula:
2793: j<i nc+((i-1)*(nlstate+ndeath-1)+j-1)*ncovmodel
2794: j>=i nc + ((i-1)*(nlstate+ndeath-1)+(j-2))*ncovmodel
2795: Computes ln(pij/pii) (lnpijopii), deduces pij/pii by exponentiation,
2796: sums on j different of i to get 1-pii/pii, deduces pii, and then all pij.
2797: Outputs ps[i][j] the probability to be observed in j being in j according to
2798: the values of the covariates cov[nc] and corresponding parameter values x[nc+shiftij]
2799: */
2800: double s1, lnpijopii;
1.126 brouard 2801: /*double t34;*/
1.164 brouard 2802: int i,j, nc, ii, jj;
1.126 brouard 2803:
1.223 brouard 2804: for(i=1; i<= nlstate; i++){
2805: for(j=1; j<i;j++){
2806: for (nc=1, lnpijopii=0.;nc <=ncovmodel; nc++){
2807: /*lnpijopii += param[i][j][nc]*cov[nc];*/
2808: lnpijopii += x[nc+((i-1)*(nlstate+ndeath-1)+j-1)*ncovmodel]*cov[nc];
2809: /* printf("Int j<i s1=%.17e, lnpijopii=%.17e\n",s1,lnpijopii); */
2810: }
2811: ps[i][j]=lnpijopii; /* In fact ln(pij/pii) */
2812: /* printf("s1=%.17e, lnpijopii=%.17e\n",s1,lnpijopii); */
2813: }
2814: for(j=i+1; j<=nlstate+ndeath;j++){
2815: for (nc=1, lnpijopii=0.;nc <=ncovmodel; nc++){
2816: /*lnpijopii += x[(i-1)*nlstate*ncovmodel+(j-2)*ncovmodel+nc+(i-1)*(ndeath-1)*ncovmodel]*cov[nc];*/
2817: lnpijopii += x[nc + ((i-1)*(nlstate+ndeath-1)+(j-2))*ncovmodel]*cov[nc];
2818: /* printf("Int j>i s1=%.17e, lnpijopii=%.17e %lx %lx\n",s1,lnpijopii,s1,lnpijopii); */
2819: }
2820: ps[i][j]=lnpijopii; /* In fact ln(pij/pii) */
2821: }
2822: }
1.218 brouard 2823:
1.223 brouard 2824: for(i=1; i<= nlstate; i++){
2825: s1=0;
2826: for(j=1; j<i; j++){
2827: s1+=exp(ps[i][j]); /* In fact sums pij/pii */
2828: /*printf("debug1 %d %d ps=%lf exp(ps)=%lf s1+=%lf\n",i,j,ps[i][j],exp(ps[i][j]),s1); */
2829: }
2830: for(j=i+1; j<=nlstate+ndeath; j++){
2831: s1+=exp(ps[i][j]); /* In fact sums pij/pii */
2832: /*printf("debug2 %d %d ps=%lf exp(ps)=%lf s1+=%lf\n",i,j,ps[i][j],exp(ps[i][j]),s1); */
2833: }
2834: /* s1= sum_{j<>i} pij/pii=(1-pii)/pii and thus pii is known from s1 */
2835: ps[i][i]=1./(s1+1.);
2836: /* Computing other pijs */
2837: for(j=1; j<i; j++)
2838: ps[i][j]= exp(ps[i][j])*ps[i][i];
2839: for(j=i+1; j<=nlstate+ndeath; j++)
2840: ps[i][j]= exp(ps[i][j])*ps[i][i];
2841: /* ps[i][nlstate+1]=1.-s1- ps[i][i];*/ /* Sum should be 1 */
2842: } /* end i */
1.218 brouard 2843:
1.223 brouard 2844: for(ii=nlstate+1; ii<= nlstate+ndeath; ii++){
2845: for(jj=1; jj<= nlstate+ndeath; jj++){
2846: ps[ii][jj]=0;
2847: ps[ii][ii]=1;
2848: }
2849: }
1.218 brouard 2850:
2851:
1.223 brouard 2852: /* for(ii=1; ii<= nlstate+ndeath; ii++){ */
2853: /* for(jj=1; jj<= nlstate+ndeath; jj++){ */
2854: /* printf(" pmij ps[%d][%d]=%lf ",ii,jj,ps[ii][jj]); */
2855: /* } */
2856: /* printf("\n "); */
2857: /* } */
2858: /* printf("\n ");printf("%lf ",cov[2]);*/
2859: /*
2860: for(i=1; i<= npar; i++) printf("%f ",x[i]);
1.218 brouard 2861: goto end;*/
1.223 brouard 2862: return ps;
1.126 brouard 2863: }
2864:
1.218 brouard 2865: /*************** backward transition probabilities ***************/
2866:
2867: /* double **bmij(double **ps, double *cov, int ncovmodel, double *x, int nlstate, double ***prevacurrent, double ageminpar, double agemaxpar, double ***dnewm, double **doldm, double **dsavm, int ij ) */
2868: /* double **bmij(double **ps, double *cov, int ncovmodel, double *x, int nlstate, double ***prevacurrent, double ***dnewm, double **doldm, double **dsavm, int ij ) */
2869: double **bmij(double **ps, double *cov, int ncovmodel, double *x, int nlstate, double ***prevacurrent, int ij )
2870: {
1.222 brouard 2871: /* Computes the backward probability at age agefin and covariate ij
2872: * and returns in **ps as well as **bmij.
2873: */
1.218 brouard 2874: int i, ii, j,k;
1.222 brouard 2875:
2876: double **out, **pmij();
2877: double sumnew=0.;
1.218 brouard 2878: double agefin;
1.222 brouard 2879:
2880: double **dnewm, **dsavm, **doldm;
2881: double **bbmij;
2882:
1.218 brouard 2883: doldm=ddoldms; /* global pointers */
1.222 brouard 2884: dnewm=ddnewms;
2885: dsavm=ddsavms;
2886:
2887: agefin=cov[2];
2888: /* bmij *//* age is cov[2], ij is included in cov, but we need for
2889: the observed prevalence (with this covariate ij) */
2890: dsavm=pmij(pmmij,cov,ncovmodel,x,nlstate);
2891: /* We do have the matrix Px in savm and we need pij */
2892: for (j=1;j<=nlstate+ndeath;j++){
2893: sumnew=0.; /* w1 p11 + w2 p21 only on live states */
2894: for (ii=1;ii<=nlstate;ii++){
2895: sumnew+=dsavm[ii][j]*prevacurrent[(int)agefin][ii][ij];
2896: } /* sumnew is (N11+N21)/N..= N.1/N.. = sum on i of w_i pij */
2897: for (ii=1;ii<=nlstate+ndeath;ii++){
2898: if(sumnew >= 1.e-10){
2899: /* if(agefin >= agemaxpar && agefin <= agemaxpar+stepm/YEARM){ */
2900: /* doldm[ii][j]=(ii==j ? 1./sumnew : 0.0); */
2901: /* }else if(agefin >= agemaxpar+stepm/YEARM){ */
2902: /* doldm[ii][j]=(ii==j ? 1./sumnew : 0.0); */
2903: /* }else */
2904: doldm[ii][j]=(ii==j ? 1./sumnew : 0.0);
2905: }else{
1.242 brouard 2906: ;
2907: /* printf("ii=%d, i=%d, doldm=%lf dsavm=%lf, probs=%lf, sumnew=%lf,agefin=%d\n",ii,j,doldm[ii][j],dsavm[ii][j],prevacurrent[(int)agefin][ii][ij],sumnew, (int)agefin); */
1.222 brouard 2908: }
2909: } /*End ii */
2910: } /* End j, At the end doldm is diag[1/(w_1p1i+w_2 p2i)] */
2911: /* left Product of this diag matrix by dsavm=Px (newm=dsavm*doldm) */
2912: bbmij=matprod2(dnewm, dsavm,1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, doldm); /* Bug Valgrind */
2913: /* dsavm=doldm; /\* dsavm is now diag [1/(w_1p1i+w_2 p2i)] but can be overwritten*\/ */
2914: /* doldm=dnewm; /\* doldm is now Px * diag [1/(w_1p1i+w_2 p2i)] *\/ */
2915: /* dnewm=dsavm; /\* doldm is now Px * diag [1/(w_1p1i+w_2 p2i)] *\/ */
2916: /* left Product of this matrix by diag matrix of prevalences (savm) */
2917: for (j=1;j<=nlstate+ndeath;j++){
2918: for (ii=1;ii<=nlstate+ndeath;ii++){
2919: dsavm[ii][j]=(ii==j ? prevacurrent[(int)agefin][ii][ij] : 0.0);
2920: }
2921: } /* End j, At the end oldm is diag[1/(w_1p1i+w_2 p2i)] */
2922: ps=matprod2(doldm, dsavm,1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, dnewm); /* Bug Valgrind */
2923: /* newm or out is now diag[w_i] * Px * diag [1/(w_1p1i+w_2 p2i)] */
2924: /* end bmij */
2925: return ps;
1.218 brouard 2926: }
1.217 brouard 2927: /*************** transition probabilities ***************/
2928:
1.218 brouard 2929: double **bpmij(double **ps, double *cov, int ncovmodel, double *x, int nlstate )
1.217 brouard 2930: {
2931: /* According to parameters values stored in x and the covariate's values stored in cov,
2932: computes the probability to be observed in state j being in state i by appying the
2933: model to the ncovmodel covariates (including constant and age).
2934: lnpijopii=ln(pij/pii)= aij+bij*age+cij*v1+dij*v2+... = sum_nc=1^ncovmodel xij(nc)*cov[nc]
2935: and, according on how parameters are entered, the position of the coefficient xij(nc) of the
2936: ncth covariate in the global vector x is given by the formula:
2937: j<i nc+((i-1)*(nlstate+ndeath-1)+j-1)*ncovmodel
2938: j>=i nc + ((i-1)*(nlstate+ndeath-1)+(j-2))*ncovmodel
2939: Computes ln(pij/pii) (lnpijopii), deduces pij/pii by exponentiation,
2940: sums on j different of i to get 1-pii/pii, deduces pii, and then all pij.
2941: Outputs ps[i][j] the probability to be observed in j being in j according to
2942: the values of the covariates cov[nc] and corresponding parameter values x[nc+shiftij]
2943: */
2944: double s1, lnpijopii;
2945: /*double t34;*/
2946: int i,j, nc, ii, jj;
2947:
1.234 brouard 2948: for(i=1; i<= nlstate; i++){
2949: for(j=1; j<i;j++){
2950: for (nc=1, lnpijopii=0.;nc <=ncovmodel; nc++){
2951: /*lnpijopii += param[i][j][nc]*cov[nc];*/
2952: lnpijopii += x[nc+((i-1)*(nlstate+ndeath-1)+j-1)*ncovmodel]*cov[nc];
2953: /* printf("Int j<i s1=%.17e, lnpijopii=%.17e\n",s1,lnpijopii); */
2954: }
2955: ps[i][j]=lnpijopii; /* In fact ln(pij/pii) */
2956: /* printf("s1=%.17e, lnpijopii=%.17e\n",s1,lnpijopii); */
2957: }
2958: for(j=i+1; j<=nlstate+ndeath;j++){
2959: for (nc=1, lnpijopii=0.;nc <=ncovmodel; nc++){
2960: /*lnpijopii += x[(i-1)*nlstate*ncovmodel+(j-2)*ncovmodel+nc+(i-1)*(ndeath-1)*ncovmodel]*cov[nc];*/
2961: lnpijopii += x[nc + ((i-1)*(nlstate+ndeath-1)+(j-2))*ncovmodel]*cov[nc];
2962: /* printf("Int j>i s1=%.17e, lnpijopii=%.17e %lx %lx\n",s1,lnpijopii,s1,lnpijopii); */
2963: }
2964: ps[i][j]=lnpijopii; /* In fact ln(pij/pii) */
2965: }
2966: }
2967:
2968: for(i=1; i<= nlstate; i++){
2969: s1=0;
2970: for(j=1; j<i; j++){
2971: s1+=exp(ps[i][j]); /* In fact sums pij/pii */
2972: /*printf("debug1 %d %d ps=%lf exp(ps)=%lf s1+=%lf\n",i,j,ps[i][j],exp(ps[i][j]),s1); */
2973: }
2974: for(j=i+1; j<=nlstate+ndeath; j++){
2975: s1+=exp(ps[i][j]); /* In fact sums pij/pii */
2976: /*printf("debug2 %d %d ps=%lf exp(ps)=%lf s1+=%lf\n",i,j,ps[i][j],exp(ps[i][j]),s1); */
2977: }
2978: /* s1= sum_{j<>i} pij/pii=(1-pii)/pii and thus pii is known from s1 */
2979: ps[i][i]=1./(s1+1.);
2980: /* Computing other pijs */
2981: for(j=1; j<i; j++)
2982: ps[i][j]= exp(ps[i][j])*ps[i][i];
2983: for(j=i+1; j<=nlstate+ndeath; j++)
2984: ps[i][j]= exp(ps[i][j])*ps[i][i];
2985: /* ps[i][nlstate+1]=1.-s1- ps[i][i];*/ /* Sum should be 1 */
2986: } /* end i */
2987:
2988: for(ii=nlstate+1; ii<= nlstate+ndeath; ii++){
2989: for(jj=1; jj<= nlstate+ndeath; jj++){
2990: ps[ii][jj]=0;
2991: ps[ii][ii]=1;
2992: }
2993: }
2994: /* Added for backcast */ /* Transposed matrix too */
2995: for(jj=1; jj<= nlstate+ndeath; jj++){
2996: s1=0.;
2997: for(ii=1; ii<= nlstate+ndeath; ii++){
2998: s1+=ps[ii][jj];
2999: }
3000: for(ii=1; ii<= nlstate; ii++){
3001: ps[ii][jj]=ps[ii][jj]/s1;
3002: }
3003: }
3004: /* Transposition */
3005: for(jj=1; jj<= nlstate+ndeath; jj++){
3006: for(ii=jj; ii<= nlstate+ndeath; ii++){
3007: s1=ps[ii][jj];
3008: ps[ii][jj]=ps[jj][ii];
3009: ps[jj][ii]=s1;
3010: }
3011: }
3012: /* for(ii=1; ii<= nlstate+ndeath; ii++){ */
3013: /* for(jj=1; jj<= nlstate+ndeath; jj++){ */
3014: /* printf(" pmij ps[%d][%d]=%lf ",ii,jj,ps[ii][jj]); */
3015: /* } */
3016: /* printf("\n "); */
3017: /* } */
3018: /* printf("\n ");printf("%lf ",cov[2]);*/
3019: /*
3020: for(i=1; i<= npar; i++) printf("%f ",x[i]);
3021: goto end;*/
3022: return ps;
1.217 brouard 3023: }
3024:
3025:
1.126 brouard 3026: /**************** Product of 2 matrices ******************/
3027:
1.145 brouard 3028: double **matprod2(double **out, double **in,int nrl, int nrh, int ncl, int nch, int ncolol, int ncoloh, double **b)
1.126 brouard 3029: {
3030: /* Computes the matrix product of in(1,nrh-nrl+1)(1,nch-ncl+1) times
3031: b(1,nch-ncl+1)(1,ncoloh-ncolol+1) into out(...) */
3032: /* in, b, out are matrice of pointers which should have been initialized
3033: before: only the contents of out is modified. The function returns
3034: a pointer to pointers identical to out */
1.145 brouard 3035: int i, j, k;
1.126 brouard 3036: for(i=nrl; i<= nrh; i++)
1.145 brouard 3037: for(k=ncolol; k<=ncoloh; k++){
3038: out[i][k]=0.;
3039: for(j=ncl; j<=nch; j++)
3040: out[i][k] +=in[i][j]*b[j][k];
3041: }
1.126 brouard 3042: return out;
3043: }
3044:
3045:
3046: /************* Higher Matrix Product ***************/
3047:
1.235 brouard 3048: double ***hpxij(double ***po, int nhstepm, double age, int hstepm, double *x, int nlstate, int stepm, double **oldm, double **savm, int ij, int nres )
1.126 brouard 3049: {
1.218 brouard 3050: /* Computes the transition matrix starting at age 'age' and combination of covariate values corresponding to ij over
1.126 brouard 3051: 'nhstepm*hstepm*stepm' months (i.e. until
3052: age (in years) age+nhstepm*hstepm*stepm/12) by multiplying
3053: nhstepm*hstepm matrices.
3054: Output is stored in matrix po[i][j][h] for h every 'hstepm' step
3055: (typically every 2 years instead of every month which is too big
3056: for the memory).
3057: Model is determined by parameters x and covariates have to be
3058: included manually here.
3059:
3060: */
3061:
3062: int i, j, d, h, k;
1.131 brouard 3063: double **out, cov[NCOVMAX+1];
1.126 brouard 3064: double **newm;
1.187 brouard 3065: double agexact;
1.214 brouard 3066: double agebegin, ageend;
1.126 brouard 3067:
3068: /* Hstepm could be zero and should return the unit matrix */
3069: for (i=1;i<=nlstate+ndeath;i++)
3070: for (j=1;j<=nlstate+ndeath;j++){
3071: oldm[i][j]=(i==j ? 1.0 : 0.0);
3072: po[i][j][0]=(i==j ? 1.0 : 0.0);
3073: }
3074: /* Even if hstepm = 1, at least one multiplication by the unit matrix */
3075: for(h=1; h <=nhstepm; h++){
3076: for(d=1; d <=hstepm; d++){
3077: newm=savm;
3078: /* Covariates have to be included here again */
3079: cov[1]=1.;
1.214 brouard 3080: agexact=age+((h-1)*hstepm + (d-1))*stepm/YEARM; /* age just before transition */
1.187 brouard 3081: cov[2]=agexact;
3082: if(nagesqr==1)
1.227 brouard 3083: cov[3]= agexact*agexact;
1.235 brouard 3084: for (k=1; k<=nsd;k++) { /* For single dummy covariates only */
3085: /* Here comes the value of the covariate 'ij' after renumbering k with single dummy covariates */
3086: cov[2+nagesqr+TvarsDind[k]]=nbcode[TvarsD[k]][codtabm(ij,k)];
3087: /* printf("hpxij Dummy combi=%d k=%d TvarsD[%d]=V%d TvarsDind[%d]=%d nbcode=%d cov=%lf codtabm(%d,Tvar[%d])=%d \n",ij,k, k, TvarsD[k],k,TvarsDind[k],nbcode[TvarsD[k]][codtabm(ij,k)],cov[2+nagesqr+TvarsDind[k]], ij, k, codtabm(ij,k)); */
3088: }
3089: for (k=1; k<=nsq;k++) { /* For single varying covariates only */
3090: /* Here comes the value of quantitative after renumbering k with single quantitative covariates */
3091: cov[2+nagesqr+TvarsQind[k]]=Tqresult[nres][k];
3092: /* printf("hPxij Quantitative k=%d TvarsQind[%d]=%d, TvarsQ[%d]=V%d,Tqresult[%d][%d]=%f\n",k,k,TvarsQind[k],k,TvarsQ[k],nres,k,Tqresult[nres][k]); */
3093: }
3094: for (k=1; k<=cptcovage;k++){
3095: if(Dummy[Tvar[Tage[k]]]){
3096: cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2];
3097: } else{
3098: cov[2+nagesqr+Tage[k]]=Tqresult[nres][k];
3099: }
3100: /* printf("hPxij Age combi=%d k=%d Tage[%d]=V%d Tqresult[%d][%d]=%f\n",ij,k,k,Tage[k],nres,k,Tqresult[nres][k]); */
3101: }
3102: for (k=1; k<=cptcovprod;k++){ /* */
3103: /* printf("hPxij Prod ij=%d k=%d Tprod[%d]=%d Tvard[%d][1]=V%d, Tvard[%d][2]=V%d\n",ij,k,k,Tprod[k], k,Tvard[k][1], k,Tvard[k][2]); */
3104: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * nbcode[Tvard[k][2]][codtabm(ij,k)];
3105: }
3106: /* for (k=1; k<=cptcovn;k++) */
3107: /* cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(ij,k)]; */
3108: /* for (k=1; k<=cptcovage;k++) /\* Should start at cptcovn+1 *\/ */
3109: /* cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2]; */
3110: /* for (k=1; k<=cptcovprod;k++) /\* Useless because included in cptcovn *\/ */
3111: /* cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)]*nbcode[Tvard[k][2]][codtabm(ij,k)]; */
1.227 brouard 3112:
3113:
1.126 brouard 3114: /*printf("hxi cptcov=%d cptcode=%d\n",cptcov,cptcode);*/
3115: /*printf("h=%d d=%d age=%f cov=%f\n",h,d,age,cov[2]);*/
1.218 brouard 3116: /* right multiplication of oldm by the current matrix */
1.126 brouard 3117: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath,
3118: pmij(pmmij,cov,ncovmodel,x,nlstate));
1.217 brouard 3119: /* if((int)age == 70){ */
3120: /* printf(" Forward hpxij age=%d agexact=%f d=%d nhstepm=%d hstepm=%d\n", (int) age, agexact, d, nhstepm, hstepm); */
3121: /* for(i=1; i<=nlstate+ndeath; i++) { */
3122: /* printf("%d pmmij ",i); */
3123: /* for(j=1;j<=nlstate+ndeath;j++) { */
3124: /* printf("%f ",pmmij[i][j]); */
3125: /* } */
3126: /* printf(" oldm "); */
3127: /* for(j=1;j<=nlstate+ndeath;j++) { */
3128: /* printf("%f ",oldm[i][j]); */
3129: /* } */
3130: /* printf("\n"); */
3131: /* } */
3132: /* } */
1.126 brouard 3133: savm=oldm;
3134: oldm=newm;
3135: }
3136: for(i=1; i<=nlstate+ndeath; i++)
3137: for(j=1;j<=nlstate+ndeath;j++) {
1.218 brouard 3138: po[i][j][h]=newm[i][j];
3139: /*if(h==nhstepm) printf("po[%d][%d][%d]=%f ",i,j,h,po[i][j][h]);*/
1.126 brouard 3140: }
1.128 brouard 3141: /*printf("h=%d ",h);*/
1.126 brouard 3142: } /* end h */
1.218 brouard 3143: /* printf("\n H=%d \n",h); */
1.126 brouard 3144: return po;
3145: }
3146:
1.217 brouard 3147: /************* Higher Back Matrix Product ***************/
1.218 brouard 3148: /* double ***hbxij(double ***po, int nhstepm, double age, int hstepm, double *x, double ***prevacurrent, int nlstate, int stepm, double **oldm, double **savm, double **dnewm, double **doldm, double **dsavm, int ij ) */
1.222 brouard 3149: double ***hbxij(double ***po, int nhstepm, double age, int hstepm, double *x, double ***prevacurrent, int nlstate, int stepm, int ij )
1.217 brouard 3150: {
1.218 brouard 3151: /* Computes the transition matrix starting at age 'age' over
1.217 brouard 3152: 'nhstepm*hstepm*stepm' months (i.e. until
1.218 brouard 3153: age (in years) age+nhstepm*hstepm*stepm/12) by multiplying
3154: nhstepm*hstepm matrices.
3155: Output is stored in matrix po[i][j][h] for h every 'hstepm' step
3156: (typically every 2 years instead of every month which is too big
1.217 brouard 3157: for the memory).
1.218 brouard 3158: Model is determined by parameters x and covariates have to be
3159: included manually here.
1.217 brouard 3160:
1.222 brouard 3161: */
1.217 brouard 3162:
3163: int i, j, d, h, k;
3164: double **out, cov[NCOVMAX+1];
3165: double **newm;
3166: double agexact;
3167: double agebegin, ageend;
1.222 brouard 3168: double **oldm, **savm;
1.217 brouard 3169:
1.222 brouard 3170: oldm=oldms;savm=savms;
1.217 brouard 3171: /* Hstepm could be zero and should return the unit matrix */
3172: for (i=1;i<=nlstate+ndeath;i++)
3173: for (j=1;j<=nlstate+ndeath;j++){
3174: oldm[i][j]=(i==j ? 1.0 : 0.0);
3175: po[i][j][0]=(i==j ? 1.0 : 0.0);
3176: }
3177: /* Even if hstepm = 1, at least one multiplication by the unit matrix */
3178: for(h=1; h <=nhstepm; h++){
3179: for(d=1; d <=hstepm; d++){
3180: newm=savm;
3181: /* Covariates have to be included here again */
3182: cov[1]=1.;
3183: agexact=age-((h-1)*hstepm + (d-1))*stepm/YEARM; /* age just before transition */
3184: /* agexact=age+((h-1)*hstepm + (d-1))*stepm/YEARM; /\* age just before transition *\/ */
3185: cov[2]=agexact;
3186: if(nagesqr==1)
1.222 brouard 3187: cov[3]= agexact*agexact;
1.218 brouard 3188: for (k=1; k<=cptcovn;k++)
1.222 brouard 3189: cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(ij,k)];
3190: /* cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(ij,Tvar[k])]; */
1.217 brouard 3191: for (k=1; k<=cptcovage;k++) /* Should start at cptcovn+1 */
1.222 brouard 3192: /* cov[2+Tage[k]]=cov[2+Tage[k]]*cov[2]; */
3193: cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2];
3194: /* cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,Tvar[Tage[k]])]*cov[2]; */
1.217 brouard 3195: for (k=1; k<=cptcovprod;k++) /* Useless because included in cptcovn */
1.222 brouard 3196: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)]*nbcode[Tvard[k][2]][codtabm(ij,k)];
3197: /* cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,Tvard[k][1])]*nbcode[Tvard[k][2]][codtabm(ij,Tvard[k][2])]; */
1.218 brouard 3198:
3199:
1.217 brouard 3200: /*printf("hxi cptcov=%d cptcode=%d\n",cptcov,cptcode);*/
3201: /*printf("h=%d d=%d age=%f cov=%f\n",h,d,age,cov[2]);*/
1.218 brouard 3202: /* Careful transposed matrix */
1.222 brouard 3203: /* age is in cov[2] */
1.218 brouard 3204: /* out=matprod2(newm, bmij(pmmij,cov,ncovmodel,x,nlstate,prevacurrent, dnewm, doldm, dsavm,ij),\ */
1.222 brouard 3205: /* 1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, oldm); */
1.218 brouard 3206: out=matprod2(newm, bmij(pmmij,cov,ncovmodel,x,nlstate,prevacurrent,ij),\
1.222 brouard 3207: 1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, oldm);
1.217 brouard 3208: /* if((int)age == 70){ */
3209: /* printf(" Backward hbxij age=%d agexact=%f d=%d nhstepm=%d hstepm=%d\n", (int) age, agexact, d, nhstepm, hstepm); */
3210: /* for(i=1; i<=nlstate+ndeath; i++) { */
3211: /* printf("%d pmmij ",i); */
3212: /* for(j=1;j<=nlstate+ndeath;j++) { */
3213: /* printf("%f ",pmmij[i][j]); */
3214: /* } */
3215: /* printf(" oldm "); */
3216: /* for(j=1;j<=nlstate+ndeath;j++) { */
3217: /* printf("%f ",oldm[i][j]); */
3218: /* } */
3219: /* printf("\n"); */
3220: /* } */
3221: /* } */
3222: savm=oldm;
3223: oldm=newm;
3224: }
3225: for(i=1; i<=nlstate+ndeath; i++)
3226: for(j=1;j<=nlstate+ndeath;j++) {
1.222 brouard 3227: po[i][j][h]=newm[i][j];
3228: /*if(h==nhstepm) printf("po[%d][%d][%d]=%f ",i,j,h,po[i][j][h]);*/
1.217 brouard 3229: }
3230: /*printf("h=%d ",h);*/
3231: } /* end h */
1.222 brouard 3232: /* printf("\n H=%d \n",h); */
1.217 brouard 3233: return po;
3234: }
3235:
3236:
1.162 brouard 3237: #ifdef NLOPT
3238: double myfunc(unsigned n, const double *p1, double *grad, void *pd){
3239: double fret;
3240: double *xt;
3241: int j;
3242: myfunc_data *d2 = (myfunc_data *) pd;
3243: /* xt = (p1-1); */
3244: xt=vector(1,n);
3245: for (j=1;j<=n;j++) xt[j]=p1[j-1]; /* xt[1]=p1[0] */
3246:
3247: fret=(d2->function)(xt); /* p xt[1]@8 is fine */
3248: /* fret=(*func)(xt); /\* p xt[1]@8 is fine *\/ */
3249: printf("Function = %.12lf ",fret);
3250: for (j=1;j<=n;j++) printf(" %d %.8lf", j, xt[j]);
3251: printf("\n");
3252: free_vector(xt,1,n);
3253: return fret;
3254: }
3255: #endif
1.126 brouard 3256:
3257: /*************** log-likelihood *************/
3258: double func( double *x)
3259: {
1.226 brouard 3260: int i, ii, j, k, mi, d, kk;
3261: int ioffset=0;
3262: double l, ll[NLSTATEMAX+1], cov[NCOVMAX+1];
3263: double **out;
3264: double lli; /* Individual log likelihood */
3265: int s1, s2;
1.228 brouard 3266: int iv=0, iqv=0, itv=0, iqtv=0 ; /* Index of varying covariate, fixed quantitative cov, time varying covariate, quantitative time varying covariate */
1.226 brouard 3267: double bbh, survp;
3268: long ipmx;
3269: double agexact;
3270: /*extern weight */
3271: /* We are differentiating ll according to initial status */
3272: /* for (i=1;i<=npar;i++) printf("%f ", x[i]);*/
3273: /*for(i=1;i<imx;i++)
3274: printf(" %d\n",s[4][i]);
3275: */
1.162 brouard 3276:
1.226 brouard 3277: ++countcallfunc;
1.162 brouard 3278:
1.226 brouard 3279: cov[1]=1.;
1.126 brouard 3280:
1.226 brouard 3281: for(k=1; k<=nlstate; k++) ll[k]=0.;
1.224 brouard 3282: ioffset=0;
1.226 brouard 3283: if(mle==1){
3284: for (i=1,ipmx=0, sw=0.; i<=imx; i++){
3285: /* Computes the values of the ncovmodel covariates of the model
3286: depending if the covariates are fixed or varying (age dependent) and stores them in cov[]
3287: Then computes with function pmij which return a matrix p[i][j] giving the elementary probability
3288: to be observed in j being in i according to the model.
3289: */
1.243 brouard 3290: ioffset=2+nagesqr ;
1.233 brouard 3291: /* Fixed */
1.234 brouard 3292: for (k=1; k<=ncovf;k++){ /* Simple and product fixed covariates without age* products */
3293: cov[ioffset+TvarFind[k]]=covar[Tvar[TvarFind[k]]][i];/* V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1, only V1 is fixed (k=6)*/
3294: }
1.226 brouard 3295: /* In model V2+V1*V4+age*V3+V3*V2 Tvar[1] is V2, Tvar[2=V1*V4]
3296: is 6, Tvar[3=age*V3] should not be computed because of age Tvar[4=V3*V2]
3297: has been calculated etc */
3298: /* For an individual i, wav[i] gives the number of effective waves */
3299: /* We compute the contribution to Likelihood of each effective transition
3300: mw[mi][i] is real wave of the mi th effectve wave */
3301: /* Then statuses are computed at each begin and end of an effective wave s1=s[ mw[mi][i] ][i];
3302: s2=s[mw[mi+1][i]][i];
3303: And the iv th varying covariate is the cotvar[mw[mi+1][i]][iv][i]
3304: But if the variable is not in the model TTvar[iv] is the real variable effective in the model:
3305: meaning that decodemodel should be used cotvar[mw[mi+1][i]][TTvar[iv]][i]
3306: */
3307: for(mi=1; mi<= wav[i]-1; mi++){
1.234 brouard 3308: for(k=1; k <= ncovv ; k++){ /* Varying covariates (single and product but no age )*/
1.242 brouard 3309: /* cov[ioffset+TvarVind[k]]=cotvar[mw[mi][i]][Tvar[TvarVind[k]]][i]; */
3310: cov[ioffset+TvarVind[k]]=cotvar[mw[mi][i]][Tvar[TvarVind[k]]-ncovcol-nqv][i];
1.234 brouard 3311: }
3312: for (ii=1;ii<=nlstate+ndeath;ii++)
3313: for (j=1;j<=nlstate+ndeath;j++){
3314: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
3315: savm[ii][j]=(ii==j ? 1.0 : 0.0);
3316: }
3317: for(d=0; d<dh[mi][i]; d++){
3318: newm=savm;
3319: agexact=agev[mw[mi][i]][i]+d*stepm/YEARM;
3320: cov[2]=agexact;
3321: if(nagesqr==1)
3322: cov[3]= agexact*agexact; /* Should be changed here */
3323: for (kk=1; kk<=cptcovage;kk++) {
1.242 brouard 3324: if(!FixedV[Tvar[Tage[kk]]])
1.234 brouard 3325: cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact; /* Tage[kk] gives the data-covariate associated with age */
1.242 brouard 3326: else
3327: cov[Tage[kk]+2+nagesqr]=cotvar[mw[mi][i]][Tvar[Tage[kk]]-ncovcol-nqv][i]*agexact;
1.234 brouard 3328: }
3329: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,
3330: 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate));
3331: savm=oldm;
3332: oldm=newm;
3333: } /* end mult */
3334:
3335: /*lli=log(out[s[mw[mi][i]][i]][s[mw[mi+1][i]][i]]);*/ /* Original formula */
3336: /* But now since version 0.9 we anticipate for bias at large stepm.
3337: * If stepm is larger than one month (smallest stepm) and if the exact delay
3338: * (in months) between two waves is not a multiple of stepm, we rounded to
3339: * the nearest (and in case of equal distance, to the lowest) interval but now
3340: * we keep into memory the bias bh[mi][i] and also the previous matrix product
3341: * (i.e to dh[mi][i]-1) saved in 'savm'. Then we inter(extra)polate the
3342: * probability in order to take into account the bias as a fraction of the way
1.231 brouard 3343: * from savm to out if bh is negative or even beyond if bh is positive. bh varies
3344: * -stepm/2 to stepm/2 .
3345: * For stepm=1 the results are the same as for previous versions of Imach.
3346: * For stepm > 1 the results are less biased than in previous versions.
3347: */
1.234 brouard 3348: s1=s[mw[mi][i]][i];
3349: s2=s[mw[mi+1][i]][i];
3350: bbh=(double)bh[mi][i]/(double)stepm;
3351: /* bias bh is positive if real duration
3352: * is higher than the multiple of stepm and negative otherwise.
3353: */
3354: /* lli= (savm[s1][s2]>1.e-8 ?(1.+bbh)*log(out[s1][s2])- bbh*log(savm[s1][s2]):log((1.+bbh)*out[s1][s2]));*/
3355: if( s2 > nlstate){
3356: /* i.e. if s2 is a death state and if the date of death is known
3357: then the contribution to the likelihood is the probability to
3358: die between last step unit time and current step unit time,
3359: which is also equal to probability to die before dh
3360: minus probability to die before dh-stepm .
3361: In version up to 0.92 likelihood was computed
3362: as if date of death was unknown. Death was treated as any other
3363: health state: the date of the interview describes the actual state
3364: and not the date of a change in health state. The former idea was
3365: to consider that at each interview the state was recorded
3366: (healthy, disable or death) and IMaCh was corrected; but when we
3367: introduced the exact date of death then we should have modified
3368: the contribution of an exact death to the likelihood. This new
3369: contribution is smaller and very dependent of the step unit
3370: stepm. It is no more the probability to die between last interview
3371: and month of death but the probability to survive from last
3372: interview up to one month before death multiplied by the
3373: probability to die within a month. Thanks to Chris
3374: Jackson for correcting this bug. Former versions increased
3375: mortality artificially. The bad side is that we add another loop
3376: which slows down the processing. The difference can be up to 10%
3377: lower mortality.
3378: */
3379: /* If, at the beginning of the maximization mostly, the
3380: cumulative probability or probability to be dead is
3381: constant (ie = 1) over time d, the difference is equal to
3382: 0. out[s1][3] = savm[s1][3]: probability, being at state
3383: s1 at precedent wave, to be dead a month before current
3384: wave is equal to probability, being at state s1 at
3385: precedent wave, to be dead at mont of the current
3386: wave. Then the observed probability (that this person died)
3387: is null according to current estimated parameter. In fact,
3388: it should be very low but not zero otherwise the log go to
3389: infinity.
3390: */
1.183 brouard 3391: /* #ifdef INFINITYORIGINAL */
3392: /* lli=log(out[s1][s2] - savm[s1][s2]); */
3393: /* #else */
3394: /* if ((out[s1][s2] - savm[s1][s2]) < mytinydouble) */
3395: /* lli=log(mytinydouble); */
3396: /* else */
3397: /* lli=log(out[s1][s2] - savm[s1][s2]); */
3398: /* #endif */
1.226 brouard 3399: lli=log(out[s1][s2] - savm[s1][s2]);
1.216 brouard 3400:
1.226 brouard 3401: } else if ( s2==-1 ) { /* alive */
3402: for (j=1,survp=0. ; j<=nlstate; j++)
3403: survp += (1.+bbh)*out[s1][j]- bbh*savm[s1][j];
3404: /*survp += out[s1][j]; */
3405: lli= log(survp);
3406: }
3407: else if (s2==-4) {
3408: for (j=3,survp=0. ; j<=nlstate; j++)
3409: survp += (1.+bbh)*out[s1][j]- bbh*savm[s1][j];
3410: lli= log(survp);
3411: }
3412: else if (s2==-5) {
3413: for (j=1,survp=0. ; j<=2; j++)
3414: survp += (1.+bbh)*out[s1][j]- bbh*savm[s1][j];
3415: lli= log(survp);
3416: }
3417: else{
3418: lli= log((1.+bbh)*out[s1][s2]- bbh*savm[s1][s2]); /* linear interpolation */
3419: /* lli= (savm[s1][s2]>(double)1.e-8 ?log((1.+bbh)*out[s1][s2]- bbh*(savm[s1][s2])):log((1.+bbh)*out[s1][s2]));*/ /* linear interpolation */
3420: }
3421: /*lli=(1.+bbh)*log(out[s1][s2])- bbh*log(savm[s1][s2]);*/
3422: /*if(lli ==000.0)*/
3423: /*printf("bbh= %f lli=%f savm=%f out=%f %d\n",bbh,lli,savm[s1][s2], out[s[mw[mi][i]][i]][s[mw[mi+1][i]][i]],i); */
3424: ipmx +=1;
3425: sw += weight[i];
3426: ll[s[mw[mi][i]][i]] += 2*weight[i]*lli;
3427: /* if (lli < log(mytinydouble)){ */
3428: /* printf("Close to inf lli = %.10lf < %.10lf i= %d mi= %d, s[%d][i]=%d s1=%d s2=%d\n", lli,log(mytinydouble), i, mi,mw[mi][i], s[mw[mi][i]][i], s1,s2); */
3429: /* fprintf(ficlog,"Close to inf lli = %.10lf i= %d mi= %d, s[mw[mi][i]][i]=%d\n", lli, i, mi,s[mw[mi][i]][i]); */
3430: /* } */
3431: } /* end of wave */
3432: } /* end of individual */
3433: } else if(mle==2){
3434: for (i=1,ipmx=0, sw=0.; i<=imx; i++){
3435: for (k=1; k<=cptcovn;k++) cov[2+nagesqr+k]=covar[Tvar[k]][i];
3436: for(mi=1; mi<= wav[i]-1; mi++){
3437: for (ii=1;ii<=nlstate+ndeath;ii++)
3438: for (j=1;j<=nlstate+ndeath;j++){
3439: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
3440: savm[ii][j]=(ii==j ? 1.0 : 0.0);
3441: }
3442: for(d=0; d<=dh[mi][i]; d++){
3443: newm=savm;
3444: agexact=agev[mw[mi][i]][i]+d*stepm/YEARM;
3445: cov[2]=agexact;
3446: if(nagesqr==1)
3447: cov[3]= agexact*agexact;
3448: for (kk=1; kk<=cptcovage;kk++) {
3449: cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact;
3450: }
3451: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,
3452: 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate));
3453: savm=oldm;
3454: oldm=newm;
3455: } /* end mult */
3456:
3457: s1=s[mw[mi][i]][i];
3458: s2=s[mw[mi+1][i]][i];
3459: bbh=(double)bh[mi][i]/(double)stepm;
3460: lli= (savm[s1][s2]>(double)1.e-8 ?log((1.+bbh)*out[s1][s2]- bbh*(savm[s1][s2])):log((1.+bbh)*out[s1][s2])); /* linear interpolation */
3461: ipmx +=1;
3462: sw += weight[i];
3463: ll[s[mw[mi][i]][i]] += 2*weight[i]*lli;
3464: } /* end of wave */
3465: } /* end of individual */
3466: } else if(mle==3){ /* exponential inter-extrapolation */
3467: for (i=1,ipmx=0, sw=0.; i<=imx; i++){
3468: for (k=1; k<=cptcovn;k++) cov[2+nagesqr+k]=covar[Tvar[k]][i];
3469: for(mi=1; mi<= wav[i]-1; mi++){
3470: for (ii=1;ii<=nlstate+ndeath;ii++)
3471: for (j=1;j<=nlstate+ndeath;j++){
3472: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
3473: savm[ii][j]=(ii==j ? 1.0 : 0.0);
3474: }
3475: for(d=0; d<dh[mi][i]; d++){
3476: newm=savm;
3477: agexact=agev[mw[mi][i]][i]+d*stepm/YEARM;
3478: cov[2]=agexact;
3479: if(nagesqr==1)
3480: cov[3]= agexact*agexact;
3481: for (kk=1; kk<=cptcovage;kk++) {
3482: cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact;
3483: }
3484: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,
3485: 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate));
3486: savm=oldm;
3487: oldm=newm;
3488: } /* end mult */
3489:
3490: s1=s[mw[mi][i]][i];
3491: s2=s[mw[mi+1][i]][i];
3492: bbh=(double)bh[mi][i]/(double)stepm;
3493: lli= (savm[s1][s2]>1.e-8 ?(1.+bbh)*log(out[s1][s2])- bbh*log(savm[s1][s2]):log((1.+bbh)*out[s1][s2])); /* exponential inter-extrapolation */
3494: ipmx +=1;
3495: sw += weight[i];
3496: ll[s[mw[mi][i]][i]] += 2*weight[i]*lli;
3497: } /* end of wave */
3498: } /* end of individual */
3499: }else if (mle==4){ /* ml=4 no inter-extrapolation */
3500: for (i=1,ipmx=0, sw=0.; i<=imx; i++){
3501: for (k=1; k<=cptcovn;k++) cov[2+nagesqr+k]=covar[Tvar[k]][i];
3502: for(mi=1; mi<= wav[i]-1; mi++){
3503: for (ii=1;ii<=nlstate+ndeath;ii++)
3504: for (j=1;j<=nlstate+ndeath;j++){
3505: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
3506: savm[ii][j]=(ii==j ? 1.0 : 0.0);
3507: }
3508: for(d=0; d<dh[mi][i]; d++){
3509: newm=savm;
3510: agexact=agev[mw[mi][i]][i]+d*stepm/YEARM;
3511: cov[2]=agexact;
3512: if(nagesqr==1)
3513: cov[3]= agexact*agexact;
3514: for (kk=1; kk<=cptcovage;kk++) {
3515: cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact;
3516: }
1.126 brouard 3517:
1.226 brouard 3518: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,
3519: 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate));
3520: savm=oldm;
3521: oldm=newm;
3522: } /* end mult */
3523:
3524: s1=s[mw[mi][i]][i];
3525: s2=s[mw[mi+1][i]][i];
3526: if( s2 > nlstate){
3527: lli=log(out[s1][s2] - savm[s1][s2]);
3528: } else if ( s2==-1 ) { /* alive */
3529: for (j=1,survp=0. ; j<=nlstate; j++)
3530: survp += out[s1][j];
3531: lli= log(survp);
3532: }else{
3533: lli=log(out[s[mw[mi][i]][i]][s[mw[mi+1][i]][i]]); /* Original formula */
3534: }
3535: ipmx +=1;
3536: sw += weight[i];
3537: ll[s[mw[mi][i]][i]] += 2*weight[i]*lli;
1.126 brouard 3538: /* printf("i=%6d s1=%1d s2=%1d mi=%1d mw=%1d dh=%3d prob=%10.6f w=%6.4f out=%10.6f sav=%10.6f\n",i,s1,s2,mi,mw[mi][i],dh[mi][i],exp(lli),weight[i],out[s1][s2],savm[s1][s2]); */
1.226 brouard 3539: } /* end of wave */
3540: } /* end of individual */
3541: }else{ /* ml=5 no inter-extrapolation no jackson =0.8a */
3542: for (i=1,ipmx=0, sw=0.; i<=imx; i++){
3543: for (k=1; k<=cptcovn;k++) cov[2+nagesqr+k]=covar[Tvar[k]][i];
3544: for(mi=1; mi<= wav[i]-1; mi++){
3545: for (ii=1;ii<=nlstate+ndeath;ii++)
3546: for (j=1;j<=nlstate+ndeath;j++){
3547: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
3548: savm[ii][j]=(ii==j ? 1.0 : 0.0);
3549: }
3550: for(d=0; d<dh[mi][i]; d++){
3551: newm=savm;
3552: agexact=agev[mw[mi][i]][i]+d*stepm/YEARM;
3553: cov[2]=agexact;
3554: if(nagesqr==1)
3555: cov[3]= agexact*agexact;
3556: for (kk=1; kk<=cptcovage;kk++) {
3557: cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact;
3558: }
1.126 brouard 3559:
1.226 brouard 3560: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,
3561: 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate));
3562: savm=oldm;
3563: oldm=newm;
3564: } /* end mult */
3565:
3566: s1=s[mw[mi][i]][i];
3567: s2=s[mw[mi+1][i]][i];
3568: lli=log(out[s[mw[mi][i]][i]][s[mw[mi+1][i]][i]]); /* Original formula */
3569: ipmx +=1;
3570: sw += weight[i];
3571: ll[s[mw[mi][i]][i]] += 2*weight[i]*lli;
3572: /*printf("i=%6d s1=%1d s2=%1d mi=%1d mw=%1d dh=%3d prob=%10.6f w=%6.4f out=%10.6f sav=%10.6f\n",i,s1,s2,mi,mw[mi][i],dh[mi][i],exp(lli),weight[i],out[s1][s2],savm[s1][s2]);*/
3573: } /* end of wave */
3574: } /* end of individual */
3575: } /* End of if */
3576: for(k=1,l=0.; k<=nlstate; k++) l += ll[k];
3577: /* printf("l1=%f l2=%f ",ll[1],ll[2]); */
3578: l= l*ipmx/sw; /* To get the same order of magnitude as if weight=1 for every body */
3579: return -l;
1.126 brouard 3580: }
3581:
3582: /*************** log-likelihood *************/
3583: double funcone( double *x)
3584: {
1.228 brouard 3585: /* Same as func but slower because of a lot of printf and if */
1.126 brouard 3586: int i, ii, j, k, mi, d, kk;
1.228 brouard 3587: int ioffset=0;
1.131 brouard 3588: double l, ll[NLSTATEMAX+1], cov[NCOVMAX+1];
1.126 brouard 3589: double **out;
3590: double lli; /* Individual log likelihood */
3591: double llt;
3592: int s1, s2;
1.228 brouard 3593: int iv=0, iqv=0, itv=0, iqtv=0 ; /* Index of varying covariate, fixed quantitative cov, time varying covariate, quantitative time varying covariate */
3594:
1.126 brouard 3595: double bbh, survp;
1.187 brouard 3596: double agexact;
1.214 brouard 3597: double agebegin, ageend;
1.126 brouard 3598: /*extern weight */
3599: /* We are differentiating ll according to initial status */
3600: /* for (i=1;i<=npar;i++) printf("%f ", x[i]);*/
3601: /*for(i=1;i<imx;i++)
3602: printf(" %d\n",s[4][i]);
3603: */
3604: cov[1]=1.;
3605:
3606: for(k=1; k<=nlstate; k++) ll[k]=0.;
1.224 brouard 3607: ioffset=0;
3608: for (i=1,ipmx=0, sw=0.; i<=imx; i++){
1.243 brouard 3609: /* ioffset=2+nagesqr+cptcovage; */
3610: ioffset=2+nagesqr;
1.232 brouard 3611: /* Fixed */
1.224 brouard 3612: /* for (k=1; k<=cptcovn;k++) cov[2+nagesqr+k]=covar[Tvar[k]][i]; */
1.232 brouard 3613: /* for (k=1; k<=ncoveff;k++){ /\* Simple and product fixed Dummy covariates without age* products *\/ */
3614: for (k=1; k<=ncovf;k++){ /* Simple and product fixed covariates without age* products */
3615: cov[ioffset+TvarFind[k]]=covar[Tvar[TvarFind[k]]][i];/* V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1, only V1 is fixed (k=6)*/
3616: /* cov[ioffset+TvarFind[1]]=covar[Tvar[TvarFind[1]]][i]; */
3617: /* cov[2+6]=covar[Tvar[6]][i]; */
3618: /* cov[2+6]=covar[2][i]; V2 */
3619: /* cov[TvarFind[2]]=covar[Tvar[TvarFind[2]]][i]; */
3620: /* cov[2+7]=covar[Tvar[7]][i]; */
3621: /* cov[2+7]=covar[7][i]; V7=V1*V2 */
3622: /* cov[TvarFind[3]]=covar[Tvar[TvarFind[3]]][i]; */
3623: /* cov[2+9]=covar[Tvar[9]][i]; */
3624: /* cov[2+9]=covar[1][i]; V1 */
1.225 brouard 3625: }
1.232 brouard 3626: /* for (k=1; k<=nqfveff;k++){ /\* Simple and product fixed Quantitative covariates without age* products *\/ */
3627: /* cov[++ioffset]=coqvar[TvarFQ[k]][i];/\* V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1, only V2 and V1*V2 is fixed (k=6 and 7?)*\/ */
3628: /* } */
1.231 brouard 3629: /* for(iqv=1; iqv <= nqfveff; iqv++){ /\* Quantitative fixed covariates *\/ */
3630: /* cov[++ioffset]=coqvar[Tvar[iqv]][i]; /\* Only V2 k=6 and V1*V2 7 *\/ */
3631: /* } */
1.225 brouard 3632:
1.233 brouard 3633:
3634: for(mi=1; mi<= wav[i]-1; mi++){ /* Varying with waves */
1.232 brouard 3635: /* Wave varying (but not age varying) */
3636: for(k=1; k <= ncovv ; k++){ /* Varying covariates (single and product but no age )*/
1.242 brouard 3637: /* cov[ioffset+TvarVind[k]]=cotvar[mw[mi][i]][Tvar[TvarVind[k]]][i]; */
3638: cov[ioffset+TvarVind[k]]=cotvar[mw[mi][i]][Tvar[TvarVind[k]]-ncovcol-nqv][i];
3639: }
1.232 brouard 3640: /* for(itv=1; itv <= ntveff; itv++){ /\* Varying dummy covariates (single??)*\/ */
1.242 brouard 3641: /* iv= Tvar[Tmodelind[ioffset-2-nagesqr-cptcovage+itv]]-ncovcol-nqv; /\* Counting the # varying covariate from 1 to ntveff *\/ */
3642: /* cov[ioffset+iv]=cotvar[mw[mi][i]][iv][i]; */
3643: /* k=ioffset-2-nagesqr-cptcovage+itv; /\* position in simple model *\/ */
3644: /* cov[ioffset+itv]=cotvar[mw[mi][i]][TmodelInvind[itv]][i]; */
3645: /* printf(" i=%d,mi=%d,itv=%d,TmodelInvind[itv]=%d,cotvar[mw[mi][i]][TmodelInvind[itv]][i]=%f\n", i, mi, itv, TmodelInvind[itv],cotvar[mw[mi][i]][TmodelInvind[itv]][i]); */
1.232 brouard 3646: /* for(iqtv=1; iqtv <= nqtveff; iqtv++){ /\* Varying quantitatives covariates *\/ */
1.242 brouard 3647: /* iv=TmodelInvQind[iqtv]; /\* Counting the # varying covariate from 1 to ntveff *\/ */
3648: /* /\* printf(" i=%d,mi=%d,iqtv=%d,TmodelInvQind[iqtv]=%d,cotqvar[mw[mi][i]][TmodelInvQind[iqtv]][i]=%f\n", i, mi, iqtv, TmodelInvQind[iqtv],cotqvar[mw[mi][i]][TmodelInvQind[iqtv]][i]); *\/ */
3649: /* cov[ioffset+ntveff+iqtv]=cotqvar[mw[mi][i]][TmodelInvQind[iqtv]][i]; */
1.232 brouard 3650: /* } */
1.126 brouard 3651: for (ii=1;ii<=nlstate+ndeath;ii++)
1.242 brouard 3652: for (j=1;j<=nlstate+ndeath;j++){
3653: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
3654: savm[ii][j]=(ii==j ? 1.0 : 0.0);
3655: }
1.214 brouard 3656:
3657: agebegin=agev[mw[mi][i]][i]; /* Age at beginning of effective wave */
3658: ageend=agev[mw[mi][i]][i] + (dh[mi][i])*stepm/YEARM; /* Age at end of effective wave and at the end of transition */
3659: for(d=0; d<dh[mi][i]; d++){ /* Delay between two effective waves */
1.247 brouard 3660: /* for(d=0; d<=0; d++){ /\* Delay between two effective waves Only one matrix to speed up*\/ */
1.242 brouard 3661: /*dh[m][i] or dh[mw[mi][i]][i] is the delay between two effective waves m=mw[mi][i]
3662: and mw[mi+1][i]. dh depends on stepm.*/
3663: newm=savm;
1.247 brouard 3664: agexact=agev[mw[mi][i]][i]+d*stepm/YEARM; /* Here d is needed */
1.242 brouard 3665: cov[2]=agexact;
3666: if(nagesqr==1)
3667: cov[3]= agexact*agexact;
3668: for (kk=1; kk<=cptcovage;kk++) {
3669: if(!FixedV[Tvar[Tage[kk]]])
3670: cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact;
3671: else
3672: cov[Tage[kk]+2+nagesqr]=cotvar[mw[mi][i]][Tvar[Tage[kk]]-ncovcol-nqv][i]*agexact;
3673: }
3674: /* printf("i=%d,mi=%d,d=%d,mw[mi][i]=%d\n",i, mi,d,mw[mi][i]); */
3675: /* savm=pmij(pmmij,cov,ncovmodel,x,nlstate); */
3676: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,
3677: 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate));
3678: /* out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath, */
3679: /* 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate)); */
3680: savm=oldm;
3681: oldm=newm;
1.126 brouard 3682: } /* end mult */
3683:
3684: s1=s[mw[mi][i]][i];
3685: s2=s[mw[mi+1][i]][i];
1.217 brouard 3686: /* if(s2==-1){ */
3687: /* printf(" s1=%d, s2=%d i=%d \n", s1, s2, i); */
3688: /* /\* exit(1); *\/ */
3689: /* } */
1.126 brouard 3690: bbh=(double)bh[mi][i]/(double)stepm;
3691: /* bias is positive if real duration
3692: * is higher than the multiple of stepm and negative otherwise.
3693: */
3694: if( s2 > nlstate && (mle <5) ){ /* Jackson */
1.242 brouard 3695: lli=log(out[s1][s2] - savm[s1][s2]);
1.216 brouard 3696: } else if ( s2==-1 ) { /* alive */
1.242 brouard 3697: for (j=1,survp=0. ; j<=nlstate; j++)
3698: survp += (1.+bbh)*out[s1][j]- bbh*savm[s1][j];
3699: lli= log(survp);
1.126 brouard 3700: }else if (mle==1){
1.242 brouard 3701: lli= log((1.+bbh)*out[s1][s2]- bbh*savm[s1][s2]); /* linear interpolation */
1.126 brouard 3702: } else if(mle==2){
1.242 brouard 3703: lli= (savm[s1][s2]>(double)1.e-8 ?log((1.+bbh)*out[s1][s2]- bbh*savm[s1][s2]):log((1.+bbh)*out[s1][s2])); /* linear interpolation */
1.126 brouard 3704: } else if(mle==3){ /* exponential inter-extrapolation */
1.242 brouard 3705: lli= (savm[s1][s2]>(double)1.e-8 ?(1.+bbh)*log(out[s1][s2])- bbh*log(savm[s1][s2]):log((1.+bbh)*out[s1][s2])); /* exponential inter-extrapolation */
1.126 brouard 3706: } else if (mle==4){ /* mle=4 no inter-extrapolation */
1.242 brouard 3707: lli=log(out[s1][s2]); /* Original formula */
1.136 brouard 3708: } else{ /* mle=0 back to 1 */
1.242 brouard 3709: lli= log((1.+bbh)*out[s1][s2]- bbh*savm[s1][s2]); /* linear interpolation */
3710: /*lli=log(out[s1][s2]); */ /* Original formula */
1.126 brouard 3711: } /* End of if */
3712: ipmx +=1;
3713: sw += weight[i];
3714: ll[s[mw[mi][i]][i]] += 2*weight[i]*lli;
1.132 brouard 3715: /*printf("i=%6d s1=%1d s2=%1d mi=%1d mw=%1d dh=%3d prob=%10.6f w=%6.4f out=%10.6f sav=%10.6f\n",i,s1,s2,mi,mw[mi][i],dh[mi][i],exp(lli),weight[i],out[s1][s2],savm[s1][s2]); */
1.126 brouard 3716: if(globpr){
1.246 brouard 3717: fprintf(ficresilk,"%09ld %6.1f %6.1f %6d %2d %2d %2d %2d %3d %15.6f %8.4f %8.3f\
1.126 brouard 3718: %11.6f %11.6f %11.6f ", \
1.242 brouard 3719: num[i], agebegin, ageend, i,s1,s2,mi,mw[mi][i],dh[mi][i],exp(lli),weight[i],weight[i]*gipmx/gsw,
3720: 2*weight[i]*lli,out[s1][s2],savm[s1][s2]);
3721: for(k=1,llt=0.,l=0.; k<=nlstate; k++){
3722: llt +=ll[k]*gipmx/gsw;
3723: fprintf(ficresilk," %10.6f",-ll[k]*gipmx/gsw);
3724: }
3725: fprintf(ficresilk," %10.6f\n", -llt);
1.126 brouard 3726: }
1.232 brouard 3727: } /* end of wave */
3728: } /* end of individual */
3729: for(k=1,l=0.; k<=nlstate; k++) l += ll[k];
3730: /* printf("l1=%f l2=%f ",ll[1],ll[2]); */
3731: l= l*ipmx/sw; /* To get the same order of magnitude as if weight=1 for every body */
3732: if(globpr==0){ /* First time we count the contributions and weights */
3733: gipmx=ipmx;
3734: gsw=sw;
3735: }
3736: return -l;
1.126 brouard 3737: }
3738:
3739:
3740: /*************** function likelione ***********/
3741: void likelione(FILE *ficres,double p[], int npar, int nlstate, int *globpri, long *ipmx, double *sw, double *fretone, double (*funcone)(double []))
3742: {
3743: /* This routine should help understanding what is done with
3744: the selection of individuals/waves and
3745: to check the exact contribution to the likelihood.
3746: Plotting could be done.
3747: */
3748: int k;
3749:
3750: if(*globpri !=0){ /* Just counts and sums, no printings */
1.201 brouard 3751: strcpy(fileresilk,"ILK_");
1.202 brouard 3752: strcat(fileresilk,fileresu);
1.126 brouard 3753: if((ficresilk=fopen(fileresilk,"w"))==NULL) {
3754: printf("Problem with resultfile: %s\n", fileresilk);
3755: fprintf(ficlog,"Problem with resultfile: %s\n", fileresilk);
3756: }
1.214 brouard 3757: fprintf(ficresilk, "#individual(line's_record) count ageb ageend s1 s2 wave# effective_wave# number_of_matrices_product pij weight weight/gpw -2ln(pij)*weight 0pij_x 0pij_(x-stepm) cumulating_loglikeli_by_health_state(reweighted=-2ll*weightXnumber_of_contribs/sum_of_weights) and_total\n");
3758: fprintf(ficresilk, "#num_i ageb agend i s1 s2 mi mw dh likeli weight %%weight 2wlli out sav ");
1.126 brouard 3759: /* i,s1,s2,mi,mw[mi][i],dh[mi][i],exp(lli),weight[i],2*weight[i]*lli,out[s1][s2],savm[s1][s2]); */
3760: for(k=1; k<=nlstate; k++)
3761: fprintf(ficresilk," -2*gipw/gsw*weight*ll[%d]++",k);
3762: fprintf(ficresilk," -2*gipw/gsw*weight*ll(total)\n");
3763: }
3764:
3765: *fretone=(*funcone)(p);
3766: if(*globpri !=0){
3767: fclose(ficresilk);
1.205 brouard 3768: if (mle ==0)
3769: fprintf(fichtm,"\n<br>File of contributions to the likelihood computed with initial parameters and mle = %d.",mle);
3770: else if(mle >=1)
3771: fprintf(fichtm,"\n<br>File of contributions to the likelihood computed with optimized parameters mle = %d.",mle);
3772: fprintf(fichtm," You should at least run with mle >= 1 to get starting values corresponding to the optimized parameters in order to visualize the real contribution of each individual/wave: <a href=\"%s\">%s</a><br>\n",subdirf(fileresilk),subdirf(fileresilk));
1.207 brouard 3773:
1.208 brouard 3774:
3775: for (k=1; k<= nlstate ; k++) {
1.211 brouard 3776: fprintf(fichtm,"<br>- Probability p<sub>%dj</sub> by origin %d and destination j. Dot's sizes are related to corresponding weight: <a href=\"%s-p%dj.png\">%s-p%dj.png</a><br> \
1.208 brouard 3777: <img src=\"%s-p%dj.png\">",k,k,subdirf2(optionfilefiname,"ILK_"),k,subdirf2(optionfilefiname,"ILK_"),k,subdirf2(optionfilefiname,"ILK_"),k);
3778: }
1.207 brouard 3779: fprintf(fichtm,"<br>- The function drawn is -2Log(L) in Log scale: by state of origin <a href=\"%s-ori.png\">%s-ori.png</a><br> \
1.204 brouard 3780: <img src=\"%s-ori.png\">",subdirf2(optionfilefiname,"ILK_"),subdirf2(optionfilefiname,"ILK_"),subdirf2(optionfilefiname,"ILK_"));
1.207 brouard 3781: fprintf(fichtm,"<br>- and by state of destination <a href=\"%s-dest.png\">%s-dest.png</a><br> \
1.204 brouard 3782: <img src=\"%s-dest.png\">",subdirf2(optionfilefiname,"ILK_"),subdirf2(optionfilefiname,"ILK_"),subdirf2(optionfilefiname,"ILK_"));
1.207 brouard 3783: fflush(fichtm);
1.205 brouard 3784: }
1.126 brouard 3785: return;
3786: }
3787:
3788:
3789: /*********** Maximum Likelihood Estimation ***************/
3790:
3791: void mlikeli(FILE *ficres,double p[], int npar, int ncovmodel, int nlstate, double ftol, double (*func)(double []))
3792: {
1.165 brouard 3793: int i,j, iter=0;
1.126 brouard 3794: double **xi;
3795: double fret;
3796: double fretone; /* Only one call to likelihood */
3797: /* char filerespow[FILENAMELENGTH];*/
1.162 brouard 3798:
3799: #ifdef NLOPT
3800: int creturn;
3801: nlopt_opt opt;
3802: /* double lb[9] = { -HUGE_VAL, -HUGE_VAL, -HUGE_VAL, -HUGE_VAL, -HUGE_VAL, -HUGE_VAL, -HUGE_VAL, -HUGE_VAL, -HUGE_VAL }; /\* lower bounds *\/ */
3803: double *lb;
3804: double minf; /* the minimum objective value, upon return */
3805: double * p1; /* Shifted parameters from 0 instead of 1 */
3806: myfunc_data dinst, *d = &dinst;
3807: #endif
3808:
3809:
1.126 brouard 3810: xi=matrix(1,npar,1,npar);
3811: for (i=1;i<=npar;i++)
3812: for (j=1;j<=npar;j++)
3813: xi[i][j]=(i==j ? 1.0 : 0.0);
3814: printf("Powell\n"); fprintf(ficlog,"Powell\n");
1.201 brouard 3815: strcpy(filerespow,"POW_");
1.126 brouard 3816: strcat(filerespow,fileres);
3817: if((ficrespow=fopen(filerespow,"w"))==NULL) {
3818: printf("Problem with resultfile: %s\n", filerespow);
3819: fprintf(ficlog,"Problem with resultfile: %s\n", filerespow);
3820: }
3821: fprintf(ficrespow,"# Powell\n# iter -2*LL");
3822: for (i=1;i<=nlstate;i++)
3823: for(j=1;j<=nlstate+ndeath;j++)
3824: if(j!=i)fprintf(ficrespow," p%1d%1d",i,j);
3825: fprintf(ficrespow,"\n");
1.162 brouard 3826: #ifdef POWELL
1.126 brouard 3827: powell(p,xi,npar,ftol,&iter,&fret,func);
1.162 brouard 3828: #endif
1.126 brouard 3829:
1.162 brouard 3830: #ifdef NLOPT
3831: #ifdef NEWUOA
3832: opt = nlopt_create(NLOPT_LN_NEWUOA,npar);
3833: #else
3834: opt = nlopt_create(NLOPT_LN_BOBYQA,npar);
3835: #endif
3836: lb=vector(0,npar-1);
3837: for (i=0;i<npar;i++) lb[i]= -HUGE_VAL;
3838: nlopt_set_lower_bounds(opt, lb);
3839: nlopt_set_initial_step1(opt, 0.1);
3840:
3841: p1= (p+1); /* p *(p+1)@8 and p *(p1)@8 are equal p1[0]=p[1] */
3842: d->function = func;
3843: printf(" Func %.12lf \n",myfunc(npar,p1,NULL,d));
3844: nlopt_set_min_objective(opt, myfunc, d);
3845: nlopt_set_xtol_rel(opt, ftol);
3846: if ((creturn=nlopt_optimize(opt, p1, &minf)) < 0) {
3847: printf("nlopt failed! %d\n",creturn);
3848: }
3849: else {
3850: printf("found minimum after %d evaluations (NLOPT=%d)\n", countcallfunc ,NLOPT);
3851: printf("found minimum at f(%g,%g) = %0.10g\n", p[0], p[1], minf);
3852: iter=1; /* not equal */
3853: }
3854: nlopt_destroy(opt);
3855: #endif
1.126 brouard 3856: free_matrix(xi,1,npar,1,npar);
3857: fclose(ficrespow);
1.203 brouard 3858: printf("\n#Number of iterations & function calls = %d & %d, -2 Log likelihood = %.12f\n",iter, countcallfunc,func(p));
3859: fprintf(ficlog,"\n#Number of iterations & function calls = %d & %d, -2 Log likelihood = %.12f\n",iter, countcallfunc,func(p));
1.180 brouard 3860: fprintf(ficres,"#Number of iterations & function calls = %d & %d, -2 Log likelihood = %.12f\n",iter, countcallfunc,func(p));
1.126 brouard 3861:
3862: }
3863:
3864: /**** Computes Hessian and covariance matrix ***/
1.203 brouard 3865: void hesscov(double **matcov, double **hess, double p[], int npar, double delti[], double ftolhess, double (*func)(double []))
1.126 brouard 3866: {
3867: double **a,**y,*x,pd;
1.203 brouard 3868: /* double **hess; */
1.164 brouard 3869: int i, j;
1.126 brouard 3870: int *indx;
3871:
3872: double hessii(double p[], double delta, int theta, double delti[],double (*func)(double []),int npar);
1.203 brouard 3873: double hessij(double p[], double **hess, double delti[], int i, int j,double (*func)(double []),int npar);
1.126 brouard 3874: void lubksb(double **a, int npar, int *indx, double b[]) ;
3875: void ludcmp(double **a, int npar, int *indx, double *d) ;
3876: double gompertz(double p[]);
1.203 brouard 3877: /* hess=matrix(1,npar,1,npar); */
1.126 brouard 3878:
3879: printf("\nCalculation of the hessian matrix. Wait...\n");
3880: fprintf(ficlog,"\nCalculation of the hessian matrix. Wait...\n");
3881: for (i=1;i<=npar;i++){
1.203 brouard 3882: printf("%d-",i);fflush(stdout);
3883: fprintf(ficlog,"%d-",i);fflush(ficlog);
1.126 brouard 3884:
3885: hess[i][i]=hessii(p,ftolhess,i,delti,func,npar);
3886:
3887: /* printf(" %f ",p[i]);
3888: printf(" %lf %lf %lf",hess[i][i],ftolhess,delti[i]);*/
3889: }
3890:
3891: for (i=1;i<=npar;i++) {
3892: for (j=1;j<=npar;j++) {
3893: if (j>i) {
1.203 brouard 3894: printf(".%d-%d",i,j);fflush(stdout);
3895: fprintf(ficlog,".%d-%d",i,j);fflush(ficlog);
3896: hess[i][j]=hessij(p,hess, delti,i,j,func,npar);
1.126 brouard 3897:
3898: hess[j][i]=hess[i][j];
3899: /*printf(" %lf ",hess[i][j]);*/
3900: }
3901: }
3902: }
3903: printf("\n");
3904: fprintf(ficlog,"\n");
3905:
3906: printf("\nInverting the hessian to get the covariance matrix. Wait...\n");
3907: fprintf(ficlog,"\nInverting the hessian to get the covariance matrix. Wait...\n");
3908:
3909: a=matrix(1,npar,1,npar);
3910: y=matrix(1,npar,1,npar);
3911: x=vector(1,npar);
3912: indx=ivector(1,npar);
3913: for (i=1;i<=npar;i++)
3914: for (j=1;j<=npar;j++) a[i][j]=hess[i][j];
3915: ludcmp(a,npar,indx,&pd);
3916:
3917: for (j=1;j<=npar;j++) {
3918: for (i=1;i<=npar;i++) x[i]=0;
3919: x[j]=1;
3920: lubksb(a,npar,indx,x);
3921: for (i=1;i<=npar;i++){
3922: matcov[i][j]=x[i];
3923: }
3924: }
3925:
3926: printf("\n#Hessian matrix#\n");
3927: fprintf(ficlog,"\n#Hessian matrix#\n");
3928: for (i=1;i<=npar;i++) {
3929: for (j=1;j<=npar;j++) {
1.203 brouard 3930: printf("%.6e ",hess[i][j]);
3931: fprintf(ficlog,"%.6e ",hess[i][j]);
1.126 brouard 3932: }
3933: printf("\n");
3934: fprintf(ficlog,"\n");
3935: }
3936:
1.203 brouard 3937: /* printf("\n#Covariance matrix#\n"); */
3938: /* fprintf(ficlog,"\n#Covariance matrix#\n"); */
3939: /* for (i=1;i<=npar;i++) { */
3940: /* for (j=1;j<=npar;j++) { */
3941: /* printf("%.6e ",matcov[i][j]); */
3942: /* fprintf(ficlog,"%.6e ",matcov[i][j]); */
3943: /* } */
3944: /* printf("\n"); */
3945: /* fprintf(ficlog,"\n"); */
3946: /* } */
3947:
1.126 brouard 3948: /* Recompute Inverse */
1.203 brouard 3949: /* for (i=1;i<=npar;i++) */
3950: /* for (j=1;j<=npar;j++) a[i][j]=matcov[i][j]; */
3951: /* ludcmp(a,npar,indx,&pd); */
3952:
3953: /* printf("\n#Hessian matrix recomputed#\n"); */
3954:
3955: /* for (j=1;j<=npar;j++) { */
3956: /* for (i=1;i<=npar;i++) x[i]=0; */
3957: /* x[j]=1; */
3958: /* lubksb(a,npar,indx,x); */
3959: /* for (i=1;i<=npar;i++){ */
3960: /* y[i][j]=x[i]; */
3961: /* printf("%.3e ",y[i][j]); */
3962: /* fprintf(ficlog,"%.3e ",y[i][j]); */
3963: /* } */
3964: /* printf("\n"); */
3965: /* fprintf(ficlog,"\n"); */
3966: /* } */
3967:
3968: /* Verifying the inverse matrix */
3969: #ifdef DEBUGHESS
3970: y=matprod2(y,hess,1,npar,1,npar,1,npar,matcov);
1.126 brouard 3971:
1.203 brouard 3972: printf("\n#Verification: multiplying the matrix of covariance by the Hessian matrix, should be unity:#\n");
3973: fprintf(ficlog,"\n#Verification: multiplying the matrix of covariance by the Hessian matrix. Should be unity:#\n");
1.126 brouard 3974:
3975: for (j=1;j<=npar;j++) {
3976: for (i=1;i<=npar;i++){
1.203 brouard 3977: printf("%.2f ",y[i][j]);
3978: fprintf(ficlog,"%.2f ",y[i][j]);
1.126 brouard 3979: }
3980: printf("\n");
3981: fprintf(ficlog,"\n");
3982: }
1.203 brouard 3983: #endif
1.126 brouard 3984:
3985: free_matrix(a,1,npar,1,npar);
3986: free_matrix(y,1,npar,1,npar);
3987: free_vector(x,1,npar);
3988: free_ivector(indx,1,npar);
1.203 brouard 3989: /* free_matrix(hess,1,npar,1,npar); */
1.126 brouard 3990:
3991:
3992: }
3993:
3994: /*************** hessian matrix ****************/
3995: double hessii(double x[], double delta, int theta, double delti[], double (*func)(double []), int npar)
1.203 brouard 3996: { /* Around values of x, computes the function func and returns the scales delti and hessian */
1.126 brouard 3997: int i;
3998: int l=1, lmax=20;
1.203 brouard 3999: double k1,k2, res, fx;
1.132 brouard 4000: double p2[MAXPARM+1]; /* identical to x */
1.126 brouard 4001: double delt=0.0001, delts, nkhi=10.,nkhif=1., khi=1.e-4;
4002: int k=0,kmax=10;
4003: double l1;
4004:
4005: fx=func(x);
4006: for (i=1;i<=npar;i++) p2[i]=x[i];
1.145 brouard 4007: for(l=0 ; l <=lmax; l++){ /* Enlarging the zone around the Maximum */
1.126 brouard 4008: l1=pow(10,l);
4009: delts=delt;
4010: for(k=1 ; k <kmax; k=k+1){
4011: delt = delta*(l1*k);
4012: p2[theta]=x[theta] +delt;
1.145 brouard 4013: k1=func(p2)-fx; /* Might be negative if too close to the theoretical maximum */
1.126 brouard 4014: p2[theta]=x[theta]-delt;
4015: k2=func(p2)-fx;
4016: /*res= (k1-2.0*fx+k2)/delt/delt; */
1.203 brouard 4017: res= (k1+k2)/delt/delt/2.; /* Divided by 2 because L and not 2*L */
1.126 brouard 4018:
1.203 brouard 4019: #ifdef DEBUGHESSII
1.126 brouard 4020: printf("%d %d k1=%.12e k2=%.12e xk1=%.12e xk2=%.12e delt=%.12e res=%.12e l=%d k=%d,fx=%.12e\n",theta,theta,k1,k2,x[theta]+delt,x[theta]-delt,delt,res, l, k,fx);
4021: fprintf(ficlog,"%d %d k1=%.12e k2=%.12e xk1=%.12e xk2=%.12e delt=%.12e res=%.12e l=%d k=%d,fx=%.12e\n",theta,theta,k1,k2,x[theta]+delt,x[theta]-delt,delt,res, l, k,fx);
4022: #endif
4023: /*if(fabs(k1-2.0*fx+k2) <1.e-13){ */
4024: if((k1 <khi/nkhi/2.) || (k2 <khi/nkhi/2.)){
4025: k=kmax;
4026: }
4027: else if((k1 >khi/nkhif) || (k2 >khi/nkhif)){ /* Keeps lastvalue before 3.84/2 KHI2 5% 1d.f. */
1.164 brouard 4028: k=kmax; l=lmax*10;
1.126 brouard 4029: }
4030: else if((k1 >khi/nkhi) || (k2 >khi/nkhi)){
4031: delts=delt;
4032: }
1.203 brouard 4033: } /* End loop k */
1.126 brouard 4034: }
4035: delti[theta]=delts;
4036: return res;
4037:
4038: }
4039:
1.203 brouard 4040: double hessij( double x[], double **hess, double delti[], int thetai,int thetaj,double (*func)(double []),int npar)
1.126 brouard 4041: {
4042: int i;
1.164 brouard 4043: int l=1, lmax=20;
1.126 brouard 4044: double k1,k2,k3,k4,res,fx;
1.132 brouard 4045: double p2[MAXPARM+1];
1.203 brouard 4046: int k, kmax=1;
4047: double v1, v2, cv12, lc1, lc2;
1.208 brouard 4048:
4049: int firstime=0;
1.203 brouard 4050:
1.126 brouard 4051: fx=func(x);
1.203 brouard 4052: for (k=1; k<=kmax; k=k+10) {
1.126 brouard 4053: for (i=1;i<=npar;i++) p2[i]=x[i];
1.203 brouard 4054: p2[thetai]=x[thetai]+delti[thetai]*k;
4055: p2[thetaj]=x[thetaj]+delti[thetaj]*k;
1.126 brouard 4056: k1=func(p2)-fx;
4057:
1.203 brouard 4058: p2[thetai]=x[thetai]+delti[thetai]*k;
4059: p2[thetaj]=x[thetaj]-delti[thetaj]*k;
1.126 brouard 4060: k2=func(p2)-fx;
4061:
1.203 brouard 4062: p2[thetai]=x[thetai]-delti[thetai]*k;
4063: p2[thetaj]=x[thetaj]+delti[thetaj]*k;
1.126 brouard 4064: k3=func(p2)-fx;
4065:
1.203 brouard 4066: p2[thetai]=x[thetai]-delti[thetai]*k;
4067: p2[thetaj]=x[thetaj]-delti[thetaj]*k;
1.126 brouard 4068: k4=func(p2)-fx;
1.203 brouard 4069: res=(k1-k2-k3+k4)/4.0/delti[thetai]/k/delti[thetaj]/k/2.; /* Because of L not 2*L */
4070: if(k1*k2*k3*k4 <0.){
1.208 brouard 4071: firstime=1;
1.203 brouard 4072: kmax=kmax+10;
1.208 brouard 4073: }
4074: if(kmax >=10 || firstime ==1){
1.246 brouard 4075: printf("Warning: directions %d-%d, you are not estimating the Hessian at the exact maximum likelihood; you could increase ftol=%.2e\n",thetai,thetaj, ftol);
4076: fprintf(ficlog,"Warning: directions %d-%d, you are not estimating the Hessian at the exact maximum likelihood; you could increase ftol=%.2e\n",thetai,thetaj, ftol);
1.203 brouard 4077: printf("%d %d k=%d, k1=%.12e k2=%.12e k3=%.12e k4=%.12e delti*k=%.12e deltj*k=%.12e, xi-de*k=%.12e xj-de*k=%.12e res=%.12e k1234=%.12e,k1-2=%.12e,k3-4=%.12e\n",thetai,thetaj,k,k1,k2,k3,k4,delti[thetai]/k,delti[thetaj]/k,x[thetai]-delti[thetai]/k,x[thetaj]-delti[thetaj]/k, res,k1-k2-k3+k4,k1-k2,k3-k4);
4078: fprintf(ficlog,"%d %d k=%d, k1=%.12e k2=%.12e k3=%.12e k4=%.12e delti*k=%.12e deltj*k=%.12e, xi-de*k=%.12e xj-de*k=%.12e res=%.12e k1234=%.12e,k1-2=%.12e,k3-4=%.12e\n",thetai,thetaj,k,k1,k2,k3,k4,delti[thetai]/k,delti[thetaj]/k,x[thetai]-delti[thetai]/k,x[thetaj]-delti[thetaj]/k, res,k1-k2-k3+k4,k1-k2,k3-k4);
4079: }
4080: #ifdef DEBUGHESSIJ
4081: v1=hess[thetai][thetai];
4082: v2=hess[thetaj][thetaj];
4083: cv12=res;
4084: /* Computing eigen value of Hessian matrix */
4085: lc1=((v1+v2)+sqrt((v1+v2)*(v1+v2) - 4*(v1*v2-cv12*cv12)))/2.;
4086: lc2=((v1+v2)-sqrt((v1+v2)*(v1+v2) - 4*(v1*v2-cv12*cv12)))/2.;
4087: if ((lc2 <0) || (lc1 <0) ){
4088: printf("Warning: sub Hessian matrix '%d%d' does not have positive eigen values \n",thetai,thetaj);
4089: fprintf(ficlog, "Warning: sub Hessian matrix '%d%d' does not have positive eigen values \n",thetai,thetaj);
4090: printf("%d %d k=%d, k1=%.12e k2=%.12e k3=%.12e k4=%.12e delti/k=%.12e deltj/k=%.12e, xi-de/k=%.12e xj-de/k=%.12e res=%.12e k1234=%.12e,k1-2=%.12e,k3-4=%.12e\n",thetai,thetaj,k,k1,k2,k3,k4,delti[thetai]/k,delti[thetaj]/k,x[thetai]-delti[thetai]/k,x[thetaj]-delti[thetaj]/k, res,k1-k2-k3+k4,k1-k2,k3-k4);
4091: fprintf(ficlog,"%d %d k=%d, k1=%.12e k2=%.12e k3=%.12e k4=%.12e delti/k=%.12e deltj/k=%.12e, xi-de/k=%.12e xj-de/k=%.12e res=%.12e k1234=%.12e,k1-2=%.12e,k3-4=%.12e\n",thetai,thetaj,k,k1,k2,k3,k4,delti[thetai]/k,delti[thetaj]/k,x[thetai]-delti[thetai]/k,x[thetaj]-delti[thetaj]/k, res,k1-k2-k3+k4,k1-k2,k3-k4);
4092: }
1.126 brouard 4093: #endif
4094: }
4095: return res;
4096: }
4097:
1.203 brouard 4098: /* Not done yet: Was supposed to fix if not exactly at the maximum */
4099: /* double hessij( double x[], double delti[], int thetai,int thetaj,double (*func)(double []),int npar) */
4100: /* { */
4101: /* int i; */
4102: /* int l=1, lmax=20; */
4103: /* double k1,k2,k3,k4,res,fx; */
4104: /* double p2[MAXPARM+1]; */
4105: /* double delt=0.0001, delts, nkhi=10.,nkhif=1., khi=1.e-4; */
4106: /* int k=0,kmax=10; */
4107: /* double l1; */
4108:
4109: /* fx=func(x); */
4110: /* for(l=0 ; l <=lmax; l++){ /\* Enlarging the zone around the Maximum *\/ */
4111: /* l1=pow(10,l); */
4112: /* delts=delt; */
4113: /* for(k=1 ; k <kmax; k=k+1){ */
4114: /* delt = delti*(l1*k); */
4115: /* for (i=1;i<=npar;i++) p2[i]=x[i]; */
4116: /* p2[thetai]=x[thetai]+delti[thetai]/k; */
4117: /* p2[thetaj]=x[thetaj]+delti[thetaj]/k; */
4118: /* k1=func(p2)-fx; */
4119:
4120: /* p2[thetai]=x[thetai]+delti[thetai]/k; */
4121: /* p2[thetaj]=x[thetaj]-delti[thetaj]/k; */
4122: /* k2=func(p2)-fx; */
4123:
4124: /* p2[thetai]=x[thetai]-delti[thetai]/k; */
4125: /* p2[thetaj]=x[thetaj]+delti[thetaj]/k; */
4126: /* k3=func(p2)-fx; */
4127:
4128: /* p2[thetai]=x[thetai]-delti[thetai]/k; */
4129: /* p2[thetaj]=x[thetaj]-delti[thetaj]/k; */
4130: /* k4=func(p2)-fx; */
4131: /* res=(k1-k2-k3+k4)/4.0/delti[thetai]*k/delti[thetaj]*k/2.; /\* Because of L not 2*L *\/ */
4132: /* #ifdef DEBUGHESSIJ */
4133: /* printf("%d %d k=%d, k1=%.12e k2=%.12e k3=%.12e k4=%.12e delti/k=%.12e deltj/k=%.12e, xi-de/k=%.12e xj-de/k=%.12e res=%.12e k1234=%.12e,k1-2=%.12e,k3-4=%.12e\n",thetai,thetaj,k,k1,k2,k3,k4,delti[thetai]/k,delti[thetaj]/k,x[thetai]-delti[thetai]/k,x[thetaj]-delti[thetaj]/k, res,k1-k2-k3+k4,k1-k2,k3-k4); */
4134: /* fprintf(ficlog,"%d %d k=%d, k1=%.12e k2=%.12e k3=%.12e k4=%.12e delti/k=%.12e deltj/k=%.12e, xi-de/k=%.12e xj-de/k=%.12e res=%.12e k1234=%.12e,k1-2=%.12e,k3-4=%.12e\n",thetai,thetaj,k,k1,k2,k3,k4,delti[thetai]/k,delti[thetaj]/k,x[thetai]-delti[thetai]/k,x[thetaj]-delti[thetaj]/k, res,k1-k2-k3+k4,k1-k2,k3-k4); */
4135: /* #endif */
4136: /* if((k1 <khi/nkhi/2.) || (k2 <khi/nkhi/2.)|| (k4 <khi/nkhi/2.)|| (k4 <khi/nkhi/2.)){ */
4137: /* k=kmax; */
4138: /* } */
4139: /* else if((k1 >khi/nkhif) || (k2 >khi/nkhif) || (k4 >khi/nkhif) || (k4 >khi/nkhif)){ /\* Keeps lastvalue before 3.84/2 KHI2 5% 1d.f. *\/ */
4140: /* k=kmax; l=lmax*10; */
4141: /* } */
4142: /* else if((k1 >khi/nkhi) || (k2 >khi/nkhi)){ */
4143: /* delts=delt; */
4144: /* } */
4145: /* } /\* End loop k *\/ */
4146: /* } */
4147: /* delti[theta]=delts; */
4148: /* return res; */
4149: /* } */
4150:
4151:
1.126 brouard 4152: /************** Inverse of matrix **************/
4153: void ludcmp(double **a, int n, int *indx, double *d)
4154: {
4155: int i,imax,j,k;
4156: double big,dum,sum,temp;
4157: double *vv;
4158:
4159: vv=vector(1,n);
4160: *d=1.0;
4161: for (i=1;i<=n;i++) {
4162: big=0.0;
4163: for (j=1;j<=n;j++)
4164: if ((temp=fabs(a[i][j])) > big) big=temp;
1.256 brouard 4165: if (big == 0.0){
4166: printf(" Singular Hessian matrix at row %d:\n",i);
4167: for (j=1;j<=n;j++) {
4168: printf(" a[%d][%d]=%f,",i,j,a[i][j]);
4169: fprintf(ficlog," a[%d][%d]=%f,",i,j,a[i][j]);
4170: }
4171: fflush(ficlog);
4172: fclose(ficlog);
4173: nrerror("Singular matrix in routine ludcmp");
4174: }
1.126 brouard 4175: vv[i]=1.0/big;
4176: }
4177: for (j=1;j<=n;j++) {
4178: for (i=1;i<j;i++) {
4179: sum=a[i][j];
4180: for (k=1;k<i;k++) sum -= a[i][k]*a[k][j];
4181: a[i][j]=sum;
4182: }
4183: big=0.0;
4184: for (i=j;i<=n;i++) {
4185: sum=a[i][j];
4186: for (k=1;k<j;k++)
4187: sum -= a[i][k]*a[k][j];
4188: a[i][j]=sum;
4189: if ( (dum=vv[i]*fabs(sum)) >= big) {
4190: big=dum;
4191: imax=i;
4192: }
4193: }
4194: if (j != imax) {
4195: for (k=1;k<=n;k++) {
4196: dum=a[imax][k];
4197: a[imax][k]=a[j][k];
4198: a[j][k]=dum;
4199: }
4200: *d = -(*d);
4201: vv[imax]=vv[j];
4202: }
4203: indx[j]=imax;
4204: if (a[j][j] == 0.0) a[j][j]=TINY;
4205: if (j != n) {
4206: dum=1.0/(a[j][j]);
4207: for (i=j+1;i<=n;i++) a[i][j] *= dum;
4208: }
4209: }
4210: free_vector(vv,1,n); /* Doesn't work */
4211: ;
4212: }
4213:
4214: void lubksb(double **a, int n, int *indx, double b[])
4215: {
4216: int i,ii=0,ip,j;
4217: double sum;
4218:
4219: for (i=1;i<=n;i++) {
4220: ip=indx[i];
4221: sum=b[ip];
4222: b[ip]=b[i];
4223: if (ii)
4224: for (j=ii;j<=i-1;j++) sum -= a[i][j]*b[j];
4225: else if (sum) ii=i;
4226: b[i]=sum;
4227: }
4228: for (i=n;i>=1;i--) {
4229: sum=b[i];
4230: for (j=i+1;j<=n;j++) sum -= a[i][j]*b[j];
4231: b[i]=sum/a[i][i];
4232: }
4233: }
4234:
4235: void pstamp(FILE *fichier)
4236: {
1.196 brouard 4237: fprintf(fichier,"# %s.%s\n#IMaCh version %s, %s\n#%s\n# %s", optionfilefiname,optionfilext,version,copyright, fullversion, strstart);
1.126 brouard 4238: }
4239:
1.253 brouard 4240: int linreg(int ifi, int ila, int *no, const double x[], const double y[], double* a, double* b, double* r, double* sa, double * sb) {
4241:
4242: /* y=a+bx regression */
4243: double sumx = 0.0; /* sum of x */
4244: double sumx2 = 0.0; /* sum of x**2 */
4245: double sumxy = 0.0; /* sum of x * y */
4246: double sumy = 0.0; /* sum of y */
4247: double sumy2 = 0.0; /* sum of y**2 */
4248: double sume2; /* sum of square or residuals */
4249: double yhat;
4250:
4251: double denom=0;
4252: int i;
4253: int ne=*no;
4254:
4255: for ( i=ifi, ne=0;i<=ila;i++) {
4256: if(!isfinite(x[i]) || !isfinite(y[i])){
4257: /* printf(" x[%d]=%f, y[%d]=%f\n",i,x[i],i,y[i]); */
4258: continue;
4259: }
4260: ne=ne+1;
4261: sumx += x[i];
4262: sumx2 += x[i]*x[i];
4263: sumxy += x[i] * y[i];
4264: sumy += y[i];
4265: sumy2 += y[i]*y[i];
4266: denom = (ne * sumx2 - sumx*sumx);
4267: /* printf("ne=%d, i=%d,x[%d]=%f, y[%d]=%f sumx=%f, sumx2=%f, sumxy=%f, sumy=%f, sumy2=%f, denom=%f\n",ne,i,i,x[i],i,y[i], sumx, sumx2,sumxy, sumy, sumy2,denom); */
4268: }
4269:
4270: denom = (ne * sumx2 - sumx*sumx);
4271: if (denom == 0) {
4272: // vertical, slope m is infinity
4273: *b = INFINITY;
4274: *a = 0;
4275: if (r) *r = 0;
4276: return 1;
4277: }
4278:
4279: *b = (ne * sumxy - sumx * sumy) / denom;
4280: *a = (sumy * sumx2 - sumx * sumxy) / denom;
4281: if (r!=NULL) {
4282: *r = (sumxy - sumx * sumy / ne) / /* compute correlation coeff */
4283: sqrt((sumx2 - sumx*sumx/ne) *
4284: (sumy2 - sumy*sumy/ne));
4285: }
4286: *no=ne;
4287: for ( i=ifi, ne=0;i<=ila;i++) {
4288: if(!isfinite(x[i]) || !isfinite(y[i])){
4289: /* printf(" x[%d]=%f, y[%d]=%f\n",i,x[i],i,y[i]); */
4290: continue;
4291: }
4292: ne=ne+1;
4293: yhat = y[i] - *a -*b* x[i];
4294: sume2 += yhat * yhat ;
4295:
4296: denom = (ne * sumx2 - sumx*sumx);
4297: /* printf("ne=%d, i=%d,x[%d]=%f, y[%d]=%f sumx=%f, sumx2=%f, sumxy=%f, sumy=%f, sumy2=%f, denom=%f\n",ne,i,i,x[i],i,y[i], sumx, sumx2,sumxy, sumy, sumy2,denom); */
4298: }
4299: *sb = sqrt(sume2/(ne-2)/(sumx2 - sumx * sumx /ne));
4300: *sa= *sb * sqrt(sumx2/ne);
4301:
4302: return 0;
4303: }
4304:
1.126 brouard 4305: /************ Frequencies ********************/
1.251 brouard 4306: void freqsummary(char fileres[], double p[], double pstart[], int iagemin, int iagemax, int **s, double **agev, int nlstate, int imx, \
1.226 brouard 4307: int *Tvaraff, int *invalidvarcomb, int **nbcode, int *ncodemax,double **mint,double **anint, char strstart[], \
4308: int firstpass, int lastpass, int stepm, int weightopt, char model[])
1.250 brouard 4309: { /* Some frequencies as well as proposing some starting values */
1.226 brouard 4310:
1.265 ! brouard 4311: int i, m, jk, j1, bool, z1,j, nj, nl, k, iv, jj=0, s1=1, s2=1;
1.226 brouard 4312: int iind=0, iage=0;
4313: int mi; /* Effective wave */
4314: int first;
4315: double ***freq; /* Frequencies */
1.253 brouard 4316: double *x, *y, a,b,r, sa, sb; /* for regression, y=b+m*x and r is the correlation coefficient */
4317: int no;
1.226 brouard 4318: double *meanq;
4319: double **meanqt;
4320: double *pp, **prop, *posprop, *pospropt;
4321: double pos=0., posproptt=0., pospropta=0., k2, dateintsum=0,k2cpt=0;
4322: char fileresp[FILENAMELENGTH], fileresphtm[FILENAMELENGTH], fileresphtmfr[FILENAMELENGTH];
4323: double agebegin, ageend;
4324:
4325: pp=vector(1,nlstate);
1.251 brouard 4326: prop=matrix(1,nlstate,iagemin-AGEMARGE,iagemax+4+AGEMARGE);
1.226 brouard 4327: posprop=vector(1,nlstate); /* Counting the number of transition starting from a live state per age */
4328: pospropt=vector(1,nlstate); /* Counting the number of transition starting from a live state */
4329: /* prop=matrix(1,nlstate,iagemin,iagemax+3); */
4330: meanq=vector(1,nqfveff); /* Number of Quantitative Fixed Variables Effective */
4331: meanqt=matrix(1,lastpass,1,nqtveff);
4332: strcpy(fileresp,"P_");
4333: strcat(fileresp,fileresu);
4334: /*strcat(fileresphtm,fileresu);*/
4335: if((ficresp=fopen(fileresp,"w"))==NULL) {
4336: printf("Problem with prevalence resultfile: %s\n", fileresp);
4337: fprintf(ficlog,"Problem with prevalence resultfile: %s\n", fileresp);
4338: exit(0);
4339: }
1.240 brouard 4340:
1.226 brouard 4341: strcpy(fileresphtm,subdirfext(optionfilefiname,"PHTM_",".htm"));
4342: if((ficresphtm=fopen(fileresphtm,"w"))==NULL) {
4343: printf("Problem with prevalence HTM resultfile '%s' with errno='%s'\n",fileresphtm,strerror(errno));
4344: fprintf(ficlog,"Problem with prevalence HTM resultfile '%s' with errno='%s'\n",fileresphtm,strerror(errno));
4345: fflush(ficlog);
4346: exit(70);
4347: }
4348: else{
4349: fprintf(ficresphtm,"<html><head>\n<title>IMaCh PHTM_ %s</title></head>\n <body><font size=\"2\">%s <br> %s</font> \
1.240 brouard 4350: <hr size=\"2\" color=\"#EC5E5E\"> \n \
1.214 brouard 4351: Title=%s <br>Datafile=%s Firstpass=%d Lastpass=%d Stepm=%d Weight=%d Model=1+age+%s<br>\n",\
1.226 brouard 4352: fileresphtm,version,fullversion,title,datafile,firstpass,lastpass,stepm, weightopt, model);
4353: }
1.237 brouard 4354: fprintf(ficresphtm,"Current page is file <a href=\"%s\">%s</a><br>\n\n<h4>Frequencies and prevalence by age at begin of transition and dummy covariate value at beginning of transition</h4>\n",fileresphtm, fileresphtm);
1.240 brouard 4355:
1.226 brouard 4356: strcpy(fileresphtmfr,subdirfext(optionfilefiname,"PHTMFR_",".htm"));
4357: if((ficresphtmfr=fopen(fileresphtmfr,"w"))==NULL) {
4358: printf("Problem with frequency table HTM resultfile '%s' with errno='%s'\n",fileresphtmfr,strerror(errno));
4359: fprintf(ficlog,"Problem with frequency table HTM resultfile '%s' with errno='%s'\n",fileresphtmfr,strerror(errno));
4360: fflush(ficlog);
4361: exit(70);
1.240 brouard 4362: } else{
1.226 brouard 4363: fprintf(ficresphtmfr,"<html><head>\n<title>IMaCh PHTM_Frequency table %s</title></head>\n <body><font size=\"2\">%s <br> %s</font> \
1.240 brouard 4364: <hr size=\"2\" color=\"#EC5E5E\"> \n \
1.214 brouard 4365: Title=%s <br>Datafile=%s Firstpass=%d Lastpass=%d Stepm=%d Weight=%d Model=1+age+%s<br>\n",\
1.226 brouard 4366: fileresphtmfr,version,fullversion,title,datafile,firstpass,lastpass,stepm, weightopt, model);
4367: }
1.240 brouard 4368: fprintf(ficresphtmfr,"Current page is file <a href=\"%s\">%s</a><br>\n\n<h4>Frequencies of all effective transitions of the model, by age at begin of transition, and covariate value at the begin of transition (if the covariate is a varying covariate) </h4>Unknown status is -1<br/>\n",fileresphtmfr, fileresphtmfr);
4369:
1.253 brouard 4370: y= vector(iagemin-AGEMARGE,iagemax+4+AGEMARGE);
4371: x= vector(iagemin-AGEMARGE,iagemax+4+AGEMARGE);
1.251 brouard 4372: freq= ma3x(-5,nlstate+ndeath,-5,nlstate+ndeath,iagemin-AGEMARGE,iagemax+4+AGEMARGE);
1.226 brouard 4373: j1=0;
1.126 brouard 4374:
1.227 brouard 4375: /* j=ncoveff; /\* Only fixed dummy covariates *\/ */
4376: j=cptcoveff; /* Only dummy covariates of the model */
1.226 brouard 4377: if (cptcovn<1) {j=1;ncodemax[1]=1;}
1.240 brouard 4378:
4379:
1.226 brouard 4380: /* Detects if a combination j1 is empty: for a multinomial variable like 3 education levels:
4381: reference=low_education V1=0,V2=0
4382: med_educ V1=1 V2=0,
4383: high_educ V1=0 V2=1
4384: Then V1=1 and V2=1 is a noisy combination that we want to exclude for the list 2**cptcoveff
4385: */
1.249 brouard 4386: dateintsum=0;
4387: k2cpt=0;
4388:
1.253 brouard 4389: if(cptcoveff == 0 )
1.265 ! brouard 4390: nl=1; /* Constant and age model only */
1.253 brouard 4391: else
4392: nl=2;
1.265 ! brouard 4393:
! 4394: /* if a constant only model, one pass to compute frequency tables and to write it on ficresp */
! 4395: /* Loop on nj=1 or 2 if dummy covariates j!=0
! 4396: * Loop on j1(1 to 2**cptcoveff) covariate combination
! 4397: * freq[s1][s2][iage] =0.
! 4398: * Loop on iind
! 4399: * ++freq[s1][s2][iage] weighted
! 4400: * end iind
! 4401: * if covariate and j!0
! 4402: * headers Variable on one line
! 4403: * endif cov j!=0
! 4404: * header of frequency table by age
! 4405: * Loop on age
! 4406: * pp[s1]+=freq[s1][s2][iage] weighted
! 4407: * pos+=freq[s1][s2][iage] weighted
! 4408: * Loop on s1 initial state
! 4409: * fprintf(ficresp
! 4410: * end s1
! 4411: * end age
! 4412: * if j!=0 computes starting values
! 4413: * end compute starting values
! 4414: * end j1
! 4415: * end nl
! 4416: */
1.253 brouard 4417: for (nj = 1; nj <= nl; nj++){ /* nj= 1 constant model, nl number of loops. */
4418: if(nj==1)
4419: j=0; /* First pass for the constant */
1.265 ! brouard 4420: else{
1.253 brouard 4421: j=cptcoveff; /* Other passes for the covariate values */
1.265 ! brouard 4422: }
1.251 brouard 4423: first=1;
1.265 ! brouard 4424: for (j1 = 1; j1 <= (int) pow(2,j); j1++){ /* Loop on all covariates combination of the model, excluding quantitatives, V4=0, V3=0 for example, fixed or varying covariates */
1.251 brouard 4425: posproptt=0.;
4426: /*printf("cptcoveff=%d Tvaraff=%d", cptcoveff,Tvaraff[1]);
4427: scanf("%d", i);*/
4428: for (i=-5; i<=nlstate+ndeath; i++)
1.265 ! brouard 4429: for (s2=-5; s2<=nlstate+ndeath; s2++)
1.251 brouard 4430: for(m=iagemin; m <= iagemax+3; m++)
1.265 ! brouard 4431: freq[i][s2][m]=0;
1.251 brouard 4432:
4433: for (i=1; i<=nlstate; i++) {
1.240 brouard 4434: for(m=iagemin; m <= iagemax+3; m++)
1.251 brouard 4435: prop[i][m]=0;
4436: posprop[i]=0;
4437: pospropt[i]=0;
4438: }
4439: /* for (z1=1; z1<= nqfveff; z1++) { */
4440: /* meanq[z1]+=0.; */
4441: /* for(m=1;m<=lastpass;m++){ */
4442: /* meanqt[m][z1]=0.; */
4443: /* } */
4444: /* } */
4445:
4446: /* dateintsum=0; */
4447: /* k2cpt=0; */
4448:
1.265 ! brouard 4449: /* For that combination of covariates j1 (V4=1 V3=0 for example), we count and print the frequencies in one pass */
1.251 brouard 4450: for (iind=1; iind<=imx; iind++) { /* For each individual iind */
4451: bool=1;
4452: if(j !=0){
4453: if(anyvaryingduminmodel==0){ /* If All fixed covariates */
4454: if (cptcoveff >0) { /* Filter is here: Must be looked at for model=V1+V2+V3+V4 */
4455: /* for (z1=1; z1<= nqfveff; z1++) { */
4456: /* meanq[z1]+=coqvar[Tvar[z1]][iind]; /\* Computes mean of quantitative with selected filter *\/ */
4457: /* } */
4458: for (z1=1; z1<=cptcoveff; z1++) { /* loops on covariates in the model */
4459: /* if(Tvaraff[z1] ==-20){ */
4460: /* /\* sumnew+=cotvar[mw[mi][iind]][z1][iind]; *\/ */
4461: /* }else if(Tvaraff[z1] ==-10){ */
4462: /* /\* sumnew+=coqvar[z1][iind]; *\/ */
4463: /* }else */
4464: if (covar[Tvaraff[z1]][iind]!= nbcode[Tvaraff[z1]][codtabm(j1,z1)]){ /* for combination j1 of covariates */
1.265 ! brouard 4465: /* Tests if the value of the covariate z1 for this individual iind responded to combination j1 (V4=1 V3=0) */
1.251 brouard 4466: bool=0; /* bool should be equal to 1 to be selected, one covariate value failed */
4467: /* printf("bool=%d i=%d, z1=%d, Tvaraff[%d]=%d, covar[Tvarff][%d]=%2f, codtabm(%d,%d)=%d, nbcode[Tvaraff][codtabm(%d,%d)=%d, j1=%d\n",
4468: bool,i,z1, z1, Tvaraff[z1],i,covar[Tvaraff[z1]][i],j1,z1,codtabm(j1,z1),
4469: j1,z1,nbcode[Tvaraff[z1]][codtabm(j1,z1)],j1);*/
4470: /* For j1=7 in V1+V2+V3+V4 = 0 1 1 0 and codtabm(7,3)=1 and nbcde[3][?]=1*/
4471: } /* Onlyf fixed */
4472: } /* end z1 */
4473: } /* cptcovn > 0 */
4474: } /* end any */
4475: }/* end j==0 */
1.265 ! brouard 4476: if (bool==1){ /* We selected an individual iind satisfying combination j1 (V4=1 V3=0) or all fixed covariates */
1.251 brouard 4477: /* for(m=firstpass; m<=lastpass; m++){ */
4478: for(mi=1; mi<wav[iind];mi++){ /* For that wave */
4479: m=mw[mi][iind];
4480: if(j!=0){
4481: if(anyvaryingduminmodel==1){ /* Some are varying covariates */
4482: for (z1=1; z1<=cptcoveff; z1++) {
4483: if( Fixed[Tmodelind[z1]]==1){
4484: iv= Tvar[Tmodelind[z1]]-ncovcol-nqv;
4485: if (cotvar[m][iv][iind]!= nbcode[Tvaraff[z1]][codtabm(j1,z1)]) /* iv=1 to ntv, right modality. If covariate's
4486: value is -1, we don't select. It differs from the
4487: constant and age model which counts them. */
4488: bool=0; /* not selected */
4489: }else if( Fixed[Tmodelind[z1]]== 0) { /* fixed */
4490: if (covar[Tvaraff[z1]][iind]!= nbcode[Tvaraff[z1]][codtabm(j1,z1)]) {
4491: bool=0;
4492: }
4493: }
4494: }
4495: }/* Some are varying covariates, we tried to speed up if all fixed covariates in the model, avoiding waves loop */
4496: } /* end j==0 */
4497: /* bool =0 we keep that guy which corresponds to the combination of dummy values */
4498: if(bool==1){
4499: /* dh[m][iind] or dh[mw[mi][iind]][iind] is the delay between two effective (mi) waves m=mw[mi][iind]
4500: and mw[mi+1][iind]. dh depends on stepm. */
4501: agebegin=agev[m][iind]; /* Age at beginning of wave before transition*/
4502: ageend=agev[m][iind]+(dh[m][iind])*stepm/YEARM; /* Age at end of wave and transition */
4503: if(m >=firstpass && m <=lastpass){
4504: k2=anint[m][iind]+(mint[m][iind]/12.);
4505: /*if ((k2>=dateprev1) && (k2<=dateprev2)) {*/
4506: if(agev[m][iind]==0) agev[m][iind]=iagemax+1; /* All ages equal to 0 are in iagemax+1 */
4507: if(agev[m][iind]==1) agev[m][iind]=iagemax+2; /* All ages equal to 1 are in iagemax+2 */
4508: if (s[m][iind]>0 && s[m][iind]<=nlstate) /* If status at wave m is known and a live state */
4509: prop[s[m][iind]][(int)agev[m][iind]] += weight[iind]; /* At age of beginning of transition, where status is known */
4510: if (m<lastpass) {
4511: /* if(s[m][iind]==4 && s[m+1][iind]==4) */
4512: /* printf(" num=%ld m=%d, iind=%d s1=%d s2=%d agev at m=%d\n", num[iind], m, iind,s[m][iind],s[m+1][iind], (int)agev[m][iind]); */
4513: if(s[m][iind]==-1)
4514: printf(" num=%ld m=%d, iind=%d s1=%d s2=%d agev at m=%d agebegin=%.2f ageend=%.2f, agemed=%d\n", num[iind], m, iind,s[m][iind],s[m+1][iind], (int)agev[m][iind],agebegin, ageend, (int)((agebegin+ageend)/2.));
4515: freq[s[m][iind]][s[m+1][iind]][(int)agev[m][iind]] += weight[iind]; /* At age of beginning of transition, where status is known */
4516: /* if((int)agev[m][iind] == 55) */
4517: /* printf("j=%d, j1=%d Age %d, iind=%d, num=%09ld m=%d\n",j,j1,(int)agev[m][iind],iind, num[iind],m); */
4518: /* freq[s[m][iind]][s[m+1][iind]][(int)((agebegin+ageend)/2.)] += weight[iind]; */
4519: freq[s[m][iind]][s[m+1][iind]][iagemax+3] += weight[iind]; /* Total is in iagemax+3 *//* At age of beginning of transition, where status is known */
1.234 brouard 4520: }
1.251 brouard 4521: } /* end if between passes */
4522: if ((agev[m][iind]>1) && (agev[m][iind]< (iagemax+3)) && (anint[m][iind]!=9999) && (mint[m][iind]!=99) && (j==0)) {
4523: dateintsum=dateintsum+k2; /* on all covariates ?*/
4524: k2cpt++;
4525: /* printf("iind=%ld dateintmean = %lf dateintsum=%lf k2cpt=%lf k2=%lf\n",iind, dateintsum/k2cpt, dateintsum,k2cpt, k2); */
1.234 brouard 4526: }
1.251 brouard 4527: }else{
4528: bool=1;
4529: }/* end bool 2 */
4530: } /* end m */
4531: } /* end bool */
4532: } /* end iind = 1 to imx */
4533: /* prop[s][age] is feeded for any initial and valid live state as well as
4534: freq[s1][s2][age] at single age of beginning the transition, for a combination j1 */
4535:
4536:
4537: /* fprintf(ficresp, "#Count between %.lf/%.lf/%.lf and %.lf/%.lf/%.lf\n",jprev1, mprev1,anprev1,jprev2, mprev2,anprev2);*/
1.265 ! brouard 4538: if(cptcoveff==0 && nj==1) /* no covariate and first pass */
! 4539: pstamp(ficresp);
1.251 brouard 4540: if (cptcoveff>0 && j!=0){
1.265 ! brouard 4541: pstamp(ficresp);
1.251 brouard 4542: printf( "\n#********** Variable ");
4543: fprintf(ficresp, "\n#********** Variable ");
4544: fprintf(ficresphtm, "\n<br/><br/><h3>********** Variable ");
4545: fprintf(ficresphtmfr, "\n<br/><br/><h3>********** Variable ");
4546: fprintf(ficlog, "\n#********** Variable ");
4547: for (z1=1; z1<=cptcoveff; z1++){
4548: if(!FixedV[Tvaraff[z1]]){
4549: printf( "V%d(fixed)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4550: fprintf(ficresp, "V%d(fixed)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4551: fprintf(ficresphtm, "V%d(fixed)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4552: fprintf(ficresphtmfr, "V%d(fixed)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4553: fprintf(ficlog, "V%d(fixed)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
1.250 brouard 4554: }else{
1.251 brouard 4555: printf( "V%d(varying)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4556: fprintf(ficresp, "V%d(varying)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4557: fprintf(ficresphtm, "V%d(varying)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4558: fprintf(ficresphtmfr, "V%d(varying)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4559: fprintf(ficlog, "V%d(varying)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4560: }
4561: }
4562: printf( "**********\n#");
4563: fprintf(ficresp, "**********\n#");
4564: fprintf(ficresphtm, "**********</h3>\n");
4565: fprintf(ficresphtmfr, "**********</h3>\n");
4566: fprintf(ficlog, "**********\n");
4567: }
4568: fprintf(ficresphtm,"<table style=\"text-align:center; border: 1px solid\">");
1.265 ! brouard 4569: if((cptcoveff==0 && nj==1)|| nj==2 ) /* no covariate and first pass */
! 4570: fprintf(ficresp, " Age");
! 4571: if(nj==2) for (z1=1; z1<=cptcoveff; z1++) fprintf(ficresp, " V%d=%d",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
1.251 brouard 4572: for(i=1; i<=nlstate;i++) {
1.265 ! brouard 4573: if((cptcoveff==0 && nj==1)|| nj==2 ) fprintf(ficresp," Prev(%d) N(%d) N ",i,i);
1.251 brouard 4574: fprintf(ficresphtm, "<th>Age</th><th>Prev(%d)</th><th>N(%d)</th><th>N</th>",i,i);
4575: }
1.265 ! brouard 4576: if((cptcoveff==0 && nj==1)|| nj==2 ) fprintf(ficresp, "\n");
1.251 brouard 4577: fprintf(ficresphtm, "\n");
4578:
4579: /* Header of frequency table by age */
4580: fprintf(ficresphtmfr,"<table style=\"text-align:center; border: 1px solid\">");
4581: fprintf(ficresphtmfr,"<th>Age</th> ");
1.265 ! brouard 4582: for(s2=-1; s2 <=nlstate+ndeath; s2++){
1.251 brouard 4583: for(m=-1; m <=nlstate+ndeath; m++){
1.265 ! brouard 4584: if(s2!=0 && m!=0)
! 4585: fprintf(ficresphtmfr,"<th>%d%d</th> ",s2,m);
1.240 brouard 4586: }
1.226 brouard 4587: }
1.251 brouard 4588: fprintf(ficresphtmfr, "\n");
4589:
4590: /* For each age */
4591: for(iage=iagemin; iage <= iagemax+3; iage++){
4592: fprintf(ficresphtm,"<tr>");
4593: if(iage==iagemax+1){
4594: fprintf(ficlog,"1");
4595: fprintf(ficresphtmfr,"<tr><th>0</th> ");
4596: }else if(iage==iagemax+2){
4597: fprintf(ficlog,"0");
4598: fprintf(ficresphtmfr,"<tr><th>Unknown</th> ");
4599: }else if(iage==iagemax+3){
4600: fprintf(ficlog,"Total");
4601: fprintf(ficresphtmfr,"<tr><th>Total</th> ");
4602: }else{
1.240 brouard 4603: if(first==1){
1.251 brouard 4604: first=0;
4605: printf("See log file for details...\n");
4606: }
4607: fprintf(ficresphtmfr,"<tr><th>%d</th> ",iage);
4608: fprintf(ficlog,"Age %d", iage);
4609: }
1.265 ! brouard 4610: for(s1=1; s1 <=nlstate ; s1++){
! 4611: for(m=-1, pp[s1]=0; m <=nlstate+ndeath ; m++)
! 4612: pp[s1] += freq[s1][m][iage];
1.251 brouard 4613: }
1.265 ! brouard 4614: for(s1=1; s1 <=nlstate ; s1++){
1.251 brouard 4615: for(m=-1, pos=0; m <=0 ; m++)
1.265 ! brouard 4616: pos += freq[s1][m][iage];
! 4617: if(pp[s1]>=1.e-10){
1.251 brouard 4618: if(first==1){
1.265 ! brouard 4619: printf(" %d.=%.0f loss[%d]=%.1f%%",s1,pp[s1],s1,100*pos/pp[s1]);
1.251 brouard 4620: }
1.265 ! brouard 4621: fprintf(ficlog," %d.=%.0f loss[%d]=%.1f%%",s1,pp[s1],s1,100*pos/pp[s1]);
1.251 brouard 4622: }else{
4623: if(first==1)
1.265 ! brouard 4624: printf(" %d.=%.0f loss[%d]=NaNQ%%",s1,pp[s1],s1);
! 4625: fprintf(ficlog," %d.=%.0f loss[%d]=NaNQ%%",s1,pp[s1],s1);
1.240 brouard 4626: }
4627: }
4628:
1.265 ! brouard 4629: for(s1=1; s1 <=nlstate ; s1++){
! 4630: /* posprop[s1]=0; */
! 4631: for(m=0, pp[s1]=0; m <=nlstate+ndeath; m++)/* Summing on all ages */
! 4632: pp[s1] += freq[s1][m][iage];
! 4633: } /* pp[s1] is the total number of transitions starting from state s1 and any ending status until this age */
! 4634:
! 4635: for(s1=1,pos=0, pospropta=0.; s1 <=nlstate ; s1++){
! 4636: pos += pp[s1]; /* pos is the total number of transitions until this age */
! 4637: posprop[s1] += prop[s1][iage]; /* prop is the number of transitions from a live state
! 4638: from s1 at age iage prop[s[m][iind]][(int)agev[m][iind]] += weight[iind] */
! 4639: pospropta += prop[s1][iage]; /* prop is the number of transitions from a live state
! 4640: from s1 at age iage prop[s[m][iind]][(int)agev[m][iind]] += weight[iind] */
! 4641: }
! 4642:
! 4643: /* Writing ficresp */
! 4644: if(cptcoveff==0 && nj==1){ /* no covariate and first pass */
! 4645: if( iage <= iagemax){
! 4646: fprintf(ficresp," %d",iage);
! 4647: }
! 4648: }else if( nj==2){
! 4649: if( iage <= iagemax){
! 4650: fprintf(ficresp," %d",iage);
! 4651: for (z1=1; z1<=cptcoveff; z1++) fprintf(ficresp, " %d %d",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
! 4652: }
1.240 brouard 4653: }
1.265 ! brouard 4654: for(s1=1; s1 <=nlstate ; s1++){
1.240 brouard 4655: if(pos>=1.e-5){
1.251 brouard 4656: if(first==1)
1.265 ! brouard 4657: printf(" %d.=%.0f prev[%d]=%.1f%%",s1,pp[s1],s1,100*pp[s1]/pos);
! 4658: fprintf(ficlog," %d.=%.0f prev[%d]=%.1f%%",s1,pp[s1],s1,100*pp[s1]/pos);
1.251 brouard 4659: }else{
4660: if(first==1)
1.265 ! brouard 4661: printf(" %d.=%.0f prev[%d]=NaNQ%%",s1,pp[s1],s1);
! 4662: fprintf(ficlog," %d.=%.0f prev[%d]=NaNQ%%",s1,pp[s1],s1);
1.251 brouard 4663: }
4664: if( iage <= iagemax){
4665: if(pos>=1.e-5){
1.265 ! brouard 4666: if(cptcoveff==0 && nj==1){ /* no covariate and first pass */
! 4667: fprintf(ficresp," %.5f %.0f %.0f",prop[s1][iage]/pospropta, prop[s1][iage],pospropta);
! 4668: }else if( nj==2){
! 4669: fprintf(ficresp," %.5f %.0f %.0f",prop[s1][iage]/pospropta, prop[s1][iage],pospropta);
! 4670: }
! 4671: fprintf(ficresphtm,"<th>%d</th><td>%.5f</td><td>%.0f</td><td>%.0f</td>",iage,prop[s1][iage]/pospropta, prop[s1][iage],pospropta);
! 4672: /*probs[iage][s1][j1]= pp[s1]/pos;*/
! 4673: /*printf("\niage=%d s1=%d j1=%d %.5f %.0f %.0f %f",iage,s1,j1,pp[s1]/pos, pp[s1],pos,probs[iage][s1][j1]);*/
! 4674: } else{
! 4675: if((cptcoveff==0 && nj==1)|| nj==2 ) fprintf(ficresp," NaNq %.0f %.0f",prop[s1][iage],pospropta);
! 4676: fprintf(ficresphtm,"<th>%d</th><td>NaNq</td><td>%.0f</td><td>%.0f</td>",iage, prop[s1][iage],pospropta);
1.251 brouard 4677: }
1.240 brouard 4678: }
1.265 ! brouard 4679: pospropt[s1] +=posprop[s1];
! 4680: } /* end loop s1 */
1.251 brouard 4681: /* pospropt=0.; */
1.265 ! brouard 4682: for(s1=-1; s1 <=nlstate+ndeath; s1++){
1.251 brouard 4683: for(m=-1; m <=nlstate+ndeath; m++){
1.265 ! brouard 4684: if(freq[s1][m][iage] !=0 ) { /* minimizing output */
1.251 brouard 4685: if(first==1){
1.265 ! brouard 4686: printf(" %d%d=%.0f",s1,m,freq[s1][m][iage]);
1.251 brouard 4687: }
1.265 ! brouard 4688: /* printf(" %d%d=%.0f",s1,m,freq[s1][m][iage]); */
! 4689: fprintf(ficlog," %d%d=%.0f",s1,m,freq[s1][m][iage]);
1.251 brouard 4690: }
1.265 ! brouard 4691: if(s1!=0 && m!=0)
! 4692: fprintf(ficresphtmfr,"<td>%.0f</td> ",freq[s1][m][iage]);
1.240 brouard 4693: }
1.265 ! brouard 4694: } /* end loop s1 */
1.251 brouard 4695: posproptt=0.;
1.265 ! brouard 4696: for(s1=1; s1 <=nlstate; s1++){
! 4697: posproptt += pospropt[s1];
1.251 brouard 4698: }
4699: fprintf(ficresphtmfr,"</tr>\n ");
1.265 ! brouard 4700: fprintf(ficresphtm,"</tr>\n");
! 4701: if((cptcoveff==0 && nj==1)|| nj==2 ) {
! 4702: if(iage <= iagemax)
! 4703: fprintf(ficresp,"\n");
1.240 brouard 4704: }
1.251 brouard 4705: if(first==1)
4706: printf("Others in log...\n");
4707: fprintf(ficlog,"\n");
4708: } /* end loop age iage */
1.265 ! brouard 4709:
1.251 brouard 4710: fprintf(ficresphtm,"<tr><th>Tot</th>");
1.265 ! brouard 4711: for(s1=1; s1 <=nlstate ; s1++){
1.251 brouard 4712: if(posproptt < 1.e-5){
1.265 ! brouard 4713: fprintf(ficresphtm,"<td>Nanq</td><td>%.0f</td><td>%.0f</td>",pospropt[s1],posproptt);
1.251 brouard 4714: }else{
1.265 ! brouard 4715: fprintf(ficresphtm,"<td>%.5f</td><td>%.0f</td><td>%.0f</td>",pospropt[s1]/posproptt,pospropt[s1],posproptt);
1.240 brouard 4716: }
1.226 brouard 4717: }
1.251 brouard 4718: fprintf(ficresphtm,"</tr>\n");
4719: fprintf(ficresphtm,"</table>\n");
4720: fprintf(ficresphtmfr,"</table>\n");
1.226 brouard 4721: if(posproptt < 1.e-5){
1.251 brouard 4722: fprintf(ficresphtm,"\n <p><b> This combination (%d) is not valid and no result will be produced</b></p>",j1);
4723: fprintf(ficresphtmfr,"\n <p><b> This combination (%d) is not valid and no result will be produced</b></p>",j1);
1.260 brouard 4724: fprintf(ficlog,"# This combination (%d) is not valid and no result will be produced\n",j1);
4725: printf("# This combination (%d) is not valid and no result will be produced\n",j1);
1.251 brouard 4726: invalidvarcomb[j1]=1;
1.226 brouard 4727: }else{
1.251 brouard 4728: fprintf(ficresphtm,"\n <p> This combination (%d) is valid and result will be produced.</p>",j1);
4729: invalidvarcomb[j1]=0;
1.226 brouard 4730: }
1.251 brouard 4731: fprintf(ficresphtmfr,"</table>\n");
4732: fprintf(ficlog,"\n");
4733: if(j!=0){
4734: printf("#Freqsummary: Starting values for combination j1=%d:\n", j1);
1.265 ! brouard 4735: for(i=1,s1=1; i <=nlstate; i++){
1.251 brouard 4736: for(k=1; k <=(nlstate+ndeath); k++){
4737: if (k != i) {
1.265 ! brouard 4738: for(jj=1; jj <=ncovmodel; jj++){ /* For counting s1 */
1.253 brouard 4739: if(jj==1){ /* Constant case (in fact cste + age) */
1.251 brouard 4740: if(j1==1){ /* All dummy covariates to zero */
4741: freq[i][k][iagemax+4]=freq[i][k][iagemax+3]; /* Stores case 0 0 0 */
4742: freq[i][i][iagemax+4]=freq[i][i][iagemax+3]; /* Stores case 0 0 0 */
1.252 brouard 4743: printf("%d%d ",i,k);
4744: fprintf(ficlog,"%d%d ",i,k);
1.265 ! brouard 4745: printf("%12.7f ln(%.0f/%.0f)= %f, OR=%f sd=%f \n",p[s1],freq[i][k][iagemax+3],freq[i][i][iagemax+3], log(freq[i][k][iagemax+3]/freq[i][i][iagemax+3]),freq[i][k][iagemax+3]/freq[i][i][iagemax+3], sqrt(1/freq[i][k][iagemax+3]+1/freq[i][i][iagemax+3]));
! 4746: fprintf(ficlog,"%12.7f ln(%.0f/%.0f)= %12.7f \n",p[s1],freq[i][k][iagemax+3],freq[i][i][iagemax+3], log(freq[i][k][iagemax+3]/freq[i][i][iagemax+3]));
! 4747: pstart[s1]= log(freq[i][k][iagemax+3]/freq[i][i][iagemax+3]);
1.251 brouard 4748: }
1.253 brouard 4749: }else if((j1==1) && (jj==2 || nagesqr==1)){ /* age or age*age parameter without covariate V4*age (to be done later) */
4750: for(iage=iagemin; iage <= iagemax+3; iage++){
4751: x[iage]= (double)iage;
4752: y[iage]= log(freq[i][k][iage]/freq[i][i][iage]);
1.265 ! brouard 4753: /* printf("i=%d, k=%d, s1=%d, j1=%d, jj=%d, y[%d]=%f\n",i,k,s1,j1,jj, iage, y[iage]); */
1.253 brouard 4754: }
4755: linreg(iagemin,iagemax,&no,x,y,&a,&b,&r, &sa, &sb ); /* y= a+b*x with standard errors */
1.265 ! brouard 4756: pstart[s1]=b;
! 4757: pstart[s1-1]=a;
1.252 brouard 4758: }else if( j1!=1 && (j1==2 || (log(j1-1.)/log(2.)-(int)(log(j1-1.)/log(2.))) <0.010) && ( TvarsDind[(int)(log(j1-1.)/log(2.))+1]+2+nagesqr == jj) && Dummy[jj-2-nagesqr]==0){ /* We want only if the position, jj, in model corresponds to unique covariate equal to 1 in j1 combination */
4759: printf("j1=%d, jj=%d, (int)(log(j1-1.)/log(2.))+1=%d, TvarsDind[(int)(log(j1-1.)/log(2.))+1]=%d\n",j1, jj,(int)(log(j1-1.)/log(2.))+1,TvarsDind[(int)(log(j1-1.)/log(2.))+1]);
4760: printf("j1=%d, jj=%d, (log(j1-1.)/log(2.))+1=%f, TvarsDind[(int)(log(j1-1.)/log(2.))+1]=%d\n",j1, jj,(log(j1-1.)/log(2.))+1,TvarsDind[(int)(log(j1-1.)/log(2.))+1]);
1.265 ! brouard 4761: pstart[s1]= log((freq[i][k][iagemax+3]/freq[i][i][iagemax+3])/(freq[i][k][iagemax+4]/freq[i][i][iagemax+4]));
1.252 brouard 4762: printf("%d%d ",i,k);
4763: fprintf(ficlog,"%d%d ",i,k);
1.265 ! brouard 4764: printf("s1=%d,i=%d,k=%d,p[%d]=%12.7f ln((%.0f/%.0f)/(%.0f/%.0f))= %f, OR=%f sd=%f \n",s1,i,k,s1,p[s1],freq[i][k][iagemax+3],freq[i][i][iagemax+3],freq[i][k][iagemax+4],freq[i][i][iagemax+4], log((freq[i][k][iagemax+3]/freq[i][i][iagemax+3])/(freq[i][k][iagemax+4]/freq[i][i][iagemax+4])),(freq[i][k][iagemax+3]/freq[i][i][iagemax+3])/(freq[i][k][iagemax+4]/freq[i][i][iagemax+4]), sqrt(1/freq[i][k][iagemax+3]+1/freq[i][i][iagemax+3]+1/freq[i][k][iagemax+4]+1/freq[i][i][iagemax+4]));
1.251 brouard 4765: }else{ /* Other cases, like quantitative fixed or varying covariates */
4766: ;
4767: }
4768: /* printf("%12.7f )", param[i][jj][k]); */
4769: /* fprintf(ficlog,"%12.7f )", param[i][jj][k]); */
1.265 ! brouard 4770: s1++;
1.251 brouard 4771: } /* end jj */
4772: } /* end k!= i */
4773: } /* end k */
1.265 ! brouard 4774: } /* end i, s1 */
1.251 brouard 4775: } /* end j !=0 */
4776: } /* end selected combination of covariate j1 */
4777: if(j==0){ /* We can estimate starting values from the occurences in each case */
4778: printf("#Freqsummary: Starting values for the constants:\n");
4779: fprintf(ficlog,"\n");
1.265 ! brouard 4780: for(i=1,s1=1; i <=nlstate; i++){
1.251 brouard 4781: for(k=1; k <=(nlstate+ndeath); k++){
4782: if (k != i) {
4783: printf("%d%d ",i,k);
4784: fprintf(ficlog,"%d%d ",i,k);
4785: for(jj=1; jj <=ncovmodel; jj++){
1.265 ! brouard 4786: pstart[s1]=p[s1]; /* Setting pstart to p values by default */
1.253 brouard 4787: if(jj==1){ /* Age has to be done */
1.265 ! brouard 4788: pstart[s1]= log(freq[i][k][iagemax+3]/freq[i][i][iagemax+3]);
! 4789: printf("%12.7f ln(%.0f/%.0f)= %12.7f ",p[s1],freq[i][k][iagemax+3],freq[i][i][iagemax+3], log(freq[i][k][iagemax+3]/freq[i][i][iagemax+3]));
! 4790: fprintf(ficlog,"%12.7f ln(%.0f/%.0f)= %12.7f ",p[s1],freq[i][k][iagemax+3],freq[i][i][iagemax+3], log(freq[i][k][iagemax+3]/freq[i][i][iagemax+3]));
1.251 brouard 4791: }
4792: /* printf("%12.7f )", param[i][jj][k]); */
4793: /* fprintf(ficlog,"%12.7f )", param[i][jj][k]); */
1.265 ! brouard 4794: s1++;
1.250 brouard 4795: }
1.251 brouard 4796: printf("\n");
4797: fprintf(ficlog,"\n");
1.250 brouard 4798: }
4799: }
4800: }
1.251 brouard 4801: printf("#Freqsummary\n");
4802: fprintf(ficlog,"\n");
1.265 ! brouard 4803: for(s1=-1; s1 <=nlstate+ndeath; s1++){
! 4804: for(s2=-1; s2 <=nlstate+ndeath; s2++){
! 4805: /* param[i]|j][k]= freq[s1][s2][iagemax+3] */
! 4806: printf(" %d%d=%.0f",s1,s2,freq[s1][s2][iagemax+3]);
! 4807: fprintf(ficlog," %d%d=%.0f",s1,s2,freq[s1][s2][iagemax+3]);
! 4808: /* if(freq[s1][s2][iage] !=0 ) { /\* minimizing output *\/ */
! 4809: /* printf(" %d%d=%.0f",s1,s2,freq[s1][s2][iagemax+3]); */
! 4810: /* fprintf(ficlog," %d%d=%.0f",s1,s2,freq[s1][s2][iagemax+3]); */
1.251 brouard 4811: /* } */
4812: }
1.265 ! brouard 4813: } /* end loop s1 */
1.251 brouard 4814:
4815: printf("\n");
4816: fprintf(ficlog,"\n");
4817: } /* end j=0 */
1.249 brouard 4818: } /* end j */
1.252 brouard 4819:
1.253 brouard 4820: if(mle == -2){ /* We want to use these values as starting values */
1.252 brouard 4821: for(i=1, jk=1; i <=nlstate; i++){
4822: for(j=1; j <=nlstate+ndeath; j++){
4823: if(j!=i){
4824: /*ca[0]= k+'a'-1;ca[1]='\0';*/
4825: printf("%1d%1d",i,j);
4826: fprintf(ficparo,"%1d%1d",i,j);
4827: for(k=1; k<=ncovmodel;k++){
4828: /* printf(" %lf",param[i][j][k]); */
4829: /* fprintf(ficparo," %lf",param[i][j][k]); */
4830: p[jk]=pstart[jk];
4831: printf(" %f ",pstart[jk]);
4832: fprintf(ficparo," %f ",pstart[jk]);
4833: jk++;
4834: }
4835: printf("\n");
4836: fprintf(ficparo,"\n");
4837: }
4838: }
4839: }
4840: } /* end mle=-2 */
1.226 brouard 4841: dateintmean=dateintsum/k2cpt;
1.240 brouard 4842:
1.226 brouard 4843: fclose(ficresp);
4844: fclose(ficresphtm);
4845: fclose(ficresphtmfr);
4846: free_vector(meanq,1,nqfveff);
4847: free_matrix(meanqt,1,lastpass,1,nqtveff);
1.253 brouard 4848: free_vector(x, iagemin-AGEMARGE, iagemax+4+AGEMARGE);
4849: free_vector(y, iagemin-AGEMARGE, iagemax+4+AGEMARGE);
1.251 brouard 4850: free_ma3x(freq,-5,nlstate+ndeath,-5,nlstate+ndeath, iagemin-AGEMARGE, iagemax+4+AGEMARGE);
1.226 brouard 4851: free_vector(pospropt,1,nlstate);
4852: free_vector(posprop,1,nlstate);
1.251 brouard 4853: free_matrix(prop,1,nlstate,iagemin-AGEMARGE, iagemax+4+AGEMARGE);
1.226 brouard 4854: free_vector(pp,1,nlstate);
4855: /* End of freqsummary */
4856: }
1.126 brouard 4857:
4858: /************ Prevalence ********************/
1.227 brouard 4859: void prevalence(double ***probs, double agemin, double agemax, int **s, double **agev, int nlstate, int imx, int *Tvar, int **nbcode, int *ncodemax,double **mint,double **anint, double dateprev1,double dateprev2, int firstpass, int lastpass)
4860: {
4861: /* Compute observed prevalence between dateprev1 and dateprev2 by counting the number of people
4862: in each health status at the date of interview (if between dateprev1 and dateprev2).
4863: We still use firstpass and lastpass as another selection.
4864: */
1.126 brouard 4865:
1.227 brouard 4866: int i, m, jk, j1, bool, z1,j, iv;
4867: int mi; /* Effective wave */
4868: int iage;
4869: double agebegin, ageend;
4870:
4871: double **prop;
4872: double posprop;
4873: double y2; /* in fractional years */
4874: int iagemin, iagemax;
4875: int first; /** to stop verbosity which is redirected to log file */
4876:
4877: iagemin= (int) agemin;
4878: iagemax= (int) agemax;
4879: /*pp=vector(1,nlstate);*/
1.251 brouard 4880: prop=matrix(1,nlstate,iagemin-AGEMARGE,iagemax+4+AGEMARGE);
1.227 brouard 4881: /* freq=ma3x(-1,nlstate+ndeath,-1,nlstate+ndeath,iagemin,iagemax+3);*/
4882: j1=0;
1.222 brouard 4883:
1.227 brouard 4884: /*j=cptcoveff;*/
4885: if (cptcovn<1) {j=1;ncodemax[1]=1;}
1.222 brouard 4886:
1.227 brouard 4887: first=1;
4888: for(j1=1; j1<= (int) pow(2,cptcoveff);j1++){ /* For each combination of covariate */
4889: for (i=1; i<=nlstate; i++)
1.251 brouard 4890: for(iage=iagemin-AGEMARGE; iage <= iagemax+4+AGEMARGE; iage++)
1.227 brouard 4891: prop[i][iage]=0.0;
4892: printf("Prevalence combination of varying and fixed dummies %d\n",j1);
4893: /* fprintf(ficlog," V%d=%d ",Tvaraff[j1],nbcode[Tvaraff[j1]][codtabm(k,j1)]); */
4894: fprintf(ficlog,"Prevalence combination of varying and fixed dummies %d\n",j1);
4895:
4896: for (i=1; i<=imx; i++) { /* Each individual */
4897: bool=1;
4898: /* for(m=firstpass; m<=lastpass; m++){/\* Other selection (we can limit to certain interviews*\/ */
4899: for(mi=1; mi<wav[i];mi++){ /* For this wave too look where individual can be counted V4=0 V3=0 */
4900: m=mw[mi][i];
4901: /* Tmodelind[z1]=k is the position of the varying covariate in the model, but which # within 1 to ntv? */
4902: /* Tvar[Tmodelind[z1]] is the n of Vn; n-ncovcol-nqv is the first time varying covariate or iv */
4903: for (z1=1; z1<=cptcoveff; z1++){
4904: if( Fixed[Tmodelind[z1]]==1){
4905: iv= Tvar[Tmodelind[z1]]-ncovcol-nqv;
4906: if (cotvar[m][iv][i]!= nbcode[Tvaraff[z1]][codtabm(j1,z1)]) /* iv=1 to ntv, right modality */
4907: bool=0;
4908: }else if( Fixed[Tmodelind[z1]]== 0) /* fixed */
4909: if (covar[Tvaraff[z1]][i]!= nbcode[Tvaraff[z1]][codtabm(j1,z1)]) {
4910: bool=0;
4911: }
4912: }
4913: if(bool==1){ /* Otherwise we skip that wave/person */
4914: agebegin=agev[m][i]; /* Age at beginning of wave before transition*/
4915: /* ageend=agev[m][i]+(dh[m][i])*stepm/YEARM; /\* Age at end of wave and transition *\/ */
4916: if(m >=firstpass && m <=lastpass){
4917: y2=anint[m][i]+(mint[m][i]/12.); /* Fractional date in year */
4918: if ((y2>=dateprev1) && (y2<=dateprev2)) { /* Here is the main selection (fractional years) */
4919: if(agev[m][i]==0) agev[m][i]=iagemax+1;
4920: if(agev[m][i]==1) agev[m][i]=iagemax+2;
1.251 brouard 4921: if((int)agev[m][i] <iagemin-AGEMARGE || (int)agev[m][i] >iagemax+4+AGEMARGE){
1.227 brouard 4922: printf("Error on individual # %d agev[m][i]=%f <%d-%d or > %d+3+%d m=%d; either change agemin or agemax or fix data\n",i, agev[m][i],iagemin,AGEMARGE, iagemax,AGEMARGE,m);
4923: exit(1);
4924: }
4925: if (s[m][i]>0 && s[m][i]<=nlstate) {
4926: /*if(i>4620) printf(" i=%d m=%d s[m][i]=%d (int)agev[m][i]=%d weight[i]=%f prop=%f\n",i,m,s[m][i],(int)agev[m][m],weight[i],prop[s[m][i]][(int)agev[m][i]]);*/
4927: prop[s[m][i]][(int)agev[m][i]] += weight[i];/* At age of beginning of transition, where status is known */
4928: prop[s[m][i]][iagemax+3] += weight[i];
4929: } /* end valid statuses */
4930: } /* end selection of dates */
4931: } /* end selection of waves */
4932: } /* end bool */
4933: } /* end wave */
4934: } /* end individual */
4935: for(i=iagemin; i <= iagemax+3; i++){
4936: for(jk=1,posprop=0; jk <=nlstate ; jk++) {
4937: posprop += prop[jk][i];
4938: }
4939:
4940: for(jk=1; jk <=nlstate ; jk++){
4941: if( i <= iagemax){
4942: if(posprop>=1.e-5){
4943: probs[i][jk][j1]= prop[jk][i]/posprop;
4944: } else{
4945: if(first==1){
4946: first=0;
4947: printf("Warning Observed prevalence probs[%d][%d][%d]=%lf because of lack of cases\nSee others in log file...\n",jk,i,j1,probs[i][jk][j1]);
4948: }
4949: }
4950: }
4951: }/* end jk */
4952: }/* end i */
1.222 brouard 4953: /*} *//* end i1 */
1.227 brouard 4954: } /* end j1 */
1.222 brouard 4955:
1.227 brouard 4956: /* free_ma3x(freq,-1,nlstate+ndeath,-1,nlstate+ndeath, iagemin, iagemax+3);*/
4957: /*free_vector(pp,1,nlstate);*/
1.251 brouard 4958: free_matrix(prop,1,nlstate, iagemin-AGEMARGE,iagemax+4+AGEMARGE);
1.227 brouard 4959: } /* End of prevalence */
1.126 brouard 4960:
4961: /************* Waves Concatenation ***************/
4962:
4963: void concatwav(int wav[], int **dh, int **bh, int **mw, int **s, double *agedc, double **agev, int firstpass, int lastpass, int imx, int nlstate, int stepm)
4964: {
4965: /* Concatenates waves: wav[i] is the number of effective (useful waves) of individual i.
4966: Death is a valid wave (if date is known).
4967: mw[mi][i] is the mi (mi=1 to wav[i]) effective wave of individual i
4968: dh[m][i] or dh[mw[mi][i]][i] is the delay between two effective waves m=mw[mi][i]
4969: and mw[mi+1][i]. dh depends on stepm.
1.227 brouard 4970: */
1.126 brouard 4971:
1.224 brouard 4972: int i=0, mi=0, m=0, mli=0;
1.126 brouard 4973: /* int j, k=0,jk, ju, jl,jmin=1e+5, jmax=-1;
4974: double sum=0., jmean=0.;*/
1.224 brouard 4975: int first=0, firstwo=0, firsthree=0, firstfour=0, firstfiv=0;
1.126 brouard 4976: int j, k=0,jk, ju, jl;
4977: double sum=0.;
4978: first=0;
1.214 brouard 4979: firstwo=0;
1.217 brouard 4980: firsthree=0;
1.218 brouard 4981: firstfour=0;
1.164 brouard 4982: jmin=100000;
1.126 brouard 4983: jmax=-1;
4984: jmean=0.;
1.224 brouard 4985:
4986: /* Treating live states */
1.214 brouard 4987: for(i=1; i<=imx; i++){ /* For simple cases and if state is death */
1.224 brouard 4988: mi=0; /* First valid wave */
1.227 brouard 4989: mli=0; /* Last valid wave */
1.126 brouard 4990: m=firstpass;
1.214 brouard 4991: while(s[m][i] <= nlstate){ /* a live state */
1.227 brouard 4992: if(m >firstpass && s[m][i]==s[m-1][i] && mint[m][i]==mint[m-1][i] && anint[m][i]==anint[m-1][i]){/* Two succesive identical information on wave m */
4993: mli=m-1;/* mw[++mi][i]=m-1; */
4994: }else if(s[m][i]>=1 || s[m][i]==-4 || s[m][i]==-5){ /* Since 0.98r4 if status=-2 vital status is really unknown, wave should be skipped */
4995: mw[++mi][i]=m;
4996: mli=m;
1.224 brouard 4997: } /* else might be a useless wave -1 and mi is not incremented and mw[mi] not updated */
4998: if(m < lastpass){ /* m < lastpass, standard case */
1.227 brouard 4999: m++; /* mi gives the "effective" current wave, m the current wave, go to next wave by incrementing m */
1.216 brouard 5000: }
1.227 brouard 5001: else{ /* m >= lastpass, eventual special issue with warning */
1.224 brouard 5002: #ifdef UNKNOWNSTATUSNOTCONTRIBUTING
1.227 brouard 5003: break;
1.224 brouard 5004: #else
1.227 brouard 5005: if(s[m][i]==-1 && (int) andc[i] == 9999 && (int)anint[m][i] != 9999){
5006: if(firsthree == 0){
1.262 brouard 5007: printf("Information! Unknown status for individual %ld line=%d occurred at last wave %d at known date %d/%d. Please, check if your unknown date of death %d/%d means a live state %d at wave %d. This case(%d)/wave(%d) contributes to the likelihood as 1-p%d%d .\nOthers in log file only\n",num[i],i,lastpass,(int)mint[m][i],(int)anint[m][i], (int) moisdc[i], (int) andc[i], s[m][i], m, i, m, s[m][i], nlstate+ndeath);
1.227 brouard 5008: firsthree=1;
5009: }
1.262 brouard 5010: fprintf(ficlog,"Information! Unknown status for individual %ld line=%d occurred at last wave %d at known date %d/%d. Please, check if your unknown date of death %d/%d means a live state %d at wave %d. This case(%d)/wave(%d) contributes to the likelihood as 1-p%d%d .\n",num[i],i,lastpass,(int)mint[m][i],(int)anint[m][i], (int) moisdc[i], (int) andc[i], s[m][i], m, i, m, s[m][i], nlstate+ndeath);
1.227 brouard 5011: mw[++mi][i]=m;
5012: mli=m;
5013: }
5014: if(s[m][i]==-2){ /* Vital status is really unknown */
5015: nbwarn++;
5016: if((int)anint[m][i] == 9999){ /* Has the vital status really been verified? */
5017: printf("Warning! Vital status for individual %ld (line=%d) at last wave %d interviewed at date %d/%d is unknown %d. Please, check if the vital status and the date of death %d/%d are really unknown. This case (%d)/wave (%d) is skipped, no contribution to likelihood.\nOthers in log file only\n",num[i],i,lastpass,(int)mint[m][i],(int)anint[m][i], s[m][i], (int) moisdc[i], (int) andc[i], i, m);
5018: fprintf(ficlog,"Warning! Vital status for individual %ld (line=%d) at last wave %d interviewed at date %d/%d is unknown %d. Please, check if the vital status and the date of death %d/%d are really unknown. This case (%d)/wave (%d) is skipped, no contribution to likelihood.\n",num[i],i,lastpass,(int)mint[m][i],(int)anint[m][i], s[m][i], (int) moisdc[i], (int) andc[i], i, m);
5019: }
5020: break;
5021: }
5022: break;
1.224 brouard 5023: #endif
1.227 brouard 5024: }/* End m >= lastpass */
1.126 brouard 5025: }/* end while */
1.224 brouard 5026:
1.227 brouard 5027: /* mi is the last effective wave, m is lastpass, mw[j][i] gives the # of j-th effective wave for individual i */
1.216 brouard 5028: /* After last pass */
1.224 brouard 5029: /* Treating death states */
1.214 brouard 5030: if (s[m][i] > nlstate){ /* In a death state */
1.227 brouard 5031: /* if( mint[m][i]==mdc[m][i] && anint[m][i]==andc[m][i]){ /\* same date of death and date of interview *\/ */
5032: /* } */
1.126 brouard 5033: mi++; /* Death is another wave */
5034: /* if(mi==0) never been interviewed correctly before death */
1.227 brouard 5035: /* Only death is a correct wave */
1.126 brouard 5036: mw[mi][i]=m;
1.257 brouard 5037: } /* else not in a death state */
1.224 brouard 5038: #ifndef DISPATCHINGKNOWNDEATHAFTERLASTWAVE
1.257 brouard 5039: else if ((int) andc[i] != 9999) { /* Date of death is known */
1.218 brouard 5040: if ((int)anint[m][i]!= 9999) { /* date of last interview is known */
1.227 brouard 5041: if((andc[i]+moisdc[i]/12.) <=(anint[m][i]+mint[m][i]/12.)){ /* death occured before last wave and status should have been death instead of -1 */
5042: nbwarn++;
5043: if(firstfiv==0){
5044: printf("Warning! Death for individual %ld line=%d occurred at %d/%d before last wave %d interviewed at %d/%d and should have been coded as death instead of '%d'. This case (%d)/wave (%d) is contributing to likelihood.\nOthers in log file only\n",num[i],i,(int) moisdc[i], (int) andc[i], lastpass,(int)mint[m][i],(int)anint[m][i], s[m][i], i,m );
5045: firstfiv=1;
5046: }else{
5047: fprintf(ficlog,"Warning! Death for individual %ld line=%d occurred at %d/%d before last wave %d interviewed at %d/%d and should have been coded as death instead of '%d'. This case (%d)/wave (%d) is contributing to likelihood.\n",num[i],i,(int) moisdc[i], (int) andc[i], lastpass,(int)mint[m][i],(int)anint[m][i], s[m][i], i,m );
5048: }
5049: }else{ /* Death occured afer last wave potential bias */
5050: nberr++;
5051: if(firstwo==0){
1.257 brouard 5052: printf("Error! Death for individual %ld line=%d occurred at %d/%d after last wave %d interviewed at %d/%d. Potential bias if other individuals are still alive at this date but ignored. This case (%d)/wave (%d) is skipped, no contribution to likelihood. Please add a new fictive wave at the date of last vital status scan, with a dead status or alive but unknown state status (-1). See documentation\nOthers in log file only\n",num[i],i,(int) moisdc[i], (int) andc[i], lastpass,(int)mint[m][i],(int)anint[m][i], i,m );
1.227 brouard 5053: firstwo=1;
5054: }
1.257 brouard 5055: fprintf(ficlog,"Error! Death for individual %ld line=%d occurred at %d/%d after last wave %d interviewed at %d/%d. Potential bias if other individuals are still alive at this date but ignored. This case (%d)/wave (%d) is skipped, no contribution to likelihood. Please add a new fictive wave at the date of last vital status scan, with a dead status or alive but unknown state status (-1). See documentation\n\n",num[i],i,(int) moisdc[i], (int) andc[i], lastpass,(int)mint[m][i],(int)anint[m][i], i,m );
1.227 brouard 5056: }
1.257 brouard 5057: }else{ /* if date of interview is unknown */
1.227 brouard 5058: /* death is known but not confirmed by death status at any wave */
5059: if(firstfour==0){
5060: printf("Error! Death for individual %ld line=%d occurred %d/%d but not confirmed by any death status for any wave, including last wave %d at unknown date %d/%d. Potential bias if other individuals are still alive at this date but ignored. This case (%d)/wave (%d) is skipped, no contribution to likelihood.\nOthers in log file only\n",num[i],i,(int) moisdc[i], (int) andc[i], lastpass,(int)mint[m][i],(int)anint[m][i], i,m );
5061: firstfour=1;
5062: }
5063: fprintf(ficlog,"Error! Death for individual %ld line=%d occurred %d/%d but not confirmed by any death status for any wave, including last wave %d at unknown date %d/%d. Potential bias if other individuals are still alive at this date but ignored. This case (%d)/wave (%d) is skipped, no contribution to likelihood.\n",num[i],i,(int) moisdc[i], (int) andc[i], lastpass,(int)mint[m][i],(int)anint[m][i], i,m );
1.214 brouard 5064: }
1.224 brouard 5065: } /* end if date of death is known */
5066: #endif
5067: wav[i]=mi; /* mi should be the last effective wave (or mli) */
5068: /* wav[i]=mw[mi][i]; */
1.126 brouard 5069: if(mi==0){
5070: nbwarn++;
5071: if(first==0){
1.227 brouard 5072: printf("Warning! No valid information for individual %ld line=%d (skipped) and may be others, see log file\n",num[i],i);
5073: first=1;
1.126 brouard 5074: }
5075: if(first==1){
1.227 brouard 5076: fprintf(ficlog,"Warning! No valid information for individual %ld line=%d (skipped)\n",num[i],i);
1.126 brouard 5077: }
5078: } /* end mi==0 */
5079: } /* End individuals */
1.214 brouard 5080: /* wav and mw are no more changed */
1.223 brouard 5081:
1.214 brouard 5082:
1.126 brouard 5083: for(i=1; i<=imx; i++){
5084: for(mi=1; mi<wav[i];mi++){
5085: if (stepm <=0)
1.227 brouard 5086: dh[mi][i]=1;
1.126 brouard 5087: else{
1.260 brouard 5088: if (s[mw[mi+1][i]][i] > nlstate) { /* A death, but what if date is unknown? */
1.227 brouard 5089: if (agedc[i] < 2*AGESUP) {
5090: j= rint(agedc[i]*12-agev[mw[mi][i]][i]*12);
5091: if(j==0) j=1; /* Survives at least one month after exam */
5092: else if(j<0){
5093: nberr++;
5094: printf("Error! Negative delay (%d to death) between waves %d and %d of individual %ld at line %d who is aged %.1f with statuses from %d to %d\n ",j,mw[mi][i],mw[mi+1][i],num[i], i,agev[mw[mi][i]][i],s[mw[mi][i]][i] ,s[mw[mi+1][i]][i]);
5095: j=1; /* Temporary Dangerous patch */
5096: printf(" We assumed that the date of interview was correct (and not the date of death) and postponed the death %d month(s) (one stepm) after the interview. You MUST fix the contradiction between dates.\n",stepm);
5097: fprintf(ficlog,"Error! Negative delay (%d to death) between waves %d and %d of individual %ld at line %d who is aged %.1f with statuses from %d to %d\n ",j,mw[mi][i],mw[mi+1][i],num[i], i,agev[mw[mi][i]][i],s[mw[mi][i]][i] ,s[mw[mi+1][i]][i]);
5098: fprintf(ficlog," We assumed that the date of interview was correct (and not the date of death) and postponed the death %d month(s) (one stepm) after the interview. You MUST fix the contradiction between dates.\n",stepm);
5099: }
5100: k=k+1;
5101: if (j >= jmax){
5102: jmax=j;
5103: ijmax=i;
5104: }
5105: if (j <= jmin){
5106: jmin=j;
5107: ijmin=i;
5108: }
5109: sum=sum+j;
5110: /*if (j<0) printf("j=%d num=%d \n",j,i);*/
5111: /* printf("%d %d %d %d\n", s[mw[mi][i]][i] ,s[mw[mi+1][i]][i],j,i);*/
5112: }
5113: }
5114: else{
5115: j= rint( (agev[mw[mi+1][i]][i]*12 - agev[mw[mi][i]][i]*12));
1.126 brouard 5116: /* if (j<0) printf("%d %lf %lf %d %d %d\n", i,agev[mw[mi+1][i]][i], agev[mw[mi][i]][i],j,s[mw[mi][i]][i] ,s[mw[mi+1][i]][i]); */
1.223 brouard 5117:
1.227 brouard 5118: k=k+1;
5119: if (j >= jmax) {
5120: jmax=j;
5121: ijmax=i;
5122: }
5123: else if (j <= jmin){
5124: jmin=j;
5125: ijmin=i;
5126: }
5127: /* if (j<10) printf("j=%d jmin=%d num=%d ",j,jmin,i); */
5128: /*printf("%d %lf %d %d %d\n", i,agev[mw[mi][i]][i],j,s[mw[mi][i]][i] ,s[mw[mi+1][i]][i]);*/
5129: if(j<0){
5130: nberr++;
5131: printf("Error! Negative delay (%d) between waves %d and %d of individual %ld at line %d who is aged %.1f with statuses from %d to %d\n ",j,mw[mi][i],mw[mi+1][i],num[i], i,agev[mw[mi][i]][i],s[mw[mi][i]][i] ,s[mw[mi+1][i]][i]);
5132: fprintf(ficlog,"Error! Negative delay (%d) between waves %d and %d of individual %ld at line %d who is aged %.1f with statuses from %d to %d\n ",j,mw[mi][i],mw[mi+1][i],num[i], i,agev[mw[mi][i]][i],s[mw[mi][i]][i] ,s[mw[mi+1][i]][i]);
5133: }
5134: sum=sum+j;
5135: }
5136: jk= j/stepm;
5137: jl= j -jk*stepm;
5138: ju= j -(jk+1)*stepm;
5139: if(mle <=1){ /* only if we use a the linear-interpoloation pseudo-likelihood */
5140: if(jl==0){
5141: dh[mi][i]=jk;
5142: bh[mi][i]=0;
5143: }else{ /* We want a negative bias in order to only have interpolation ie
5144: * to avoid the price of an extra matrix product in likelihood */
5145: dh[mi][i]=jk+1;
5146: bh[mi][i]=ju;
5147: }
5148: }else{
5149: if(jl <= -ju){
5150: dh[mi][i]=jk;
5151: bh[mi][i]=jl; /* bias is positive if real duration
5152: * is higher than the multiple of stepm and negative otherwise.
5153: */
5154: }
5155: else{
5156: dh[mi][i]=jk+1;
5157: bh[mi][i]=ju;
5158: }
5159: if(dh[mi][i]==0){
5160: dh[mi][i]=1; /* At least one step */
5161: bh[mi][i]=ju; /* At least one step */
5162: /* printf(" bh=%d ju=%d jl=%d dh=%d jk=%d stepm=%d %d\n",bh[mi][i],ju,jl,dh[mi][i],jk,stepm,i);*/
5163: }
5164: } /* end if mle */
1.126 brouard 5165: }
5166: } /* end wave */
5167: }
5168: jmean=sum/k;
5169: printf("Delay (in months) between two waves Min=%d (for indiviudal %ld) Max=%d (%ld) Mean=%f\n\n ",jmin, num[ijmin], jmax, num[ijmax], jmean);
1.141 brouard 5170: fprintf(ficlog,"Delay (in months) between two waves Min=%d (for indiviudal %d) Max=%d (%d) Mean=%f\n\n ",jmin, ijmin, jmax, ijmax, jmean);
1.227 brouard 5171: }
1.126 brouard 5172:
5173: /*********** Tricode ****************************/
1.220 brouard 5174: void tricode(int *cptcov, int *Tvar, int **nbcode, int imx, int *Ndum)
1.242 brouard 5175: {
5176: /**< Uses cptcovn+2*cptcovprod as the number of covariates */
5177: /* Tvar[i]=atoi(stre); find 'n' in Vn and stores in Tvar. If model=V2+V1 Tvar[1]=2 and Tvar[2]=1
5178: * Boring subroutine which should only output nbcode[Tvar[j]][k]
5179: * Tvar[5] in V2+V1+V3*age+V2*V4 is 4 (V4) even it is a time varying or quantitative variable
5180: * nbcode[Tvar[5]][1]= nbcode[4][1]=0, nbcode[4][2]=1 (usually);
5181: */
1.130 brouard 5182:
1.242 brouard 5183: int ij=1, k=0, j=0, i=0, maxncov=NCOVMAX;
5184: int modmaxcovj=0; /* Modality max of covariates j */
5185: int cptcode=0; /* Modality max of covariates j */
5186: int modmincovj=0; /* Modality min of covariates j */
1.145 brouard 5187:
5188:
1.242 brouard 5189: /* cptcoveff=0; */
5190: /* *cptcov=0; */
1.126 brouard 5191:
1.242 brouard 5192: for (k=1; k <= maxncov; k++) ncodemax[k]=0; /* Horrible constant again replaced by NCOVMAX */
1.126 brouard 5193:
1.242 brouard 5194: /* Loop on covariates without age and products and no quantitative variable */
5195: /* for (j=1; j<=(cptcovs); j++) { /\* From model V1 + V2*age+ V3 + V3*V4 keeps V1 + V3 = 2 only *\/ */
5196: for (k=1; k<=cptcovt; k++) { /* From model V1 + V2*age + V3 + V3*V4 keeps V1 + V3 = 2 only */
5197: for (j=-1; (j < maxncov); j++) Ndum[j]=0;
5198: if(Dummy[k]==0 && Typevar[k] !=1){ /* Dummy covariate and not age product */
5199: switch(Fixed[k]) {
5200: case 0: /* Testing on fixed dummy covariate, simple or product of fixed */
5201: for (i=1; i<=imx; i++) { /* Loop on individuals: reads the data file to get the maximum value of the modality of this covariate Vj*/
5202: ij=(int)(covar[Tvar[k]][i]);
5203: /* ij=0 or 1 or -1. Value of the covariate Tvar[j] for individual i
5204: * If product of Vn*Vm, still boolean *:
5205: * If it was coded 1, 2, 3, 4 should be splitted into 3 boolean variables
5206: * 1 => 0 0 0, 2 => 0 0 1, 3 => 0 1 1, 4=1 0 0 */
5207: /* Finds for covariate j, n=Tvar[j] of Vn . ij is the
5208: modality of the nth covariate of individual i. */
5209: if (ij > modmaxcovj)
5210: modmaxcovj=ij;
5211: else if (ij < modmincovj)
5212: modmincovj=ij;
5213: if ((ij < -1) && (ij > NCOVMAX)){
5214: printf( "Error: minimal is less than -1 or maximal is bigger than %d. Exiting. \n", NCOVMAX );
5215: exit(1);
5216: }else
5217: Ndum[ij]++; /*counts and stores the occurence of this modality 0, 1, -1*/
5218: /* If coded 1, 2, 3 , counts the number of 1 Ndum[1], number of 2, Ndum[2], etc */
5219: /*printf("i=%d ij=%d Ndum[ij]=%d imx=%d",i,ij,Ndum[ij],imx);*/
5220: /* getting the maximum value of the modality of the covariate
5221: (should be 0 or 1 now) Tvar[j]. If V=sex and male is coded 0 and
5222: female ies 1, then modmaxcovj=1.
5223: */
5224: } /* end for loop on individuals i */
5225: printf(" Minimal and maximal values of %d th (fixed) covariate V%d: min=%d max=%d \n", k, Tvar[k], modmincovj, modmaxcovj);
5226: fprintf(ficlog," Minimal and maximal values of %d th (fixed) covariate V%d: min=%d max=%d \n", k, Tvar[k], modmincovj, modmaxcovj);
5227: cptcode=modmaxcovj;
5228: /* Ndum[0] = frequency of 0 for model-covariate j, Ndum[1] frequency of 1 etc. */
5229: /*for (i=0; i<=cptcode; i++) {*/
5230: for (j=modmincovj; j<=modmaxcovj; j++) { /* j=-1 ? 0 and 1*//* For each value j of the modality of model-cov k */
5231: printf("Frequencies of (fixed) covariate %d ie V%d with value %d: %d\n", k, Tvar[k], j, Ndum[j]);
5232: fprintf(ficlog, "Frequencies of (fixed) covariate %d ie V%d with value %d: %d\n", k, Tvar[k], j, Ndum[j]);
5233: if( Ndum[j] != 0 ){ /* Counts if nobody answered modality j ie empty modality, we skip it and reorder */
5234: if( j != -1){
5235: ncodemax[k]++; /* ncodemax[k]= Number of modalities of the k th
5236: covariate for which somebody answered excluding
5237: undefined. Usually 2: 0 and 1. */
5238: }
5239: ncodemaxwundef[k]++; /* ncodemax[j]= Number of modalities of the k th
5240: covariate for which somebody answered including
5241: undefined. Usually 3: -1, 0 and 1. */
5242: } /* In fact ncodemax[k]=2 (dichotom. variables only) but it could be more for
5243: * historical reasons: 3 if coded 1, 2, 3 and 4 and Ndum[2]=0 */
5244: } /* Ndum[-1] number of undefined modalities */
1.231 brouard 5245:
1.242 brouard 5246: /* j is a covariate, n=Tvar[j] of Vn; Fills nbcode */
5247: /* For covariate j, modalities could be 1, 2, 3, 4, 5, 6, 7. */
5248: /* If Ndum[1]=0, Ndum[2]=0, Ndum[3]= 635, Ndum[4]=0, Ndum[5]=0, Ndum[6]=27, Ndum[7]=125; */
5249: /* modmincovj=3; modmaxcovj = 7; */
5250: /* There are only 3 modalities non empty 3, 6, 7 (or 2 if 27 is too few) : ncodemax[j]=3; */
5251: /* which will be coded 0, 1, 2 which in binary on 2=3-1 digits are 0=00 1=01, 2=10; */
5252: /* defining two dummy variables: variables V1_1 and V1_2.*/
5253: /* nbcode[Tvar[j]][ij]=k; */
5254: /* nbcode[Tvar[j]][1]=0; */
5255: /* nbcode[Tvar[j]][2]=1; */
5256: /* nbcode[Tvar[j]][3]=2; */
5257: /* To be continued (not working yet). */
5258: ij=0; /* ij is similar to i but can jump over null modalities */
5259: for (i=modmincovj; i<=modmaxcovj; i++) { /* i= 1 to 2 for dichotomous, or from 1 to 3 or from -1 or 0 to 1 currently*/
5260: if (Ndum[i] == 0) { /* If nobody responded to this modality k */
5261: break;
5262: }
5263: ij++;
5264: nbcode[Tvar[k]][ij]=i; /* stores the original value of modality i in an array nbcode, ij modality from 1 to last non-nul modality. nbcode[1][1]=0 nbcode[1][2]=1*/
5265: cptcode = ij; /* New max modality for covar j */
5266: } /* end of loop on modality i=-1 to 1 or more */
5267: break;
5268: case 1: /* Testing on varying covariate, could be simple and
5269: * should look at waves or product of fixed *
5270: * varying. No time to test -1, assuming 0 and 1 only */
5271: ij=0;
5272: for(i=0; i<=1;i++){
5273: nbcode[Tvar[k]][++ij]=i;
5274: }
5275: break;
5276: default:
5277: break;
5278: } /* end switch */
5279: } /* end dummy test */
5280:
5281: /* for (k=0; k<= cptcode; k++) { /\* k=-1 ? k=0 to 1 *\//\* Could be 1 to 4 *\//\* cptcode=modmaxcovj *\/ */
5282: /* /\*recode from 0 *\/ */
5283: /* k is a modality. If we have model=V1+V1*sex */
5284: /* then: nbcode[1][1]=0 ; nbcode[1][2]=1; nbcode[2][1]=0 ; nbcode[2][2]=1; */
5285: /* But if some modality were not used, it is recoded from 0 to a newer modmaxcovj=cptcode *\/ */
5286: /* } */
5287: /* /\* cptcode = ij; *\/ /\* New max modality for covar j *\/ */
5288: /* if (ij > ncodemax[j]) { */
5289: /* printf( " Error ij=%d > ncodemax[%d]=%d\n", ij, j, ncodemax[j]); */
5290: /* fprintf(ficlog, " Error ij=%d > ncodemax[%d]=%d\n", ij, j, ncodemax[j]); */
5291: /* break; */
5292: /* } */
5293: /* } /\* end of loop on modality k *\/ */
5294: } /* end of loop on model-covariate j. nbcode[Tvarj][1]=0 and nbcode[Tvarj][2]=1 sets the value of covariate j*/
5295:
5296: for (k=-1; k< maxncov; k++) Ndum[k]=0;
5297: /* Look at fixed dummy (single or product) covariates to check empty modalities */
5298: for (i=1; i<=ncovmodel-2-nagesqr; i++) { /* -2, cste and age and eventually age*age */
5299: /* Listing of all covariables in statement model to see if some covariates appear twice. For example, V1 appears twice in V1+V1*V2.*/
5300: ij=Tvar[i]; /* Tvar 5,4,3,6,5,7,1,4 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V4*age */
5301: Ndum[ij]++; /* Count the # of 1, 2 etc: {1,1,1,2,2,1,1} because V1 once, V2 once, two V4 and V5 in above */
5302: /* V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1, {2, 1, 1, 1, 2, 1, 1, 0, 0} */
5303: } /* V4+V3+V5, Ndum[1]@5={0, 0, 1, 1, 1} */
5304:
5305: ij=0;
5306: /* for (i=0; i<= maxncov-1; i++) { /\* modmaxcovj is unknown here. Only Ndum[2(V2),3(age*V3), 5(V3*V2) 6(V1*V4) *\/ */
5307: for (k=1; k<= cptcovt; k++) { /* modmaxcovj is unknown here. Only Ndum[2(V2),3(age*V3), 5(V3*V2) 6(V1*V4) */
5308: /*printf("Ndum[%d]=%d\n",i, Ndum[i]);*/
5309: /* if((Ndum[i]!=0) && (i<=ncovcol)){ /\* Tvar[i] <= ncovmodel ? *\/ */
5310: if(Ndum[Tvar[k]]!=0 && Dummy[k] == 0 && Typevar[k]==0){ /* Only Dummy and non empty in the model */
5311: /* If product not in single variable we don't print results */
5312: /*printf("diff Ndum[%d]=%d\n",i, Ndum[i]);*/
5313: ++ij;/* V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1, */
5314: Tvaraff[ij]=Tvar[k]; /* For printing combination *//* V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1, Tvar {5, 4, 3, 6, 5, 2, 7, 1, 1} Tvaraff={4, 3, 1} V4, V3, V1*/
5315: Tmodelind[ij]=k; /* Tmodelind: index in model of dummies Tmodelind[1]=2 V4: pos=2; V3: pos=3, V1=9 {2, 3, 9, ?, ?,} */
5316: TmodelInvind[ij]=Tvar[k]- ncovcol-nqv; /* Inverse TmodelInvind[2=V4]=2 second dummy varying cov (V4)4-1-1 {0, 2, 1, } TmodelInvind[3]=1 */
5317: if(Fixed[k]!=0)
5318: anyvaryingduminmodel=1;
5319: /* }else if((Ndum[i]!=0) && (i<=ncovcol+nqv)){ */
5320: /* Tvaraff[++ij]=-10; /\* Dont'n know how to treat quantitative variables yet *\/ */
5321: /* }else if((Ndum[i]!=0) && (i<=ncovcol+nqv+ntv)){ */
5322: /* Tvaraff[++ij]=i; /\*For printing (unclear) *\/ */
5323: /* }else if((Ndum[i]!=0) && (i<=ncovcol+nqv+ntv+nqtv)){ */
5324: /* Tvaraff[++ij]=-20; /\* Dont'n know how to treat quantitative variables yet *\/ */
5325: }
5326: } /* Tvaraff[1]@5 {3, 4, -20, 0, 0} Very strange */
5327: /* ij--; */
5328: /* cptcoveff=ij; /\*Number of total covariates*\/ */
5329: *cptcov=ij; /*Number of total real effective covariates: effective
5330: * because they can be excluded from the model and real
5331: * if in the model but excluded because missing values, but how to get k from ij?*/
5332: for(j=ij+1; j<= cptcovt; j++){
5333: Tvaraff[j]=0;
5334: Tmodelind[j]=0;
5335: }
5336: for(j=ntveff+1; j<= cptcovt; j++){
5337: TmodelInvind[j]=0;
5338: }
5339: /* To be sorted */
5340: ;
5341: }
1.126 brouard 5342:
1.145 brouard 5343:
1.126 brouard 5344: /*********** Health Expectancies ****************/
5345:
1.235 brouard 5346: void evsij(double ***eij, double x[], int nlstate, int stepm, int bage, int fage, double **oldm, double **savm, int cij, int estepm,char strstart[], int nres )
1.126 brouard 5347:
5348: {
5349: /* Health expectancies, no variances */
1.164 brouard 5350: int i, j, nhstepm, hstepm, h, nstepm;
1.126 brouard 5351: int nhstepma, nstepma; /* Decreasing with age */
5352: double age, agelim, hf;
5353: double ***p3mat;
5354: double eip;
5355:
1.238 brouard 5356: /* pstamp(ficreseij); */
1.126 brouard 5357: fprintf(ficreseij,"# (a) Life expectancies by health status at initial age and (b) health expectancies by health status at initial age\n");
5358: fprintf(ficreseij,"# Age");
5359: for(i=1; i<=nlstate;i++){
5360: for(j=1; j<=nlstate;j++){
5361: fprintf(ficreseij," e%1d%1d ",i,j);
5362: }
5363: fprintf(ficreseij," e%1d. ",i);
5364: }
5365: fprintf(ficreseij,"\n");
5366:
5367:
5368: if(estepm < stepm){
5369: printf ("Problem %d lower than %d\n",estepm, stepm);
5370: }
5371: else hstepm=estepm;
5372: /* We compute the life expectancy from trapezoids spaced every estepm months
5373: * This is mainly to measure the difference between two models: for example
5374: * if stepm=24 months pijx are given only every 2 years and by summing them
5375: * we are calculating an estimate of the Life Expectancy assuming a linear
5376: * progression in between and thus overestimating or underestimating according
5377: * to the curvature of the survival function. If, for the same date, we
5378: * estimate the model with stepm=1 month, we can keep estepm to 24 months
5379: * to compare the new estimate of Life expectancy with the same linear
5380: * hypothesis. A more precise result, taking into account a more precise
5381: * curvature will be obtained if estepm is as small as stepm. */
5382:
5383: /* For example we decided to compute the life expectancy with the smallest unit */
5384: /* hstepm beeing the number of stepms, if hstepm=1 the length of hstepm is stepm.
5385: nhstepm is the number of hstepm from age to agelim
5386: nstepm is the number of stepm from age to agelin.
5387: Look at hpijx to understand the reason of that which relies in memory size
5388: and note for a fixed period like estepm months */
5389: /* We decided (b) to get a life expectancy respecting the most precise curvature of the
5390: survival function given by stepm (the optimization length). Unfortunately it
5391: means that if the survival funtion is printed only each two years of age and if
5392: you sum them up and add 1 year (area under the trapezoids) you won't get the same
5393: results. So we changed our mind and took the option of the best precision.
5394: */
5395: hstepm=hstepm/stepm; /* Typically in stepm units, if stepm=6 & estepm=24 , = 24/6 months = 4 */
5396:
5397: agelim=AGESUP;
5398: /* If stepm=6 months */
5399: /* Computed by stepm unit matrices, product of hstepm matrices, stored
5400: in an array of nhstepm length: nhstepm=10, hstepm=4, stepm=6 months */
5401:
5402: /* nhstepm age range expressed in number of stepm */
5403: nstepm=(int) rint((agelim-bage)*YEARM/stepm); /* Biggest nstepm */
5404: /* Typically if 20 years nstepm = 20*12/6=40 stepm */
5405: /* if (stepm >= YEARM) hstepm=1;*/
5406: nhstepm = nstepm/hstepm;/* Expressed in hstepm, typically nhstepm=40/4=10 */
5407: p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5408:
5409: for (age=bage; age<=fage; age ++){
5410: nstepma=(int) rint((agelim-bage)*YEARM/stepm); /* Biggest nstepm */
5411: /* Typically if 20 years nstepm = 20*12/6=40 stepm */
5412: /* if (stepm >= YEARM) hstepm=1;*/
5413: nhstepma = nstepma/hstepm;/* Expressed in hstepm, typically nhstepma=40/4=10 */
5414:
5415: /* If stepm=6 months */
5416: /* Computed by stepm unit matrices, product of hstepma matrices, stored
5417: in an array of nhstepma length: nhstepma=10, hstepm=4, stepm=6 months */
5418:
1.235 brouard 5419: hpxij(p3mat,nhstepma,age,hstepm,x,nlstate,stepm,oldm, savm, cij, nres);
1.126 brouard 5420:
5421: hf=hstepm*stepm/YEARM; /* Duration of hstepm expressed in year unit. */
5422:
5423: printf("%d|",(int)age);fflush(stdout);
5424: fprintf(ficlog,"%d|",(int)age);fflush(ficlog);
5425:
5426: /* Computing expectancies */
5427: for(i=1; i<=nlstate;i++)
5428: for(j=1; j<=nlstate;j++)
5429: for (h=0, eij[i][j][(int)age]=0; h<=nhstepm-1; h++){
5430: eij[i][j][(int)age] += (p3mat[i][j][h]+p3mat[i][j][h+1])/2.0*hf;
5431:
5432: /* if((int)age==70)printf("i=%2d,j=%2d,h=%2d,age=%3d,%9.4f,%9.4f,%9.4f\n",i,j,h,(int)age,p3mat[i][j][h],hf,eij[i][j][(int)age]);*/
5433:
5434: }
5435:
5436: fprintf(ficreseij,"%3.0f",age );
5437: for(i=1; i<=nlstate;i++){
5438: eip=0;
5439: for(j=1; j<=nlstate;j++){
5440: eip +=eij[i][j][(int)age];
5441: fprintf(ficreseij,"%9.4f", eij[i][j][(int)age] );
5442: }
5443: fprintf(ficreseij,"%9.4f", eip );
5444: }
5445: fprintf(ficreseij,"\n");
5446:
5447: }
5448: free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5449: printf("\n");
5450: fprintf(ficlog,"\n");
5451:
5452: }
5453:
1.235 brouard 5454: void cvevsij(double ***eij, double x[], int nlstate, int stepm, int bage, int fage, double **oldm, double **savm, int cij, int estepm,double delti[],double **matcov,char strstart[], int nres )
1.126 brouard 5455:
5456: {
5457: /* Covariances of health expectancies eij and of total life expectancies according
1.222 brouard 5458: to initial status i, ei. .
1.126 brouard 5459: */
5460: int i, j, nhstepm, hstepm, h, nstepm, k, cptj, cptj2, i2, j2, ij, ji;
5461: int nhstepma, nstepma; /* Decreasing with age */
5462: double age, agelim, hf;
5463: double ***p3matp, ***p3matm, ***varhe;
5464: double **dnewm,**doldm;
5465: double *xp, *xm;
5466: double **gp, **gm;
5467: double ***gradg, ***trgradg;
5468: int theta;
5469:
5470: double eip, vip;
5471:
5472: varhe=ma3x(1,nlstate*nlstate,1,nlstate*nlstate,(int) bage, (int) fage);
5473: xp=vector(1,npar);
5474: xm=vector(1,npar);
5475: dnewm=matrix(1,nlstate*nlstate,1,npar);
5476: doldm=matrix(1,nlstate*nlstate,1,nlstate*nlstate);
5477:
5478: pstamp(ficresstdeij);
5479: fprintf(ficresstdeij,"# Health expectancies with standard errors\n");
5480: fprintf(ficresstdeij,"# Age");
5481: for(i=1; i<=nlstate;i++){
5482: for(j=1; j<=nlstate;j++)
5483: fprintf(ficresstdeij," e%1d%1d (SE)",i,j);
5484: fprintf(ficresstdeij," e%1d. ",i);
5485: }
5486: fprintf(ficresstdeij,"\n");
5487:
5488: pstamp(ficrescveij);
5489: fprintf(ficrescveij,"# Subdiagonal matrix of covariances of health expectancies by age: cov(eij,ekl)\n");
5490: fprintf(ficrescveij,"# Age");
5491: for(i=1; i<=nlstate;i++)
5492: for(j=1; j<=nlstate;j++){
5493: cptj= (j-1)*nlstate+i;
5494: for(i2=1; i2<=nlstate;i2++)
5495: for(j2=1; j2<=nlstate;j2++){
5496: cptj2= (j2-1)*nlstate+i2;
5497: if(cptj2 <= cptj)
5498: fprintf(ficrescveij," %1d%1d,%1d%1d",i,j,i2,j2);
5499: }
5500: }
5501: fprintf(ficrescveij,"\n");
5502:
5503: if(estepm < stepm){
5504: printf ("Problem %d lower than %d\n",estepm, stepm);
5505: }
5506: else hstepm=estepm;
5507: /* We compute the life expectancy from trapezoids spaced every estepm months
5508: * This is mainly to measure the difference between two models: for example
5509: * if stepm=24 months pijx are given only every 2 years and by summing them
5510: * we are calculating an estimate of the Life Expectancy assuming a linear
5511: * progression in between and thus overestimating or underestimating according
5512: * to the curvature of the survival function. If, for the same date, we
5513: * estimate the model with stepm=1 month, we can keep estepm to 24 months
5514: * to compare the new estimate of Life expectancy with the same linear
5515: * hypothesis. A more precise result, taking into account a more precise
5516: * curvature will be obtained if estepm is as small as stepm. */
5517:
5518: /* For example we decided to compute the life expectancy with the smallest unit */
5519: /* hstepm beeing the number of stepms, if hstepm=1 the length of hstepm is stepm.
5520: nhstepm is the number of hstepm from age to agelim
5521: nstepm is the number of stepm from age to agelin.
5522: Look at hpijx to understand the reason of that which relies in memory size
5523: and note for a fixed period like estepm months */
5524: /* We decided (b) to get a life expectancy respecting the most precise curvature of the
5525: survival function given by stepm (the optimization length). Unfortunately it
5526: means that if the survival funtion is printed only each two years of age and if
5527: you sum them up and add 1 year (area under the trapezoids) you won't get the same
5528: results. So we changed our mind and took the option of the best precision.
5529: */
5530: hstepm=hstepm/stepm; /* Typically in stepm units, if stepm=6 & estepm=24 , = 24/6 months = 4 */
5531:
5532: /* If stepm=6 months */
5533: /* nhstepm age range expressed in number of stepm */
5534: agelim=AGESUP;
5535: nstepm=(int) rint((agelim-bage)*YEARM/stepm);
5536: /* Typically if 20 years nstepm = 20*12/6=40 stepm */
5537: /* if (stepm >= YEARM) hstepm=1;*/
5538: nhstepm = nstepm/hstepm;/* Expressed in hstepm, typically nhstepm=40/4=10 */
5539:
5540: p3matp=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5541: p3matm=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5542: gradg=ma3x(0,nhstepm,1,npar,1,nlstate*nlstate);
5543: trgradg =ma3x(0,nhstepm,1,nlstate*nlstate,1,npar);
5544: gp=matrix(0,nhstepm,1,nlstate*nlstate);
5545: gm=matrix(0,nhstepm,1,nlstate*nlstate);
5546:
5547: for (age=bage; age<=fage; age ++){
5548: nstepma=(int) rint((agelim-bage)*YEARM/stepm); /* Biggest nstepm */
5549: /* Typically if 20 years nstepm = 20*12/6=40 stepm */
5550: /* if (stepm >= YEARM) hstepm=1;*/
5551: nhstepma = nstepma/hstepm;/* Expressed in hstepm, typically nhstepma=40/4=10 */
1.218 brouard 5552:
1.126 brouard 5553: /* If stepm=6 months */
5554: /* Computed by stepm unit matrices, product of hstepma matrices, stored
5555: in an array of nhstepma length: nhstepma=10, hstepm=4, stepm=6 months */
5556:
5557: hf=hstepm*stepm/YEARM; /* Duration of hstepm expressed in year unit. */
1.218 brouard 5558:
1.126 brouard 5559: /* Computing Variances of health expectancies */
5560: /* Gradient is computed with plus gp and minus gm. Code is duplicated in order to
5561: decrease memory allocation */
5562: for(theta=1; theta <=npar; theta++){
5563: for(i=1; i<=npar; i++){
1.222 brouard 5564: xp[i] = x[i] + (i==theta ?delti[theta]:0);
5565: xm[i] = x[i] - (i==theta ?delti[theta]:0);
1.126 brouard 5566: }
1.235 brouard 5567: hpxij(p3matp,nhstepm,age,hstepm,xp,nlstate,stepm,oldm,savm, cij, nres);
5568: hpxij(p3matm,nhstepm,age,hstepm,xm,nlstate,stepm,oldm,savm, cij, nres);
1.218 brouard 5569:
1.126 brouard 5570: for(j=1; j<= nlstate; j++){
1.222 brouard 5571: for(i=1; i<=nlstate; i++){
5572: for(h=0; h<=nhstepm-1; h++){
5573: gp[h][(j-1)*nlstate + i] = (p3matp[i][j][h]+p3matp[i][j][h+1])/2.;
5574: gm[h][(j-1)*nlstate + i] = (p3matm[i][j][h]+p3matm[i][j][h+1])/2.;
5575: }
5576: }
1.126 brouard 5577: }
1.218 brouard 5578:
1.126 brouard 5579: for(ij=1; ij<= nlstate*nlstate; ij++)
1.222 brouard 5580: for(h=0; h<=nhstepm-1; h++){
5581: gradg[h][theta][ij]= (gp[h][ij]-gm[h][ij])/2./delti[theta];
5582: }
1.126 brouard 5583: }/* End theta */
5584:
5585:
5586: for(h=0; h<=nhstepm-1; h++)
5587: for(j=1; j<=nlstate*nlstate;j++)
1.222 brouard 5588: for(theta=1; theta <=npar; theta++)
5589: trgradg[h][j][theta]=gradg[h][theta][j];
1.126 brouard 5590:
1.218 brouard 5591:
1.222 brouard 5592: for(ij=1;ij<=nlstate*nlstate;ij++)
1.126 brouard 5593: for(ji=1;ji<=nlstate*nlstate;ji++)
1.222 brouard 5594: varhe[ij][ji][(int)age] =0.;
1.218 brouard 5595:
1.222 brouard 5596: printf("%d|",(int)age);fflush(stdout);
5597: fprintf(ficlog,"%d|",(int)age);fflush(ficlog);
5598: for(h=0;h<=nhstepm-1;h++){
1.126 brouard 5599: for(k=0;k<=nhstepm-1;k++){
1.222 brouard 5600: matprod2(dnewm,trgradg[h],1,nlstate*nlstate,1,npar,1,npar,matcov);
5601: matprod2(doldm,dnewm,1,nlstate*nlstate,1,npar,1,nlstate*nlstate,gradg[k]);
5602: for(ij=1;ij<=nlstate*nlstate;ij++)
5603: for(ji=1;ji<=nlstate*nlstate;ji++)
5604: varhe[ij][ji][(int)age] += doldm[ij][ji]*hf*hf;
1.126 brouard 5605: }
5606: }
1.218 brouard 5607:
1.126 brouard 5608: /* Computing expectancies */
1.235 brouard 5609: hpxij(p3matm,nhstepm,age,hstepm,x,nlstate,stepm,oldm, savm, cij,nres);
1.126 brouard 5610: for(i=1; i<=nlstate;i++)
5611: for(j=1; j<=nlstate;j++)
1.222 brouard 5612: for (h=0, eij[i][j][(int)age]=0; h<=nhstepm-1; h++){
5613: eij[i][j][(int)age] += (p3matm[i][j][h]+p3matm[i][j][h+1])/2.0*hf;
1.218 brouard 5614:
1.222 brouard 5615: /* if((int)age==70)printf("i=%2d,j=%2d,h=%2d,age=%3d,%9.4f,%9.4f,%9.4f\n",i,j,h,(int)age,p3mat[i][j][h],hf,eij[i][j][(int)age]);*/
1.218 brouard 5616:
1.222 brouard 5617: }
1.218 brouard 5618:
1.126 brouard 5619: fprintf(ficresstdeij,"%3.0f",age );
5620: for(i=1; i<=nlstate;i++){
5621: eip=0.;
5622: vip=0.;
5623: for(j=1; j<=nlstate;j++){
1.222 brouard 5624: eip += eij[i][j][(int)age];
5625: for(k=1; k<=nlstate;k++) /* Sum on j and k of cov(eij,eik) */
5626: vip += varhe[(j-1)*nlstate+i][(k-1)*nlstate+i][(int)age];
5627: fprintf(ficresstdeij," %9.4f (%.4f)", eij[i][j][(int)age], sqrt(varhe[(j-1)*nlstate+i][(j-1)*nlstate+i][(int)age]) );
1.126 brouard 5628: }
5629: fprintf(ficresstdeij," %9.4f (%.4f)", eip, sqrt(vip));
5630: }
5631: fprintf(ficresstdeij,"\n");
1.218 brouard 5632:
1.126 brouard 5633: fprintf(ficrescveij,"%3.0f",age );
5634: for(i=1; i<=nlstate;i++)
5635: for(j=1; j<=nlstate;j++){
1.222 brouard 5636: cptj= (j-1)*nlstate+i;
5637: for(i2=1; i2<=nlstate;i2++)
5638: for(j2=1; j2<=nlstate;j2++){
5639: cptj2= (j2-1)*nlstate+i2;
5640: if(cptj2 <= cptj)
5641: fprintf(ficrescveij," %.4f", varhe[cptj][cptj2][(int)age]);
5642: }
1.126 brouard 5643: }
5644: fprintf(ficrescveij,"\n");
1.218 brouard 5645:
1.126 brouard 5646: }
5647: free_matrix(gm,0,nhstepm,1,nlstate*nlstate);
5648: free_matrix(gp,0,nhstepm,1,nlstate*nlstate);
5649: free_ma3x(gradg,0,nhstepm,1,npar,1,nlstate*nlstate);
5650: free_ma3x(trgradg,0,nhstepm,1,nlstate*nlstate,1,npar);
5651: free_ma3x(p3matm,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5652: free_ma3x(p3matp,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5653: printf("\n");
5654: fprintf(ficlog,"\n");
1.218 brouard 5655:
1.126 brouard 5656: free_vector(xm,1,npar);
5657: free_vector(xp,1,npar);
5658: free_matrix(dnewm,1,nlstate*nlstate,1,npar);
5659: free_matrix(doldm,1,nlstate*nlstate,1,nlstate*nlstate);
5660: free_ma3x(varhe,1,nlstate*nlstate,1,nlstate*nlstate,(int) bage, (int)fage);
5661: }
1.218 brouard 5662:
1.126 brouard 5663: /************ Variance ******************/
1.235 brouard 5664: void varevsij(char optionfilefiname[], double ***vareij, double **matcov, double x[], double delti[], int nlstate, int stepm, double bage, double fage, double **oldm, double **savm, double **prlim, double ftolpl, int *ncvyearp, int ij, int estepm, int cptcov, int cptcod, int popbased, int mobilav, char strstart[], int nres)
1.218 brouard 5665: {
5666: /* Variance of health expectancies */
5667: /* double **prevalim(double **prlim, int nlstate, double *xp, double age, double **oldm, double ** savm,double ftolpl);*/
5668: /* double **newm;*/
5669: /* int movingaverage(double ***probs, double bage,double fage, double ***mobaverage, int mobilav)*/
5670:
5671: /* int movingaverage(); */
5672: double **dnewm,**doldm;
5673: double **dnewmp,**doldmp;
5674: int i, j, nhstepm, hstepm, h, nstepm ;
5675: int k;
5676: double *xp;
5677: double **gp, **gm; /* for var eij */
5678: double ***gradg, ***trgradg; /*for var eij */
5679: double **gradgp, **trgradgp; /* for var p point j */
5680: double *gpp, *gmp; /* for var p point j */
5681: double **varppt; /* for var p point j nlstate to nlstate+ndeath */
5682: double ***p3mat;
5683: double age,agelim, hf;
5684: /* double ***mobaverage; */
5685: int theta;
5686: char digit[4];
5687: char digitp[25];
5688:
5689: char fileresprobmorprev[FILENAMELENGTH];
5690:
5691: if(popbased==1){
5692: if(mobilav!=0)
5693: strcpy(digitp,"-POPULBASED-MOBILAV_");
5694: else strcpy(digitp,"-POPULBASED-NOMOBIL_");
5695: }
5696: else
5697: strcpy(digitp,"-STABLBASED_");
1.126 brouard 5698:
1.218 brouard 5699: /* if (mobilav!=0) { */
5700: /* mobaverage= ma3x(1, AGESUP,1,NCOVMAX, 1,NCOVMAX); */
5701: /* if (movingaverage(probs, bage, fage, mobaverage,mobilav)!=0){ */
5702: /* fprintf(ficlog," Error in movingaverage mobilav=%d\n",mobilav); */
5703: /* printf(" Error in movingaverage mobilav=%d\n",mobilav); */
5704: /* } */
5705: /* } */
5706:
5707: strcpy(fileresprobmorprev,"PRMORPREV-");
5708: sprintf(digit,"%-d",ij);
5709: /*printf("DIGIT=%s, ij=%d ijr=%-d|\n",digit, ij,ij);*/
5710: strcat(fileresprobmorprev,digit); /* Tvar to be done */
5711: strcat(fileresprobmorprev,digitp); /* Popbased or not, mobilav or not */
5712: strcat(fileresprobmorprev,fileresu);
5713: if((ficresprobmorprev=fopen(fileresprobmorprev,"w"))==NULL) {
5714: printf("Problem with resultfile: %s\n", fileresprobmorprev);
5715: fprintf(ficlog,"Problem with resultfile: %s\n", fileresprobmorprev);
5716: }
5717: printf("Computing total mortality p.j=w1*p1j+w2*p2j+..: result on file '%s' \n",fileresprobmorprev);
5718: fprintf(ficlog,"Computing total mortality p.j=w1*p1j+w2*p2j+..: result on file '%s' \n",fileresprobmorprev);
5719: pstamp(ficresprobmorprev);
5720: fprintf(ficresprobmorprev,"# probabilities of dying before estepm=%d months for people of exact age and weighted probabilities w1*p1j+w2*p2j+... stand dev in()\n",estepm);
1.238 brouard 5721: fprintf(ficresprobmorprev,"# Selected quantitative variables and dummies");
5722: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
5723: fprintf(ficresprobmorprev," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
5724: }
5725: for(j=1;j<=cptcoveff;j++)
5726: fprintf(ficresprobmorprev,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(ij,j)]);
5727: fprintf(ficresprobmorprev,"\n");
5728:
1.218 brouard 5729: fprintf(ficresprobmorprev,"# Age cov=%-d",ij);
5730: for(j=nlstate+1; j<=(nlstate+ndeath);j++){
5731: fprintf(ficresprobmorprev," p.%-d SE",j);
5732: for(i=1; i<=nlstate;i++)
5733: fprintf(ficresprobmorprev," w%1d p%-d%-d",i,i,j);
5734: }
5735: fprintf(ficresprobmorprev,"\n");
5736:
5737: fprintf(ficgp,"\n# Routine varevsij");
5738: fprintf(ficgp,"\nunset title \n");
5739: /* fprintf(fichtm, "#Local time at start: %s", strstart);*/
5740: fprintf(fichtm,"\n<li><h4> Computing probabilities of dying over estepm months as a weighted average (i.e global mortality independent of initial healh state)</h4></li>\n");
5741: fprintf(fichtm,"\n<br>%s <br>\n",digitp);
5742: /* } */
5743: varppt = matrix(nlstate+1,nlstate+ndeath,nlstate+1,nlstate+ndeath);
5744: pstamp(ficresvij);
5745: fprintf(ficresvij,"# Variance and covariance of health expectancies e.j \n# (weighted average of eij where weights are ");
5746: if(popbased==1)
5747: fprintf(ficresvij,"the age specific prevalence observed (cross-sectionally) in the population i.e cross-sectionally\n in each health state (popbased=1) (mobilav=%d\n",mobilav);
5748: else
5749: fprintf(ficresvij,"the age specific period (stable) prevalences in each health state \n");
5750: fprintf(ficresvij,"# Age");
5751: for(i=1; i<=nlstate;i++)
5752: for(j=1; j<=nlstate;j++)
5753: fprintf(ficresvij," Cov(e.%1d, e.%1d)",i,j);
5754: fprintf(ficresvij,"\n");
5755:
5756: xp=vector(1,npar);
5757: dnewm=matrix(1,nlstate,1,npar);
5758: doldm=matrix(1,nlstate,1,nlstate);
5759: dnewmp= matrix(nlstate+1,nlstate+ndeath,1,npar);
5760: doldmp= matrix(nlstate+1,nlstate+ndeath,nlstate+1,nlstate+ndeath);
5761:
5762: gradgp=matrix(1,npar,nlstate+1,nlstate+ndeath);
5763: gpp=vector(nlstate+1,nlstate+ndeath);
5764: gmp=vector(nlstate+1,nlstate+ndeath);
5765: trgradgp =matrix(nlstate+1,nlstate+ndeath,1,npar); /* mu or p point j*/
1.126 brouard 5766:
1.218 brouard 5767: if(estepm < stepm){
5768: printf ("Problem %d lower than %d\n",estepm, stepm);
5769: }
5770: else hstepm=estepm;
5771: /* For example we decided to compute the life expectancy with the smallest unit */
5772: /* hstepm beeing the number of stepms, if hstepm=1 the length of hstepm is stepm.
5773: nhstepm is the number of hstepm from age to agelim
5774: nstepm is the number of stepm from age to agelim.
5775: Look at function hpijx to understand why because of memory size limitations,
5776: we decided (b) to get a life expectancy respecting the most precise curvature of the
5777: survival function given by stepm (the optimization length). Unfortunately it
5778: means that if the survival funtion is printed every two years of age and if
5779: you sum them up and add 1 year (area under the trapezoids) you won't get the same
5780: results. So we changed our mind and took the option of the best precision.
5781: */
5782: hstepm=hstepm/stepm; /* Typically in stepm units, if stepm=6 & estepm=24 , = 24/6 months = 4 */
5783: agelim = AGESUP;
5784: for (age=bage; age<=fage; age ++){ /* If stepm=6 months */
5785: nstepm=(int) rint((agelim-age)*YEARM/stepm); /* Typically 20 years = 20*12/6=40 */
5786: nhstepm = nstepm/hstepm;/* Expressed in hstepm, typically nhstepm=40/4=10 */
5787: p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5788: gradg=ma3x(0,nhstepm,1,npar,1,nlstate);
5789: gp=matrix(0,nhstepm,1,nlstate);
5790: gm=matrix(0,nhstepm,1,nlstate);
5791:
5792:
5793: for(theta=1; theta <=npar; theta++){
5794: for(i=1; i<=npar; i++){ /* Computes gradient x + delta*/
5795: xp[i] = x[i] + (i==theta ?delti[theta]:0);
5796: }
5797:
1.242 brouard 5798: prevalim(prlim,nlstate,xp,age,oldm,savm,ftolpl,ncvyearp,ij, nres);
1.218 brouard 5799:
5800: if (popbased==1) {
5801: if(mobilav ==0){
5802: for(i=1; i<=nlstate;i++)
5803: prlim[i][i]=probs[(int)age][i][ij];
5804: }else{ /* mobilav */
5805: for(i=1; i<=nlstate;i++)
5806: prlim[i][i]=mobaverage[(int)age][i][ij];
5807: }
5808: }
5809:
1.235 brouard 5810: hpxij(p3mat,nhstepm,age,hstepm,xp,nlstate,stepm,oldm,savm, ij,nres); /* Returns p3mat[i][j][h] for h=1 to nhstepm */
1.218 brouard 5811: for(j=1; j<= nlstate; j++){
5812: for(h=0; h<=nhstepm; h++){
5813: for(i=1, gp[h][j]=0.;i<=nlstate;i++)
5814: gp[h][j] += prlim[i][i]*p3mat[i][j][h];
5815: }
5816: }
5817: /* Next for computing probability of death (h=1 means
5818: computed over hstepm matrices product = hstepm*stepm months)
5819: as a weighted average of prlim.
5820: */
5821: for(j=nlstate+1;j<=nlstate+ndeath;j++){
5822: for(i=1,gpp[j]=0.; i<= nlstate; i++)
5823: gpp[j] += prlim[i][i]*p3mat[i][j][1];
5824: }
5825: /* end probability of death */
5826:
5827: for(i=1; i<=npar; i++) /* Computes gradient x - delta */
5828: xp[i] = x[i] - (i==theta ?delti[theta]:0);
5829:
1.242 brouard 5830: prevalim(prlim,nlstate,xp,age,oldm,savm,ftolpl,ncvyearp, ij, nres);
1.218 brouard 5831:
5832: if (popbased==1) {
5833: if(mobilav ==0){
5834: for(i=1; i<=nlstate;i++)
5835: prlim[i][i]=probs[(int)age][i][ij];
5836: }else{ /* mobilav */
5837: for(i=1; i<=nlstate;i++)
5838: prlim[i][i]=mobaverage[(int)age][i][ij];
5839: }
5840: }
5841:
1.235 brouard 5842: hpxij(p3mat,nhstepm,age,hstepm,xp,nlstate,stepm,oldm,savm, ij,nres);
1.218 brouard 5843:
5844: for(j=1; j<= nlstate; j++){ /* Sum of wi * eij = e.j */
5845: for(h=0; h<=nhstepm; h++){
5846: for(i=1, gm[h][j]=0.;i<=nlstate;i++)
5847: gm[h][j] += prlim[i][i]*p3mat[i][j][h];
5848: }
5849: }
5850: /* This for computing probability of death (h=1 means
5851: computed over hstepm matrices product = hstepm*stepm months)
5852: as a weighted average of prlim.
5853: */
5854: for(j=nlstate+1;j<=nlstate+ndeath;j++){
5855: for(i=1,gmp[j]=0.; i<= nlstate; i++)
5856: gmp[j] += prlim[i][i]*p3mat[i][j][1];
5857: }
5858: /* end probability of death */
5859:
5860: for(j=1; j<= nlstate; j++) /* vareij */
5861: for(h=0; h<=nhstepm; h++){
5862: gradg[h][theta][j]= (gp[h][j]-gm[h][j])/2./delti[theta];
5863: }
5864:
5865: for(j=nlstate+1; j<= nlstate+ndeath; j++){ /* var mu */
5866: gradgp[theta][j]= (gpp[j]-gmp[j])/2./delti[theta];
5867: }
5868:
5869: } /* End theta */
5870:
5871: trgradg =ma3x(0,nhstepm,1,nlstate,1,npar); /* veij */
5872:
5873: for(h=0; h<=nhstepm; h++) /* veij */
5874: for(j=1; j<=nlstate;j++)
5875: for(theta=1; theta <=npar; theta++)
5876: trgradg[h][j][theta]=gradg[h][theta][j];
5877:
5878: for(j=nlstate+1; j<=nlstate+ndeath;j++) /* mu */
5879: for(theta=1; theta <=npar; theta++)
5880: trgradgp[j][theta]=gradgp[theta][j];
5881:
5882:
5883: hf=hstepm*stepm/YEARM; /* Duration of hstepm expressed in year unit. */
5884: for(i=1;i<=nlstate;i++)
5885: for(j=1;j<=nlstate;j++)
5886: vareij[i][j][(int)age] =0.;
5887:
5888: for(h=0;h<=nhstepm;h++){
5889: for(k=0;k<=nhstepm;k++){
5890: matprod2(dnewm,trgradg[h],1,nlstate,1,npar,1,npar,matcov);
5891: matprod2(doldm,dnewm,1,nlstate,1,npar,1,nlstate,gradg[k]);
5892: for(i=1;i<=nlstate;i++)
5893: for(j=1;j<=nlstate;j++)
5894: vareij[i][j][(int)age] += doldm[i][j]*hf*hf;
5895: }
5896: }
5897:
5898: /* pptj */
5899: matprod2(dnewmp,trgradgp,nlstate+1,nlstate+ndeath,1,npar,1,npar,matcov);
5900: matprod2(doldmp,dnewmp,nlstate+1,nlstate+ndeath,1,npar,nlstate+1,nlstate+ndeath,gradgp);
5901: for(j=nlstate+1;j<=nlstate+ndeath;j++)
5902: for(i=nlstate+1;i<=nlstate+ndeath;i++)
5903: varppt[j][i]=doldmp[j][i];
5904: /* end ppptj */
5905: /* x centered again */
5906:
1.242 brouard 5907: prevalim(prlim,nlstate,x,age,oldm,savm,ftolpl,ncvyearp,ij, nres);
1.218 brouard 5908:
5909: if (popbased==1) {
5910: if(mobilav ==0){
5911: for(i=1; i<=nlstate;i++)
5912: prlim[i][i]=probs[(int)age][i][ij];
5913: }else{ /* mobilav */
5914: for(i=1; i<=nlstate;i++)
5915: prlim[i][i]=mobaverage[(int)age][i][ij];
5916: }
5917: }
5918:
5919: /* This for computing probability of death (h=1 means
5920: computed over hstepm (estepm) matrices product = hstepm*stepm months)
5921: as a weighted average of prlim.
5922: */
1.235 brouard 5923: hpxij(p3mat,nhstepm,age,hstepm,x,nlstate,stepm,oldm,savm, ij, nres);
1.218 brouard 5924: for(j=nlstate+1;j<=nlstate+ndeath;j++){
5925: for(i=1,gmp[j]=0.;i<= nlstate; i++)
5926: gmp[j] += prlim[i][i]*p3mat[i][j][1];
5927: }
5928: /* end probability of death */
5929:
5930: fprintf(ficresprobmorprev,"%3d %d ",(int) age, ij);
5931: for(j=nlstate+1; j<=(nlstate+ndeath);j++){
5932: fprintf(ficresprobmorprev," %11.3e %11.3e",gmp[j], sqrt(varppt[j][j]));
5933: for(i=1; i<=nlstate;i++){
5934: fprintf(ficresprobmorprev," %11.3e %11.3e ",prlim[i][i],p3mat[i][j][1]);
5935: }
5936: }
5937: fprintf(ficresprobmorprev,"\n");
5938:
5939: fprintf(ficresvij,"%.0f ",age );
5940: for(i=1; i<=nlstate;i++)
5941: for(j=1; j<=nlstate;j++){
5942: fprintf(ficresvij," %.4f", vareij[i][j][(int)age]);
5943: }
5944: fprintf(ficresvij,"\n");
5945: free_matrix(gp,0,nhstepm,1,nlstate);
5946: free_matrix(gm,0,nhstepm,1,nlstate);
5947: free_ma3x(gradg,0,nhstepm,1,npar,1,nlstate);
5948: free_ma3x(trgradg,0,nhstepm,1,nlstate,1,npar);
5949: free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5950: } /* End age */
5951: free_vector(gpp,nlstate+1,nlstate+ndeath);
5952: free_vector(gmp,nlstate+1,nlstate+ndeath);
5953: free_matrix(gradgp,1,npar,nlstate+1,nlstate+ndeath);
5954: free_matrix(trgradgp,nlstate+1,nlstate+ndeath,1,npar); /* mu or p point j*/
5955: /* fprintf(ficgp,"\nunset parametric;unset label; set ter png small size 320, 240"); */
5956: fprintf(ficgp,"\nunset parametric;unset label; set ter svg size 640, 480");
5957: /* for(j=nlstate+1; j<= nlstate+ndeath; j++){ *//* Only the first actually */
5958: fprintf(ficgp,"\n set log y; unset log x;set xlabel \"Age\"; set ylabel \"Force of mortality (year-1)\";");
5959: fprintf(ficgp,"\nset out \"%s%s.svg\";",subdirf3(optionfilefiname,"VARMUPTJGR-",digitp),digit);
5960: /* fprintf(ficgp,"\n plot \"%s\" u 1:($3*%6.3f) not w l 1 ",fileresprobmorprev,YEARM/estepm); */
5961: /* fprintf(ficgp,"\n replot \"%s\" u 1:(($3+1.96*$4)*%6.3f) t \"95\%% interval\" w l 2 ",fileresprobmorprev,YEARM/estepm); */
5962: /* fprintf(ficgp,"\n replot \"%s\" u 1:(($3-1.96*$4)*%6.3f) not w l 2 ",fileresprobmorprev,YEARM/estepm); */
5963: fprintf(ficgp,"\n plot \"%s\" u 1:($3) not w l lt 1 ",subdirf(fileresprobmorprev));
5964: fprintf(ficgp,"\n replot \"%s\" u 1:(($3+1.96*$4)) t \"95%% interval\" w l lt 2 ",subdirf(fileresprobmorprev));
5965: fprintf(ficgp,"\n replot \"%s\" u 1:(($3-1.96*$4)) not w l lt 2 ",subdirf(fileresprobmorprev));
5966: fprintf(fichtm,"\n<br> File (multiple files are possible if covariates are present): <A href=\"%s\">%s</a>\n",subdirf(fileresprobmorprev),subdirf(fileresprobmorprev));
5967: fprintf(fichtm,"\n<br> Probability is computed over estepm=%d months. <br> <img src=\"%s%s.svg\"> <br>\n", estepm,subdirf3(optionfilefiname,"VARMUPTJGR-",digitp),digit);
5968: /* fprintf(fichtm,"\n<br> Probability is computed over estepm=%d months and then divided by estepm and multiplied by %.0f in order to have the probability to die over a year <br> <img src=\"varmuptjgr%s%s.svg\"> <br>\n", stepm,YEARM,digitp,digit);
1.126 brouard 5969: */
1.218 brouard 5970: /* fprintf(ficgp,"\nset out \"varmuptjgr%s%s%s.svg\";replot;",digitp,optionfilefiname,digit); */
5971: fprintf(ficgp,"\nset out;\nset out \"%s%s.svg\";replot;set out;\n",subdirf3(optionfilefiname,"VARMUPTJGR-",digitp),digit);
1.126 brouard 5972:
1.218 brouard 5973: free_vector(xp,1,npar);
5974: free_matrix(doldm,1,nlstate,1,nlstate);
5975: free_matrix(dnewm,1,nlstate,1,npar);
5976: free_matrix(doldmp,nlstate+1,nlstate+ndeath,nlstate+1,nlstate+ndeath);
5977: free_matrix(dnewmp,nlstate+1,nlstate+ndeath,1,npar);
5978: free_matrix(varppt,nlstate+1,nlstate+ndeath,nlstate+1,nlstate+ndeath);
5979: /* if (mobilav!=0) free_ma3x(mobaverage,1, AGESUP,1,NCOVMAX, 1,NCOVMAX); */
5980: fclose(ficresprobmorprev);
5981: fflush(ficgp);
5982: fflush(fichtm);
5983: } /* end varevsij */
1.126 brouard 5984:
5985: /************ Variance of prevlim ******************/
1.235 brouard 5986: void varprevlim(char fileres[], double **varpl, double **matcov, double x[], double delti[], int nlstate, int stepm, double bage, double fage, double **oldm, double **savm, double **prlim, double ftolpl, int *ncvyearp, int ij, char strstart[], int nres)
1.126 brouard 5987: {
1.205 brouard 5988: /* Variance of prevalence limit for each state ij using current parameters x[] and estimates of neighbourhood give by delti*/
1.126 brouard 5989: /* double **prevalim(double **prlim, int nlstate, double *xp, double age, double **oldm, double **savm,double ftolpl);*/
1.164 brouard 5990:
1.126 brouard 5991: double **dnewm,**doldm;
5992: int i, j, nhstepm, hstepm;
5993: double *xp;
5994: double *gp, *gm;
5995: double **gradg, **trgradg;
1.208 brouard 5996: double **mgm, **mgp;
1.126 brouard 5997: double age,agelim;
5998: int theta;
5999:
6000: pstamp(ficresvpl);
6001: fprintf(ficresvpl,"# Standard deviation of period (stable) prevalences \n");
1.241 brouard 6002: fprintf(ficresvpl,"# Age ");
6003: if(nresult >=1)
6004: fprintf(ficresvpl," Result# ");
1.126 brouard 6005: for(i=1; i<=nlstate;i++)
6006: fprintf(ficresvpl," %1d-%1d",i,i);
6007: fprintf(ficresvpl,"\n");
6008:
6009: xp=vector(1,npar);
6010: dnewm=matrix(1,nlstate,1,npar);
6011: doldm=matrix(1,nlstate,1,nlstate);
6012:
6013: hstepm=1*YEARM; /* Every year of age */
6014: hstepm=hstepm/stepm; /* Typically in stepm units, if j= 2 years, = 2/6 months = 4 */
6015: agelim = AGESUP;
6016: for (age=bage; age<=fage; age ++){ /* If stepm=6 months */
6017: nhstepm=(int) rint((agelim-age)*YEARM/stepm); /* Typically 20 years = 20*12/6=40 */
6018: if (stepm >= YEARM) hstepm=1;
6019: nhstepm = nhstepm/hstepm; /* Typically 40/4=10 */
6020: gradg=matrix(1,npar,1,nlstate);
1.208 brouard 6021: mgp=matrix(1,npar,1,nlstate);
6022: mgm=matrix(1,npar,1,nlstate);
1.126 brouard 6023: gp=vector(1,nlstate);
6024: gm=vector(1,nlstate);
6025:
6026: for(theta=1; theta <=npar; theta++){
6027: for(i=1; i<=npar; i++){ /* Computes gradient */
6028: xp[i] = x[i] + (i==theta ?delti[theta]:0);
6029: }
1.209 brouard 6030: if((int)age==79 ||(int)age== 80 ||(int)age== 81 )
1.235 brouard 6031: prevalim(prlim,nlstate,xp,age,oldm,savm,ftolpl,ncvyearp,ij,nres);
1.209 brouard 6032: else
1.235 brouard 6033: prevalim(prlim,nlstate,xp,age,oldm,savm,ftolpl,ncvyearp,ij,nres);
1.208 brouard 6034: for(i=1;i<=nlstate;i++){
1.126 brouard 6035: gp[i] = prlim[i][i];
1.208 brouard 6036: mgp[theta][i] = prlim[i][i];
6037: }
1.126 brouard 6038: for(i=1; i<=npar; i++) /* Computes gradient */
6039: xp[i] = x[i] - (i==theta ?delti[theta]:0);
1.209 brouard 6040: if((int)age==79 ||(int)age== 80 ||(int)age== 81 )
1.235 brouard 6041: prevalim(prlim,nlstate,xp,age,oldm,savm,ftolpl,ncvyearp,ij,nres);
1.209 brouard 6042: else
1.235 brouard 6043: prevalim(prlim,nlstate,xp,age,oldm,savm,ftolpl,ncvyearp,ij,nres);
1.208 brouard 6044: for(i=1;i<=nlstate;i++){
1.126 brouard 6045: gm[i] = prlim[i][i];
1.208 brouard 6046: mgm[theta][i] = prlim[i][i];
6047: }
1.126 brouard 6048: for(i=1;i<=nlstate;i++)
6049: gradg[theta][i]= (gp[i]-gm[i])/2./delti[theta];
1.209 brouard 6050: /* gradg[theta][2]= -gradg[theta][1]; */ /* For testing if nlstate=2 */
1.126 brouard 6051: } /* End theta */
6052:
6053: trgradg =matrix(1,nlstate,1,npar);
6054:
6055: for(j=1; j<=nlstate;j++)
6056: for(theta=1; theta <=npar; theta++)
6057: trgradg[j][theta]=gradg[theta][j];
1.209 brouard 6058: /* if((int)age==79 ||(int)age== 80 ||(int)age== 81 ){ */
6059: /* printf("\nmgm mgp %d ",(int)age); */
6060: /* for(j=1; j<=nlstate;j++){ */
6061: /* printf(" %d ",j); */
6062: /* for(theta=1; theta <=npar; theta++) */
6063: /* printf(" %d %lf %lf",theta,mgm[theta][j],mgp[theta][j]); */
6064: /* printf("\n "); */
6065: /* } */
6066: /* } */
6067: /* if((int)age==79 ||(int)age== 80 ||(int)age== 81 ){ */
6068: /* printf("\n gradg %d ",(int)age); */
6069: /* for(j=1; j<=nlstate;j++){ */
6070: /* printf("%d ",j); */
6071: /* for(theta=1; theta <=npar; theta++) */
6072: /* printf("%d %lf ",theta,gradg[theta][j]); */
6073: /* printf("\n "); */
6074: /* } */
6075: /* } */
1.126 brouard 6076:
6077: for(i=1;i<=nlstate;i++)
6078: varpl[i][(int)age] =0.;
1.209 brouard 6079: if((int)age==79 ||(int)age== 80 ||(int)age== 81){
1.205 brouard 6080: matprod2(dnewm,trgradg,1,nlstate,1,npar,1,npar,matcov);
6081: matprod2(doldm,dnewm,1,nlstate,1,npar,1,nlstate,gradg);
6082: }else{
1.126 brouard 6083: matprod2(dnewm,trgradg,1,nlstate,1,npar,1,npar,matcov);
6084: matprod2(doldm,dnewm,1,nlstate,1,npar,1,nlstate,gradg);
1.205 brouard 6085: }
1.126 brouard 6086: for(i=1;i<=nlstate;i++)
6087: varpl[i][(int)age] = doldm[i][i]; /* Covariances are useless */
6088:
6089: fprintf(ficresvpl,"%.0f ",age );
1.241 brouard 6090: if(nresult >=1)
6091: fprintf(ficresvpl,"%d ",nres );
1.126 brouard 6092: for(i=1; i<=nlstate;i++)
6093: fprintf(ficresvpl," %.5f (%.5f)",prlim[i][i],sqrt(varpl[i][(int)age]));
6094: fprintf(ficresvpl,"\n");
6095: free_vector(gp,1,nlstate);
6096: free_vector(gm,1,nlstate);
1.208 brouard 6097: free_matrix(mgm,1,npar,1,nlstate);
6098: free_matrix(mgp,1,npar,1,nlstate);
1.126 brouard 6099: free_matrix(gradg,1,npar,1,nlstate);
6100: free_matrix(trgradg,1,nlstate,1,npar);
6101: } /* End age */
6102:
6103: free_vector(xp,1,npar);
6104: free_matrix(doldm,1,nlstate,1,npar);
6105: free_matrix(dnewm,1,nlstate,1,nlstate);
6106:
6107: }
6108:
6109: /************ Variance of one-step probabilities ******************/
6110: void varprob(char optionfilefiname[], double **matcov, double x[], double delti[], int nlstate, double bage, double fage, int ij, int *Tvar, int **nbcode, int *ncodemax, char strstart[])
1.222 brouard 6111: {
6112: int i, j=0, k1, l1, tj;
6113: int k2, l2, j1, z1;
6114: int k=0, l;
6115: int first=1, first1, first2;
6116: double cv12, mu1, mu2, lc1, lc2, v12, v21, v11, v22,v1,v2, c12, tnalp;
6117: double **dnewm,**doldm;
6118: double *xp;
6119: double *gp, *gm;
6120: double **gradg, **trgradg;
6121: double **mu;
6122: double age, cov[NCOVMAX+1];
6123: double std=2.0; /* Number of standard deviation wide of confidence ellipsoids */
6124: int theta;
6125: char fileresprob[FILENAMELENGTH];
6126: char fileresprobcov[FILENAMELENGTH];
6127: char fileresprobcor[FILENAMELENGTH];
6128: double ***varpij;
6129:
6130: strcpy(fileresprob,"PROB_");
6131: strcat(fileresprob,fileres);
6132: if((ficresprob=fopen(fileresprob,"w"))==NULL) {
6133: printf("Problem with resultfile: %s\n", fileresprob);
6134: fprintf(ficlog,"Problem with resultfile: %s\n", fileresprob);
6135: }
6136: strcpy(fileresprobcov,"PROBCOV_");
6137: strcat(fileresprobcov,fileresu);
6138: if((ficresprobcov=fopen(fileresprobcov,"w"))==NULL) {
6139: printf("Problem with resultfile: %s\n", fileresprobcov);
6140: fprintf(ficlog,"Problem with resultfile: %s\n", fileresprobcov);
6141: }
6142: strcpy(fileresprobcor,"PROBCOR_");
6143: strcat(fileresprobcor,fileresu);
6144: if((ficresprobcor=fopen(fileresprobcor,"w"))==NULL) {
6145: printf("Problem with resultfile: %s\n", fileresprobcor);
6146: fprintf(ficlog,"Problem with resultfile: %s\n", fileresprobcor);
6147: }
6148: printf("Computing standard deviation of one-step probabilities: result on file '%s' \n",fileresprob);
6149: fprintf(ficlog,"Computing standard deviation of one-step probabilities: result on file '%s' \n",fileresprob);
6150: printf("Computing matrix of variance covariance of one-step probabilities: result on file '%s' \n",fileresprobcov);
6151: fprintf(ficlog,"Computing matrix of variance covariance of one-step probabilities: result on file '%s' \n",fileresprobcov);
6152: printf("and correlation matrix of one-step probabilities: result on file '%s' \n",fileresprobcor);
6153: fprintf(ficlog,"and correlation matrix of one-step probabilities: result on file '%s' \n",fileresprobcor);
6154: pstamp(ficresprob);
6155: fprintf(ficresprob,"#One-step probabilities and stand. devi in ()\n");
6156: fprintf(ficresprob,"# Age");
6157: pstamp(ficresprobcov);
6158: fprintf(ficresprobcov,"#One-step probabilities and covariance matrix\n");
6159: fprintf(ficresprobcov,"# Age");
6160: pstamp(ficresprobcor);
6161: fprintf(ficresprobcor,"#One-step probabilities and correlation matrix\n");
6162: fprintf(ficresprobcor,"# Age");
1.126 brouard 6163:
6164:
1.222 brouard 6165: for(i=1; i<=nlstate;i++)
6166: for(j=1; j<=(nlstate+ndeath);j++){
6167: fprintf(ficresprob," p%1d-%1d (SE)",i,j);
6168: fprintf(ficresprobcov," p%1d-%1d ",i,j);
6169: fprintf(ficresprobcor," p%1d-%1d ",i,j);
6170: }
6171: /* fprintf(ficresprob,"\n");
6172: fprintf(ficresprobcov,"\n");
6173: fprintf(ficresprobcor,"\n");
6174: */
6175: xp=vector(1,npar);
6176: dnewm=matrix(1,(nlstate)*(nlstate+ndeath),1,npar);
6177: doldm=matrix(1,(nlstate)*(nlstate+ndeath),1,(nlstate)*(nlstate+ndeath));
6178: mu=matrix(1,(nlstate)*(nlstate+ndeath), (int) bage, (int)fage);
6179: varpij=ma3x(1,nlstate*(nlstate+ndeath),1,nlstate*(nlstate+ndeath),(int) bage, (int) fage);
6180: first=1;
6181: fprintf(ficgp,"\n# Routine varprob");
6182: fprintf(fichtm,"\n<li><h4> Computing and drawing one step probabilities with their confidence intervals</h4></li>\n");
6183: fprintf(fichtm,"\n");
6184:
6185: fprintf(fichtm,"\n<li><h4> <a href=\"%s\">Matrix of variance-covariance of one-step probabilities (drawings)</a></h4> this page is important in order to visualize confidence intervals and especially correlation between disability and recovery, or more generally, way in and way back.</li>\n",optionfilehtmcov);
6186: fprintf(fichtmcov,"Current page is file <a href=\"%s\">%s</a><br>\n\n<h4>Matrix of variance-covariance of pairs of step probabilities</h4>\n",optionfilehtmcov, optionfilehtmcov);
6187: fprintf(fichtmcov,"\nEllipsoids of confidence centered on point (p<inf>ij</inf>, p<inf>kl</inf>) are estimated \
1.126 brouard 6188: and drawn. It helps understanding how is the covariance between two incidences.\
6189: They are expressed in year<sup>-1</sup> in order to be less dependent of stepm.<br>\n");
1.222 brouard 6190: fprintf(fichtmcov,"\n<br> Contour plot corresponding to x'cov<sup>-1</sup>x = 4 (where x is the column vector (pij,pkl)) are drawn. \
1.126 brouard 6191: It can be understood this way: if pij and pkl where uncorrelated the (2x2) matrix of covariance \
6192: would have been (1/(var pij), 0 , 0, 1/(var pkl)), and the confidence interval would be 2 \
6193: standard deviations wide on each axis. <br>\
6194: Now, if both incidences are correlated (usual case) we diagonalised the inverse of the covariance matrix\
6195: and made the appropriate rotation to look at the uncorrelated principal directions.<br>\
6196: To be simple, these graphs help to understand the significativity of each parameter in relation to a second other one.<br> \n");
6197:
1.222 brouard 6198: cov[1]=1;
6199: /* tj=cptcoveff; */
1.225 brouard 6200: tj = (int) pow(2,cptcoveff);
1.222 brouard 6201: if (cptcovn<1) {tj=1;ncodemax[1]=1;}
6202: j1=0;
1.224 brouard 6203: for(j1=1; j1<=tj;j1++){ /* For each valid combination of covariates or only once*/
1.222 brouard 6204: if (cptcovn>0) {
6205: fprintf(ficresprob, "\n#********** Variable ");
1.225 brouard 6206: for (z1=1; z1<=cptcoveff; z1++) fprintf(ficresprob, "V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
1.222 brouard 6207: fprintf(ficresprob, "**********\n#\n");
6208: fprintf(ficresprobcov, "\n#********** Variable ");
1.225 brouard 6209: for (z1=1; z1<=cptcoveff; z1++) fprintf(ficresprobcov, "V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
1.222 brouard 6210: fprintf(ficresprobcov, "**********\n#\n");
1.220 brouard 6211:
1.222 brouard 6212: fprintf(ficgp, "\n#********** Variable ");
1.225 brouard 6213: for (z1=1; z1<=cptcoveff; z1++) fprintf(ficgp, " V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
1.222 brouard 6214: fprintf(ficgp, "**********\n#\n");
1.220 brouard 6215:
6216:
1.222 brouard 6217: fprintf(fichtmcov, "\n<hr size=\"2\" color=\"#EC5E5E\">********** Variable ");
1.225 brouard 6218: for (z1=1; z1<=cptcoveff; z1++) fprintf(fichtm, "V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
1.222 brouard 6219: fprintf(fichtmcov, "**********\n<hr size=\"2\" color=\"#EC5E5E\">");
1.220 brouard 6220:
1.222 brouard 6221: fprintf(ficresprobcor, "\n#********** Variable ");
1.225 brouard 6222: for (z1=1; z1<=cptcoveff; z1++) fprintf(ficresprobcor, "V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
1.222 brouard 6223: fprintf(ficresprobcor, "**********\n#");
6224: if(invalidvarcomb[j1]){
6225: fprintf(ficgp,"\n#Combination (%d) ignored because no cases \n",j1);
6226: fprintf(fichtmcov,"\n<h3>Combination (%d) ignored because no cases </h3>\n",j1);
6227: continue;
6228: }
6229: }
6230: gradg=matrix(1,npar,1,(nlstate)*(nlstate+ndeath));
6231: trgradg=matrix(1,(nlstate)*(nlstate+ndeath),1,npar);
6232: gp=vector(1,(nlstate)*(nlstate+ndeath));
6233: gm=vector(1,(nlstate)*(nlstate+ndeath));
6234: for (age=bage; age<=fage; age ++){
6235: cov[2]=age;
6236: if(nagesqr==1)
6237: cov[3]= age*age;
6238: for (k=1; k<=cptcovn;k++) {
6239: cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(j1,k)];
6240: /*cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(j1,Tvar[k])];*//* j1 1 2 3 4
6241: * 1 1 1 1 1
6242: * 2 2 1 1 1
6243: * 3 1 2 1 1
6244: */
6245: /* nbcode[1][1]=0 nbcode[1][2]=1;*/
6246: }
6247: /* for (k=1; k<=cptcovage;k++) cov[2+Tage[k]]=cov[2+Tage[k]]*cov[2]; */
6248: for (k=1; k<=cptcovage;k++) cov[2+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2];
6249: for (k=1; k<=cptcovprod;k++)
6250: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)]*nbcode[Tvard[k][2]][codtabm(ij,k)];
1.220 brouard 6251:
6252:
1.222 brouard 6253: for(theta=1; theta <=npar; theta++){
6254: for(i=1; i<=npar; i++)
6255: xp[i] = x[i] + (i==theta ?delti[theta]:(double)0);
1.220 brouard 6256:
1.222 brouard 6257: pmij(pmmij,cov,ncovmodel,xp,nlstate);
1.220 brouard 6258:
1.222 brouard 6259: k=0;
6260: for(i=1; i<= (nlstate); i++){
6261: for(j=1; j<=(nlstate+ndeath);j++){
6262: k=k+1;
6263: gp[k]=pmmij[i][j];
6264: }
6265: }
1.220 brouard 6266:
1.222 brouard 6267: for(i=1; i<=npar; i++)
6268: xp[i] = x[i] - (i==theta ?delti[theta]:(double)0);
1.220 brouard 6269:
1.222 brouard 6270: pmij(pmmij,cov,ncovmodel,xp,nlstate);
6271: k=0;
6272: for(i=1; i<=(nlstate); i++){
6273: for(j=1; j<=(nlstate+ndeath);j++){
6274: k=k+1;
6275: gm[k]=pmmij[i][j];
6276: }
6277: }
1.220 brouard 6278:
1.222 brouard 6279: for(i=1; i<= (nlstate)*(nlstate+ndeath); i++)
6280: gradg[theta][i]=(gp[i]-gm[i])/(double)2./delti[theta];
6281: }
1.126 brouard 6282:
1.222 brouard 6283: for(j=1; j<=(nlstate)*(nlstate+ndeath);j++)
6284: for(theta=1; theta <=npar; theta++)
6285: trgradg[j][theta]=gradg[theta][j];
1.220 brouard 6286:
1.222 brouard 6287: matprod2(dnewm,trgradg,1,(nlstate)*(nlstate+ndeath),1,npar,1,npar,matcov);
6288: matprod2(doldm,dnewm,1,(nlstate)*(nlstate+ndeath),1,npar,1,(nlstate)*(nlstate+ndeath),gradg);
1.220 brouard 6289:
1.222 brouard 6290: pmij(pmmij,cov,ncovmodel,x,nlstate);
1.220 brouard 6291:
1.222 brouard 6292: k=0;
6293: for(i=1; i<=(nlstate); i++){
6294: for(j=1; j<=(nlstate+ndeath);j++){
6295: k=k+1;
6296: mu[k][(int) age]=pmmij[i][j];
6297: }
6298: }
6299: for(i=1;i<=(nlstate)*(nlstate+ndeath);i++)
6300: for(j=1;j<=(nlstate)*(nlstate+ndeath);j++)
6301: varpij[i][j][(int)age] = doldm[i][j];
1.220 brouard 6302:
1.222 brouard 6303: /*printf("\n%d ",(int)age);
6304: for (i=1; i<=(nlstate)*(nlstate+ndeath);i++){
6305: printf("%e [%e ;%e] ",gm[i],gm[i]-2*sqrt(doldm[i][i]),gm[i]+2*sqrt(doldm[i][i]));
6306: fprintf(ficlog,"%e [%e ;%e] ",gm[i],gm[i]-2*sqrt(doldm[i][i]),gm[i]+2*sqrt(doldm[i][i]));
6307: }*/
1.220 brouard 6308:
1.222 brouard 6309: fprintf(ficresprob,"\n%d ",(int)age);
6310: fprintf(ficresprobcov,"\n%d ",(int)age);
6311: fprintf(ficresprobcor,"\n%d ",(int)age);
1.220 brouard 6312:
1.222 brouard 6313: for (i=1; i<=(nlstate)*(nlstate+ndeath);i++)
6314: fprintf(ficresprob,"%11.3e (%11.3e) ",mu[i][(int) age],sqrt(varpij[i][i][(int)age]));
6315: for (i=1; i<=(nlstate)*(nlstate+ndeath);i++){
6316: fprintf(ficresprobcov,"%11.3e ",mu[i][(int) age]);
6317: fprintf(ficresprobcor,"%11.3e ",mu[i][(int) age]);
6318: }
6319: i=0;
6320: for (k=1; k<=(nlstate);k++){
6321: for (l=1; l<=(nlstate+ndeath);l++){
6322: i++;
6323: fprintf(ficresprobcov,"\n%d %d-%d",(int)age,k,l);
6324: fprintf(ficresprobcor,"\n%d %d-%d",(int)age,k,l);
6325: for (j=1; j<=i;j++){
6326: /* printf(" k=%d l=%d i=%d j=%d\n",k,l,i,j);fflush(stdout); */
6327: fprintf(ficresprobcov," %11.3e",varpij[i][j][(int)age]);
6328: fprintf(ficresprobcor," %11.3e",varpij[i][j][(int) age]/sqrt(varpij[i][i][(int) age])/sqrt(varpij[j][j][(int)age]));
6329: }
6330: }
6331: }/* end of loop for state */
6332: } /* end of loop for age */
6333: free_vector(gp,1,(nlstate+ndeath)*(nlstate+ndeath));
6334: free_vector(gm,1,(nlstate+ndeath)*(nlstate+ndeath));
6335: free_matrix(trgradg,1,(nlstate+ndeath)*(nlstate+ndeath),1,npar);
6336: free_matrix(gradg,1,(nlstate+ndeath)*(nlstate+ndeath),1,npar);
6337:
6338: /* Confidence intervalle of pij */
6339: /*
6340: fprintf(ficgp,"\nunset parametric;unset label");
6341: fprintf(ficgp,"\nset log y;unset log x; set xlabel \"Age\";set ylabel \"probability (year-1)\"");
6342: fprintf(ficgp,"\nset ter png small\nset size 0.65,0.65");
6343: fprintf(fichtm,"\n<br>Probability with confidence intervals expressed in year<sup>-1</sup> :<a href=\"pijgr%s.png\">pijgr%s.png</A>, ",optionfilefiname,optionfilefiname);
6344: fprintf(fichtm,"\n<br><img src=\"pijgr%s.png\"> ",optionfilefiname);
6345: fprintf(ficgp,"\nset out \"pijgr%s.png\"",optionfilefiname);
6346: fprintf(ficgp,"\nplot \"%s\" every :::%d::%d u 1:2 \"\%%lf",k1,k2,xfilevarprob);
6347: */
6348:
6349: /* Drawing ellipsoids of confidence of two variables p(k1-l1,k2-l2)*/
6350: first1=1;first2=2;
6351: for (k2=1; k2<=(nlstate);k2++){
6352: for (l2=1; l2<=(nlstate+ndeath);l2++){
6353: if(l2==k2) continue;
6354: j=(k2-1)*(nlstate+ndeath)+l2;
6355: for (k1=1; k1<=(nlstate);k1++){
6356: for (l1=1; l1<=(nlstate+ndeath);l1++){
6357: if(l1==k1) continue;
6358: i=(k1-1)*(nlstate+ndeath)+l1;
6359: if(i<=j) continue;
6360: for (age=bage; age<=fage; age ++){
6361: if ((int)age %5==0){
6362: v1=varpij[i][i][(int)age]/stepm*YEARM/stepm*YEARM;
6363: v2=varpij[j][j][(int)age]/stepm*YEARM/stepm*YEARM;
6364: cv12=varpij[i][j][(int)age]/stepm*YEARM/stepm*YEARM;
6365: mu1=mu[i][(int) age]/stepm*YEARM ;
6366: mu2=mu[j][(int) age]/stepm*YEARM;
6367: c12=cv12/sqrt(v1*v2);
6368: /* Computing eigen value of matrix of covariance */
6369: lc1=((v1+v2)+sqrt((v1+v2)*(v1+v2) - 4*(v1*v2-cv12*cv12)))/2.;
6370: lc2=((v1+v2)-sqrt((v1+v2)*(v1+v2) - 4*(v1*v2-cv12*cv12)))/2.;
6371: if ((lc2 <0) || (lc1 <0) ){
6372: if(first2==1){
6373: first1=0;
6374: printf("Strange: j1=%d One eigen value of 2x2 matrix of covariance is negative, lc1=%11.3e, lc2=%11.3e, v1=%11.3e, v2=%11.3e, cv12=%11.3e.\n It means that the matrix was not well estimated (varpij), for i=%2d, j=%2d, age=%4d .\n See files %s and %s. Probably WRONG RESULTS. See log file for details...\n", j1, lc1, lc2, v1, v2, cv12, i, j, (int)age,fileresprobcov, fileresprobcor);
6375: }
6376: fprintf(ficlog,"Strange: j1=%d One eigen value of 2x2 matrix of covariance is negative, lc1=%11.3e, lc2=%11.3e, v1=%11.3e, v2=%11.3e, cv12=%11.3e.\n It means that the matrix was not well estimated (varpij), for i=%2d, j=%2d, age=%4d .\n See files %s and %s. Probably WRONG RESULTS.\n", j1, lc1, lc2, v1, v2, cv12, i, j, (int)age,fileresprobcov, fileresprobcor);fflush(ficlog);
6377: /* lc1=fabs(lc1); */ /* If we want to have them positive */
6378: /* lc2=fabs(lc2); */
6379: }
1.220 brouard 6380:
1.222 brouard 6381: /* Eigen vectors */
6382: v11=(1./sqrt(1+(v1-lc1)*(v1-lc1)/cv12/cv12));
6383: /*v21=sqrt(1.-v11*v11); *//* error */
6384: v21=(lc1-v1)/cv12*v11;
6385: v12=-v21;
6386: v22=v11;
6387: tnalp=v21/v11;
6388: if(first1==1){
6389: first1=0;
6390: printf("%d %d%d-%d%d mu %.4e %.4e Var %.4e %.4e cor %.3f cov %.4e Eig %.3e %.3e 1stv %.3f %.3f tang %.3f\nOthers in log...\n",(int) age,k1,l1,k2,l2,mu1,mu2,v1,v2,c12,cv12,lc1,lc2,v11,v21,tnalp);
6391: }
6392: fprintf(ficlog,"%d %d%d-%d%d mu %.4e %.4e Var %.4e %.4e cor %.3f cov %.4e Eig %.3e %.3e 1stv %.3f %.3f tan %.3f\n",(int) age,k1,l1,k2,l2,mu1,mu2,v1,v2,c12,cv12,lc1,lc2,v11,v21,tnalp);
6393: /*printf(fignu*/
6394: /* mu1+ v11*lc1*cost + v12*lc2*sin(t) */
6395: /* mu2+ v21*lc1*cost + v22*lc2*sin(t) */
6396: if(first==1){
6397: first=0;
6398: fprintf(ficgp,"\n# Ellipsoids of confidence\n#\n");
6399: fprintf(ficgp,"\nset parametric;unset label");
6400: fprintf(ficgp,"\nset log y;set log x; set xlabel \"p%1d%1d (year-1)\";set ylabel \"p%1d%1d (year-1)\"",k1,l1,k2,l2);
6401: fprintf(ficgp,"\nset ter svg size 640, 480");
6402: fprintf(fichtmcov,"\n<br>Ellipsoids of confidence cov(p%1d%1d,p%1d%1d) expressed in year<sup>-1</sup>\
1.220 brouard 6403: :<a href=\"%s_%d%1d%1d-%1d%1d.svg\"> \
1.201 brouard 6404: %s_%d%1d%1d-%1d%1d.svg</A>, ",k1,l1,k2,l2,\
1.222 brouard 6405: subdirf2(optionfilefiname,"VARPIJGR_"), j1,k1,l1,k2,l2, \
6406: subdirf2(optionfilefiname,"VARPIJGR_"), j1,k1,l1,k2,l2);
6407: fprintf(fichtmcov,"\n<br><img src=\"%s_%d%1d%1d-%1d%1d.svg\"> ",subdirf2(optionfilefiname,"VARPIJGR_"), j1,k1,l1,k2,l2);
6408: fprintf(fichtmcov,"\n<br> Correlation at age %d (%.3f),",(int) age, c12);
6409: fprintf(ficgp,"\nset out \"%s_%d%1d%1d-%1d%1d.svg\"",subdirf2(optionfilefiname,"VARPIJGR_"), j1,k1,l1,k2,l2);
6410: fprintf(ficgp,"\nset label \"%d\" at %11.3e,%11.3e center",(int) age, mu1,mu2);
6411: fprintf(ficgp,"\n# Age %d, p%1d%1d - p%1d%1d",(int) age, k1,l1,k2,l2);
6412: fprintf(ficgp,"\nplot [-pi:pi] %11.3e+ %.3f*(%11.3e*%11.3e*cos(t)+%11.3e*%11.3e*sin(t)), %11.3e +%.3f*(%11.3e*%11.3e*cos(t)+%11.3e*%11.3e*sin(t)) not", \
6413: mu1,std,v11,sqrt(lc1),v12,sqrt(lc2), \
6414: mu2,std,v21,sqrt(lc1),v22,sqrt(lc2));
6415: }else{
6416: first=0;
6417: fprintf(fichtmcov," %d (%.3f),",(int) age, c12);
6418: fprintf(ficgp,"\n# Age %d, p%1d%1d - p%1d%1d",(int) age, k1,l1,k2,l2);
6419: fprintf(ficgp,"\nset label \"%d\" at %11.3e,%11.3e center",(int) age, mu1,mu2);
6420: fprintf(ficgp,"\nreplot %11.3e+ %.3f*(%11.3e*%11.3e*cos(t)+%11.3e*%11.3e*sin(t)), %11.3e +%.3f*(%11.3e*%11.3e*cos(t)+%11.3e*%11.3e*sin(t)) not", \
6421: mu1,std,v11,sqrt(lc1),v12,sqrt(lc2), \
6422: mu2,std,v21,sqrt(lc1),v22,sqrt(lc2));
6423: }/* if first */
6424: } /* age mod 5 */
6425: } /* end loop age */
6426: fprintf(ficgp,"\nset out;\nset out \"%s_%d%1d%1d-%1d%1d.svg\";replot;set out;",subdirf2(optionfilefiname,"VARPIJGR_"), j1,k1,l1,k2,l2);
6427: first=1;
6428: } /*l12 */
6429: } /* k12 */
6430: } /*l1 */
6431: }/* k1 */
6432: } /* loop on combination of covariates j1 */
6433: free_ma3x(varpij,1,nlstate,1,nlstate+ndeath,(int) bage, (int)fage);
6434: free_matrix(mu,1,(nlstate+ndeath)*(nlstate+ndeath),(int) bage, (int)fage);
6435: free_matrix(doldm,1,(nlstate)*(nlstate+ndeath),1,(nlstate)*(nlstate+ndeath));
6436: free_matrix(dnewm,1,(nlstate)*(nlstate+ndeath),1,npar);
6437: free_vector(xp,1,npar);
6438: fclose(ficresprob);
6439: fclose(ficresprobcov);
6440: fclose(ficresprobcor);
6441: fflush(ficgp);
6442: fflush(fichtmcov);
6443: }
1.126 brouard 6444:
6445:
6446: /******************* Printing html file ***********/
1.201 brouard 6447: void printinghtml(char fileresu[], char title[], char datafile[], int firstpass, \
1.126 brouard 6448: int lastpass, int stepm, int weightopt, char model[],\
6449: int imx,int jmin, int jmax, double jmeanint,char rfileres[],\
1.258 brouard 6450: int popforecast, int mobilav, int prevfcast, int mobilavproj, int backcast, int estepm , \
1.213 brouard 6451: double jprev1, double mprev1,double anprev1, double dateprev1, \
6452: double jprev2, double mprev2,double anprev2, double dateprev2){
1.237 brouard 6453: int jj1, k1, i1, cpt, k4, nres;
1.126 brouard 6454:
6455: fprintf(fichtm,"<ul><li><a href='#firstorder'>Result files (first order: no variance)</a>\n \
6456: <li><a href='#secondorder'>Result files (second order (variance)</a>\n \
6457: </ul>");
1.237 brouard 6458: fprintf(fichtm,"<ul><li> model=1+age+%s\n \
6459: </ul>", model);
1.214 brouard 6460: fprintf(fichtm,"<ul><li><h4><a name='firstorder'>Result files (first order: no variance)</a></h4>\n");
6461: fprintf(fichtm,"<li>- Observed frequency between two states (during the period defined between %.lf/%.lf/%.lf and %.lf/%.lf/%.lf): <a href=\"%s\">%s</a> (html file)<br/>\n",
6462: jprev1, mprev1,anprev1,jprev2, mprev2,anprev2,subdirfext3(optionfilefiname,"PHTMFR_",".htm"),subdirfext3(optionfilefiname,"PHTMFR_",".htm"));
6463: fprintf(fichtm,"<li> - Observed prevalence in each state (during the period defined between %.lf/%.lf/%.lf and %.lf/%.lf/%.lf): <a href=\"%s\">%s</a> (html file) ",
1.213 brouard 6464: jprev1, mprev1,anprev1,jprev2, mprev2,anprev2,subdirfext3(optionfilefiname,"PHTM_",".htm"),subdirfext3(optionfilefiname,"PHTM_",".htm"));
6465: fprintf(fichtm,", <a href=\"%s\">%s</a> (text file) <br>\n",subdirf2(fileresu,"P_"),subdirf2(fileresu,"P_"));
1.126 brouard 6466: fprintf(fichtm,"\
6467: - Estimated transition probabilities over %d (stepm) months: <a href=\"%s\">%s</a><br>\n ",
1.201 brouard 6468: stepm,subdirf2(fileresu,"PIJ_"),subdirf2(fileresu,"PIJ_"));
1.126 brouard 6469: fprintf(fichtm,"\
1.217 brouard 6470: - Estimated back transition probabilities over %d (stepm) months: <a href=\"%s\">%s</a><br>\n ",
6471: stepm,subdirf2(fileresu,"PIJB_"),subdirf2(fileresu,"PIJB_"));
6472: fprintf(fichtm,"\
1.126 brouard 6473: - Period (stable) prevalence in each health state: <a href=\"%s\">%s</a> <br>\n",
1.201 brouard 6474: subdirf2(fileresu,"PL_"),subdirf2(fileresu,"PL_"));
1.126 brouard 6475: fprintf(fichtm,"\
1.217 brouard 6476: - Period (stable) back prevalence in each health state: <a href=\"%s\">%s</a> <br>\n",
6477: subdirf2(fileresu,"PLB_"),subdirf2(fileresu,"PLB_"));
6478: fprintf(fichtm,"\
1.211 brouard 6479: - (a) Life expectancies by health status at initial age, e<sub>i.</sub> (b) health expectancies by health status at initial age, e<sub>ij</sub> . If one or more covariates are included, specific tables for each value of the covariate are output in sequences within the same file (estepm=%2d months): \
1.126 brouard 6480: <a href=\"%s\">%s</a> <br>\n",
1.201 brouard 6481: estepm,subdirf2(fileresu,"E_"),subdirf2(fileresu,"E_"));
1.211 brouard 6482: if(prevfcast==1){
6483: fprintf(fichtm,"\
6484: - Prevalence projections by age and states: \
1.201 brouard 6485: <a href=\"%s\">%s</a> <br>\n</li>", subdirf2(fileresu,"F_"),subdirf2(fileresu,"F_"));
1.211 brouard 6486: }
1.126 brouard 6487:
6488:
1.225 brouard 6489: m=pow(2,cptcoveff);
1.222 brouard 6490: if (cptcovn < 1) {m=1;ncodemax[1]=1;}
1.126 brouard 6491:
1.264 brouard 6492: fprintf(fichtm," \n<ul><li><b>Graphs</b></li><p>");
6493:
6494: jj1=0;
6495:
6496: fprintf(fichtm," \n<ul>");
6497: for(nres=1; nres <= nresult; nres++) /* For each resultline */
6498: for(k1=1; k1<=m;k1++){ /* For each combination of covariate */
6499: if(m != 1 && TKresult[nres]!= k1)
6500: continue;
6501: jj1++;
6502: if (cptcovn > 0) {
6503: fprintf(fichtm,"\n<li><a size=\"1\" color=\"#EC5E5E\" href=\"#rescov");
6504: for (cpt=1; cpt<=cptcoveff;cpt++){
6505: fprintf(fichtm,"_V%d=%d_",Tvresult[nres][cpt],(int)Tresult[nres][cpt]);
6506: }
6507: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
6508: fprintf(fichtm,"_V%d=%f_",Tvqresult[nres][k4],Tqresult[nres][k4]);
6509: }
6510: fprintf(fichtm,"\">");
6511:
6512: /* if(nqfveff+nqtveff 0) */ /* Test to be done */
6513: fprintf(fichtm,"************ Results for covariates");
6514: for (cpt=1; cpt<=cptcoveff;cpt++){
6515: fprintf(fichtm," V%d=%d ",Tvresult[nres][cpt],(int)Tresult[nres][cpt]);
6516: }
6517: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
6518: fprintf(fichtm," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
6519: }
6520: if(invalidvarcomb[k1]){
6521: fprintf(fichtm," Warning Combination (%d) ignored because no cases ",k1);
6522: continue;
6523: }
6524: fprintf(fichtm,"</a></li>");
6525: } /* cptcovn >0 */
6526: }
6527: fprintf(fichtm," \n</ul>");
6528:
1.222 brouard 6529: jj1=0;
1.237 brouard 6530:
6531: for(nres=1; nres <= nresult; nres++) /* For each resultline */
1.241 brouard 6532: for(k1=1; k1<=m;k1++){ /* For each combination of covariate */
1.253 brouard 6533: if(m != 1 && TKresult[nres]!= k1)
1.237 brouard 6534: continue;
1.220 brouard 6535:
1.222 brouard 6536: /* for(i1=1; i1<=ncodemax[k1];i1++){ */
6537: jj1++;
6538: if (cptcovn > 0) {
1.264 brouard 6539: fprintf(fichtm,"\n<p><a name=\"rescov");
6540: for (cpt=1; cpt<=cptcoveff;cpt++){
6541: fprintf(fichtm,"_V%d=%d_",Tvresult[nres][cpt],(int)Tresult[nres][cpt]);
6542: }
6543: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
6544: fprintf(fichtm,"_V%d=%f_",Tvqresult[nres][k4],Tqresult[nres][k4]);
6545: }
6546: fprintf(fichtm,"\"</a>");
6547:
1.222 brouard 6548: fprintf(fichtm,"<hr size=\"2\" color=\"#EC5E5E\">************ Results for covariates");
1.225 brouard 6549: for (cpt=1; cpt<=cptcoveff;cpt++){
1.237 brouard 6550: fprintf(fichtm," V%d=%d ",Tvresult[nres][cpt],(int)Tresult[nres][cpt]);
6551: printf(" V%d=%d ",Tvresult[nres][cpt],Tresult[nres][cpt]);fflush(stdout);
6552: /* fprintf(fichtm," V%d=%d ",Tvaraff[cpt],nbcode[Tvaraff[cpt]][codtabm(jj1,cpt)]); */
6553: /* printf(" V%d=%d ",Tvaraff[cpt],nbcode[Tvaraff[cpt]][codtabm(jj1,cpt)]);fflush(stdout); */
1.222 brouard 6554: }
1.237 brouard 6555: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
6556: fprintf(fichtm," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
6557: printf(" V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);fflush(stdout);
6558: }
6559:
1.230 brouard 6560: /* if(nqfveff+nqtveff 0) */ /* Test to be done */
1.222 brouard 6561: fprintf(fichtm," ************\n<hr size=\"2\" color=\"#EC5E5E\">");
6562: if(invalidvarcomb[k1]){
6563: fprintf(fichtm,"\n<h3>Combination (%d) ignored because no cases </h3>\n",k1);
6564: printf("\nCombination (%d) ignored because no cases \n",k1);
6565: continue;
6566: }
6567: }
6568: /* aij, bij */
1.259 brouard 6569: fprintf(fichtm,"<br>- Logit model (yours is: logit(pij)=log(pij/pii)= aij+ bij age+%s) as a function of age: <a href=\"%s_%d-1-%d.svg\">%s_%d-1-%d.svg</a><br> \
1.241 brouard 6570: <img src=\"%s_%d-1-%d.svg\">",model,subdirf2(optionfilefiname,"PE_"),k1,nres,subdirf2(optionfilefiname,"PE_"),k1,nres,subdirf2(optionfilefiname,"PE_"),k1,nres);
1.222 brouard 6571: /* Pij */
1.241 brouard 6572: fprintf(fichtm,"<br>\n- P<sub>ij</sub> or conditional probabilities to be observed in state j being in state i, %d (stepm) months before: <a href=\"%s_%d-2-%d.svg\">%s_%d-2-%d.svg</a><br> \
6573: <img src=\"%s_%d-2-%d.svg\">",stepm,subdirf2(optionfilefiname,"PE_"),k1,nres,subdirf2(optionfilefiname,"PE_"),k1,nres,subdirf2(optionfilefiname,"PE_"),k1,nres);
1.222 brouard 6574: /* Quasi-incidences */
6575: fprintf(fichtm,"<br>\n- I<sub>ij</sub> or Conditional probabilities to be observed in state j being in state i %d (stepm) months\
1.220 brouard 6576: before but expressed in per year i.e. quasi incidences if stepm is small and probabilities too, \
1.211 brouard 6577: incidence (rates) are the limit when h tends to zero of the ratio of the probability <sub>h</sub>P<sub>ij</sub> \
1.241 brouard 6578: divided by h: <sub>h</sub>P<sub>ij</sub>/h : <a href=\"%s_%d-3-%d.svg\">%s_%d-3-%d.svg</a><br> \
6579: <img src=\"%s_%d-3-%d.svg\">",stepm,subdirf2(optionfilefiname,"PE_"),k1,nres,subdirf2(optionfilefiname,"PE_"),k1,nres,subdirf2(optionfilefiname,"PE_"),k1,nres);
1.222 brouard 6580: /* Survival functions (period) in state j */
6581: for(cpt=1; cpt<=nlstate;cpt++){
1.241 brouard 6582: fprintf(fichtm,"<br>\n- Survival functions in state %d. Or probability to survive in state %d being in state (1 to %d) at different ages. <a href=\"%s_%d-%d-%d.svg\">%s_%d-%d-%d.svg</a><br> \
6583: <img src=\"%s_%d-%d-%d.svg\">", cpt, cpt, nlstate, subdirf2(optionfilefiname,"LIJ_"),cpt,k1,nres,subdirf2(optionfilefiname,"LIJ_"),cpt,k1,nres,subdirf2(optionfilefiname,"LIJ_"),cpt,k1,nres);
1.222 brouard 6584: }
6585: /* State specific survival functions (period) */
6586: for(cpt=1; cpt<=nlstate;cpt++){
6587: fprintf(fichtm,"<br>\n- Survival functions from state %d in each live state and total.\
1.220 brouard 6588: Or probability to survive in various states (1 to %d) being in state %d at different ages. \
1.241 brouard 6589: <a href=\"%s_%d-%d-%d.svg\">%s_%d%d-%d.svg</a><br> <img src=\"%s_%d-%d-%d.svg\">", cpt, nlstate, cpt, subdirf2(optionfilefiname,"LIJT_"),cpt,k1,nres,subdirf2(optionfilefiname,"LIJT_"),cpt,k1,nres,subdirf2(optionfilefiname,"LIJT_"),cpt,k1,nres);
1.222 brouard 6590: }
6591: /* Period (stable) prevalence in each health state */
6592: for(cpt=1; cpt<=nlstate;cpt++){
1.264 brouard 6593: fprintf(fichtm,"<br>\n- Convergence to period (stable) prevalence in state %d. Or probability for a person being in state (1 to %d) at different ages, to be in state %d some years after. <a href=\"%s_%d-%d-%d.svg\">%s_%d-%d-%d.svg</a><br> \
6594: <img src=\"%s_%d-%d-%d.svg\">", cpt, nlstate, cpt, subdirf2(optionfilefiname,"P_"),cpt,k1,nres,subdirf2(optionfilefiname,"P_"),cpt,k1,nres,subdirf2(optionfilefiname,"P_"),cpt,k1,nres);
1.222 brouard 6595: }
6596: if(backcast==1){
6597: /* Period (stable) back prevalence in each health state */
6598: for(cpt=1; cpt<=nlstate;cpt++){
1.264 brouard 6599: fprintf(fichtm,"<br>\n- Convergence to mixed (stable) back prevalence in state %d. Or probability for a person to be in state %d at a younger age, knowing that she/he was in state (1 to %d) at different older ages. <a href=\"%s_%d-%d-%d.svg\">%s_%d-%d-%d.svg</a><br> \
1.241 brouard 6600: <img src=\"%s_%d-%d-%d.svg\">", cpt, cpt, nlstate, subdirf2(optionfilefiname,"PB_"),cpt,k1,nres,subdirf2(optionfilefiname,"PB_"),cpt,k1,nres,subdirf2(optionfilefiname,"PB_"),cpt,k1,nres);
1.222 brouard 6601: }
1.217 brouard 6602: }
1.222 brouard 6603: if(prevfcast==1){
6604: /* Projection of prevalence up to period (stable) prevalence in each health state */
6605: for(cpt=1; cpt<=nlstate;cpt++){
1.258 brouard 6606: fprintf(fichtm,"<br>\n- Projection of cross-sectional prevalence (estimated with cases observed from %.1f to %.1f and mobil_average=%d) up to period (stable) prevalence in state %d. Or probability to be in state %d being in state (1 to %d) at different ages. <a href=\"%s_%d-%d-%d.svg\">%s_%d-%d-%d.svg</a><br> \
6607: <img src=\"%s_%d-%d-%d.svg\">", dateprev1, dateprev2, mobilavproj, cpt, cpt, nlstate, subdirf2(optionfilefiname,"PROJ_"),cpt,k1,nres,subdirf2(optionfilefiname,"PROJ_"),cpt,k1,nres,subdirf2(optionfilefiname,"PROJ_"),cpt,k1,nres);
1.222 brouard 6608: }
6609: }
1.220 brouard 6610:
1.222 brouard 6611: for(cpt=1; cpt<=nlstate;cpt++) {
1.241 brouard 6612: fprintf(fichtm,"\n<br>- Life expectancy by health state (%d) at initial age and its decomposition into health expectancies in each alive state (1 to %d) (or area under each survival functions): <a href=\"%s_%d-%d-%d.svg\">%s_%d-%d-%d.svg</a> <br> \
6613: <img src=\"%s_%d-%d-%d.svg\">",cpt,nlstate,subdirf2(optionfilefiname,"EXP_"),cpt,k1,nres,subdirf2(optionfilefiname,"EXP_"),cpt,k1,nres,subdirf2(optionfilefiname,"EXP_"),cpt,k1,nres);
1.222 brouard 6614: }
6615: /* } /\* end i1 *\/ */
6616: }/* End k1 */
6617: fprintf(fichtm,"</ul>");
1.126 brouard 6618:
1.222 brouard 6619: fprintf(fichtm,"\
1.126 brouard 6620: \n<br><li><h4> <a name='secondorder'>Result files (second order: variances)</a></h4>\n\
1.193 brouard 6621: - Parameter file with estimated parameters and covariance matrix: <a href=\"%s\">%s</a> <br> \
1.203 brouard 6622: - 95%% confidence intervals and Wald tests of the estimated parameters are in the log file if optimization has been done (mle != 0).<br> \
1.197 brouard 6623: But because parameters are usually highly correlated (a higher incidence of disability \
6624: and a higher incidence of recovery can give very close observed transition) it might \
6625: be very useful to look not only at linear confidence intervals estimated from the \
6626: variances but at the covariance matrix. And instead of looking at the estimated coefficients \
6627: (parameters) of the logistic regression, it might be more meaningful to visualize the \
6628: covariance matrix of the one-step probabilities. \
6629: See page 'Matrix of variance-covariance of one-step probabilities' below. \n", rfileres,rfileres);
1.126 brouard 6630:
1.222 brouard 6631: fprintf(fichtm," - Standard deviation of one-step probabilities: <a href=\"%s\">%s</a> <br>\n",
6632: subdirf2(fileresu,"PROB_"),subdirf2(fileresu,"PROB_"));
6633: fprintf(fichtm,"\
1.126 brouard 6634: - Variance-covariance of one-step probabilities: <a href=\"%s\">%s</a> <br>\n",
1.222 brouard 6635: subdirf2(fileresu,"PROBCOV_"),subdirf2(fileresu,"PROBCOV_"));
1.126 brouard 6636:
1.222 brouard 6637: fprintf(fichtm,"\
1.126 brouard 6638: - Correlation matrix of one-step probabilities: <a href=\"%s\">%s</a> <br>\n",
1.222 brouard 6639: subdirf2(fileresu,"PROBCOR_"),subdirf2(fileresu,"PROBCOR_"));
6640: fprintf(fichtm,"\
1.126 brouard 6641: - Variances and covariances of health expectancies by age and <b>initial health status</b> (cov(e<sup>ij</sup>,e<sup>kl</sup>)(estepm=%2d months): \
6642: <a href=\"%s\">%s</a> <br>\n</li>",
1.201 brouard 6643: estepm,subdirf2(fileresu,"CVE_"),subdirf2(fileresu,"CVE_"));
1.222 brouard 6644: fprintf(fichtm,"\
1.126 brouard 6645: - (a) Health expectancies by health status at initial age (e<sup>ij</sup>) and standard errors (in parentheses) (b) life expectancies and standard errors (e<sup>i.</sup>=e<sup>i1</sup>+e<sup>i2</sup>+...)(estepm=%2d months): \
6646: <a href=\"%s\">%s</a> <br>\n</li>",
1.201 brouard 6647: estepm,subdirf2(fileresu,"STDE_"),subdirf2(fileresu,"STDE_"));
1.222 brouard 6648: fprintf(fichtm,"\
1.128 brouard 6649: - Variances and covariances of health expectancies by age. Status (i) based health expectancies (in state j), e<sup>ij</sup> are weighted by the period prevalences in each state i (if popbased=1, an additional computation is done using the cross-sectional prevalences, i.e population based) (estepm=%d months): <a href=\"%s\">%s</a><br>\n",
1.222 brouard 6650: estepm, subdirf2(fileresu,"V_"),subdirf2(fileresu,"V_"));
6651: fprintf(fichtm,"\
1.128 brouard 6652: - Total life expectancy and total health expectancies to be spent in each health state e<sup>.j</sup> with their standard errors (if popbased=1, an additional computation is done using the cross-sectional prevalences, i.e population based) (estepm=%d months): <a href=\"%s\">%s</a> <br>\n",
1.222 brouard 6653: estepm, subdirf2(fileresu,"T_"),subdirf2(fileresu,"T_"));
6654: fprintf(fichtm,"\
1.126 brouard 6655: - Standard deviation of period (stable) prevalences: <a href=\"%s\">%s</a> <br>\n",\
1.222 brouard 6656: subdirf2(fileresu,"VPL_"),subdirf2(fileresu,"VPL_"));
1.126 brouard 6657:
6658: /* if(popforecast==1) fprintf(fichtm,"\n */
6659: /* - Prevalences forecasting: <a href=\"f%s\">f%s</a> <br>\n */
6660: /* - Population forecasting (if popforecast=1): <a href=\"pop%s\">pop%s</a> <br>\n */
6661: /* <br>",fileres,fileres,fileres,fileres); */
6662: /* else */
6663: /* fprintf(fichtm,"\n No population forecast: popforecast = %d (instead of 1) or stepm = %d (instead of 1) or model=%s (instead of .)<br><br></li>\n",popforecast, stepm, model); */
1.222 brouard 6664: fflush(fichtm);
6665: fprintf(fichtm," <ul><li><b>Graphs</b></li><p>");
1.126 brouard 6666:
1.225 brouard 6667: m=pow(2,cptcoveff);
1.222 brouard 6668: if (cptcovn < 1) {m=1;ncodemax[1]=1;}
1.126 brouard 6669:
1.222 brouard 6670: jj1=0;
1.237 brouard 6671:
1.241 brouard 6672: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
1.222 brouard 6673: for(k1=1; k1<=m;k1++){
1.253 brouard 6674: if(m != 1 && TKresult[nres]!= k1)
1.237 brouard 6675: continue;
1.222 brouard 6676: /* for(i1=1; i1<=ncodemax[k1];i1++){ */
6677: jj1++;
1.126 brouard 6678: if (cptcovn > 0) {
6679: fprintf(fichtm,"<hr size=\"2\" color=\"#EC5E5E\">************ Results for covariates");
1.225 brouard 6680: for (cpt=1; cpt<=cptcoveff;cpt++) /**< cptcoveff number of variables */
1.237 brouard 6681: fprintf(fichtm," V%d=%d ",Tvresult[nres][cpt],Tresult[nres][cpt]);
6682: /* fprintf(fichtm," V%d=%d ",Tvaraff[cpt],nbcode[Tvaraff[cpt]][codtabm(jj1,cpt)]); */
6683: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
6684: fprintf(fichtm," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
6685: }
6686:
1.126 brouard 6687: fprintf(fichtm," ************\n<hr size=\"2\" color=\"#EC5E5E\">");
1.220 brouard 6688:
1.222 brouard 6689: if(invalidvarcomb[k1]){
6690: fprintf(fichtm,"\n<h4>Combination (%d) ignored because no cases </h4>\n",k1);
6691: continue;
6692: }
1.126 brouard 6693: }
6694: for(cpt=1; cpt<=nlstate;cpt++) {
1.258 brouard 6695: fprintf(fichtm,"\n<br>- Observed (cross-sectional with mov_average=%d) and period (incidence based) \
1.241 brouard 6696: prevalence (with 95%% confidence interval) in state (%d): <a href=\"%s_%d-%d-%d.svg\"> %s_%d-%d-%d.svg</a>\n <br>\
1.258 brouard 6697: <img src=\"%s_%d-%d-%d.svg\">",mobilav,cpt,subdirf2(optionfilefiname,"V_"),cpt,k1,nres,subdirf2(optionfilefiname,"V_"),cpt,k1,nres,subdirf2(optionfilefiname,"V_"),cpt,k1,nres);
1.126 brouard 6698: }
6699: fprintf(fichtm,"\n<br>- Total life expectancy by age and \
1.128 brouard 6700: health expectancies in states (1) and (2). If popbased=1 the smooth (due to the model) \
6701: true period expectancies (those weighted with period prevalences are also\
6702: drawn in addition to the population based expectancies computed using\
1.241 brouard 6703: observed and cahotic prevalences: <a href=\"%s_%d-%d.svg\">%s_%d-%d.svg</a>\n<br>\
6704: <img src=\"%s_%d-%d.svg\">",subdirf2(optionfilefiname,"E_"),k1,nres,subdirf2(optionfilefiname,"E_"),k1,nres,subdirf2(optionfilefiname,"E_"),k1,nres);
1.222 brouard 6705: /* } /\* end i1 *\/ */
6706: }/* End k1 */
1.241 brouard 6707: }/* End nres */
1.222 brouard 6708: fprintf(fichtm,"</ul>");
6709: fflush(fichtm);
1.126 brouard 6710: }
6711:
6712: /******************* Gnuplot file **************/
1.223 brouard 6713: void printinggnuplot(char fileresu[], char optionfilefiname[], double ageminpar, double agemaxpar, double fage , int prevfcast, int backcast, char pathc[], double p[]){
1.126 brouard 6714:
6715: char dirfileres[132],optfileres[132];
1.264 brouard 6716: char gplotcondition[132], gplotlabel[132];
1.237 brouard 6717: int cpt=0,k1=0,i=0,k=0,j=0,jk=0,k2=0,k3=0,k4=0,ij=0, ijp=0, l=0;
1.211 brouard 6718: int lv=0, vlv=0, kl=0;
1.130 brouard 6719: int ng=0;
1.201 brouard 6720: int vpopbased;
1.223 brouard 6721: int ioffset; /* variable offset for columns */
1.235 brouard 6722: int nres=0; /* Index of resultline */
1.219 brouard 6723:
1.126 brouard 6724: /* if((ficgp=fopen(optionfilegnuplot,"a"))==NULL) { */
6725: /* printf("Problem with file %s",optionfilegnuplot); */
6726: /* fprintf(ficlog,"Problem with file %s",optionfilegnuplot); */
6727: /* } */
6728:
6729: /*#ifdef windows */
6730: fprintf(ficgp,"cd \"%s\" \n",pathc);
1.223 brouard 6731: /*#endif */
1.225 brouard 6732: m=pow(2,cptcoveff);
1.126 brouard 6733:
1.202 brouard 6734: /* Contribution to likelihood */
6735: /* Plot the probability implied in the likelihood */
1.223 brouard 6736: fprintf(ficgp,"\n# Contributions to the Likelihood, mle >=1. For mle=4 no interpolation, pure matrix products.\n#\n");
6737: fprintf(ficgp,"\n set log y; unset log x;set xlabel \"Age\"; set ylabel \"Likelihood (-2Log(L))\";");
6738: /* fprintf(ficgp,"\nset ter svg size 640, 480"); */ /* Too big for svg */
6739: fprintf(ficgp,"\nset ter pngcairo size 640, 480");
1.204 brouard 6740: /* nice for mle=4 plot by number of matrix products.
1.202 brouard 6741: replot "rrtest1/toto.txt" u 2:($4 == 1 && $5==2 ? $9 : 1/0):5 t "p12" with point lc 1 */
6742: /* replot exp(p1+p2*x)/(1+exp(p1+p2*x)+exp(p3+p4*x)+exp(p5+p6*x)) t "p12(x)" */
1.223 brouard 6743: /* fprintf(ficgp,"\nset out \"%s.svg\";",subdirf2(optionfilefiname,"ILK_")); */
6744: fprintf(ficgp,"\nset out \"%s-dest.png\";",subdirf2(optionfilefiname,"ILK_"));
6745: fprintf(ficgp,"\nset log y;plot \"%s\" u 2:(-$13):6 t \"All sample, transitions colored by destination\" with dots lc variable; set out;\n",subdirf(fileresilk));
6746: fprintf(ficgp,"\nset out \"%s-ori.png\";",subdirf2(optionfilefiname,"ILK_"));
6747: fprintf(ficgp,"\nset log y;plot \"%s\" u 2:(-$13):5 t \"All sample, transitions colored by origin\" with dots lc variable; set out;\n\n",subdirf(fileresilk));
6748: for (i=1; i<= nlstate ; i ++) {
6749: fprintf(ficgp,"\nset out \"%s-p%dj.png\";set ylabel \"Probability for each individual/wave\";",subdirf2(optionfilefiname,"ILK_"),i);
6750: fprintf(ficgp,"unset log;\n# plot weighted, mean weight should have point size of 0.5\n plot \"%s\"",subdirf(fileresilk));
6751: fprintf(ficgp," u 2:($5 == %d && $6==%d ? $10 : 1/0):($12/4.):6 t \"p%d%d\" with points pointtype 7 ps variable lc variable \\\n",i,1,i,1);
6752: for (j=2; j<= nlstate+ndeath ; j ++) {
6753: fprintf(ficgp,",\\\n \"\" u 2:($5 == %d && $6==%d ? $10 : 1/0):($12/4.):6 t \"p%d%d\" with points pointtype 7 ps variable lc variable ",i,j,i,j);
6754: }
6755: fprintf(ficgp,";\nset out; unset ylabel;\n");
6756: }
6757: /* unset log; plot "rrtest1_sorted_4/ILK_rrtest1_sorted_4.txt" u 2:($4 == 1 && $5==2 ? $9 : 1/0):5 t "p12" with points lc variable */
6758: /* fprintf(ficgp,"\nset log y;plot \"%s\" u 2:(-$11):3 t \"All sample, all transitions\" with dots lc variable",subdirf(fileresilk)); */
6759: /* fprintf(ficgp,"\nreplot \"%s\" u 2:($3 <= 3 ? -$11 : 1/0):3 t \"First 3 individuals\" with line lc variable", subdirf(fileresilk)); */
6760: fprintf(ficgp,"\nset out;unset log\n");
6761: /* fprintf(ficgp,"\nset out \"%s.svg\"; replot; set out; # bug gnuplot",subdirf2(optionfilefiname,"ILK_")); */
1.202 brouard 6762:
1.126 brouard 6763: strcpy(dirfileres,optionfilefiname);
6764: strcpy(optfileres,"vpl");
1.223 brouard 6765: /* 1eme*/
1.238 brouard 6766: for (cpt=1; cpt<= nlstate ; cpt ++){ /* For each live state */
6767: for (k1=1; k1<= m ; k1 ++){ /* For each valid combination of covariate */
1.236 brouard 6768: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
1.238 brouard 6769: /* plot [100000000000000000000:-100000000000000000000] "mysbiaspar/vplrmysbiaspar.txt to check */
1.253 brouard 6770: if(m != 1 && TKresult[nres]!= k1)
1.238 brouard 6771: continue;
6772: /* We are interested in selected combination by the resultline */
1.246 brouard 6773: /* printf("\n# 1st: Period (stable) prevalence with CI: 'VPL_' files and live state =%d ", cpt); */
1.238 brouard 6774: fprintf(ficgp,"\n# 1st: Period (stable) prevalence with CI: 'VPL_' files and live state =%d ", cpt);
1.264 brouard 6775: strcpy(gplotlabel,"(");
1.238 brouard 6776: for (k=1; k<=cptcoveff; k++){ /* For each covariate k get corresponding value lv for combination k1 */
6777: lv= decodtabm(k1,k,cptcoveff); /* Should be the value of the covariate corresponding to k1 combination */
6778: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6779: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6780: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6781: vlv= nbcode[Tvaraff[k]][lv]; /* vlv is the value of the covariate lv, 0 or 1 */
6782: /* For each combination of covariate k1 (V1=1, V3=0), we printed the current covariate k and its value vlv */
1.246 brouard 6783: /* printf(" V%d=%d ",Tvaraff[k],vlv); */
1.238 brouard 6784: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
1.264 brouard 6785: sprintf(gplotlabel+strlen(gplotlabel)," V%d=%d ",Tvaraff[k],vlv);
1.238 brouard 6786: }
6787: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
1.246 brouard 6788: /* printf(" V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]); */
1.238 brouard 6789: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
1.264 brouard 6790: sprintf(gplotlabel+strlen(gplotlabel)," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
6791: }
6792: strcpy(gplotlabel+strlen(gplotlabel),")");
1.246 brouard 6793: /* printf("\n#\n"); */
1.238 brouard 6794: fprintf(ficgp,"\n#\n");
6795: if(invalidvarcomb[k1]){
1.260 brouard 6796: /*k1=k1-1;*/ /* To be checked */
1.238 brouard 6797: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6798: continue;
6799: }
1.235 brouard 6800:
1.241 brouard 6801: fprintf(ficgp,"\nset out \"%s_%d-%d-%d.svg\" \n",subdirf2(optionfilefiname,"V_"),cpt,k1,nres);
6802: fprintf(ficgp,"\n#set out \"V_%s_%d-%d-%d.svg\" \n",optionfilefiname,cpt,k1,nres);
1.264 brouard 6803: fprintf(ficgp,"set label \"Alive state %d %s\" at graph 0.98,0.5 center rotate font \"Helvetica,12\"\n",cpt,gplotlabel);
1.260 brouard 6804: fprintf(ficgp,"set xlabel \"Age\" \nset ylabel \"Probability\" \nset ter svg size 640, 480\nplot [%.f:%.f] \"%s\" every :::%d::%d u 1:($2==%d ? $3:1/0) \"%%lf %%lf",ageminpar,fage,subdirf2(fileresu,"VPL_"),nres-1,nres-1,nres);
6805: /* fprintf(ficgp,"set xlabel \"Age\" \nset ylabel \"Probability\" \nset ter svg size 640, 480\nplot [%.f:%.f] \"%s\" every :::%d::%d u 1:($2==%d ? $3:1/0) \"%%lf %%lf",ageminpar,fage,subdirf2(fileresu,"VPL_"),k1-1,k1-1,nres); */
6806: /* k1-1 error should be nres-1*/
1.238 brouard 6807: for (i=1; i<= nlstate ; i ++) {
6808: if (i==cpt) fprintf(ficgp," %%lf (%%lf)");
6809: else fprintf(ficgp," %%*lf (%%*lf)");
6810: }
1.260 brouard 6811: fprintf(ficgp,"\" t\"Period (stable) prevalence\" w l lt 0,\"%s\" every :::%d::%d u 1:($2==%d ? $3+1.96*$4 : 1/0) \"%%lf %%lf",subdirf2(fileresu,"VPL_"),nres-1,nres-1,nres);
1.238 brouard 6812: for (i=1; i<= nlstate ; i ++) {
6813: if (i==cpt) fprintf(ficgp," %%lf (%%lf)");
6814: else fprintf(ficgp," %%*lf (%%*lf)");
6815: }
1.260 brouard 6816: fprintf(ficgp,"\" t\"95%% CI\" w l lt 1,\"%s\" every :::%d::%d u 1:($2==%d ? $3-1.96*$4 : 1/0) \"%%lf %%lf",subdirf2(fileresu,"VPL_"),nres-1,nres-1,nres);
1.238 brouard 6817: for (i=1; i<= nlstate ; i ++) {
6818: if (i==cpt) fprintf(ficgp," %%lf (%%lf)");
6819: else fprintf(ficgp," %%*lf (%%*lf)");
6820: }
1.265 ! brouard 6821: /* fprintf(ficgp,"\" t\"\" w l lt 1,\"%s\" every :::%d::%d u 1:($%d) t\"Observed prevalence\" w l lt 2",subdirf2(fileresu,"P_"),k1-1,k1-1,2+4*(cpt-1)); */
! 6822:
! 6823: fprintf(ficgp,"\" t\"\" w l lt 1,\"%s\" u 1:((",subdirf2(fileresu,"P_"));
! 6824: if(cptcoveff ==0){
! 6825: fprintf(ficgp,"$%d)) t 'Observed prevalence in state %d' with line lt 3", 2+(cpt-1), cpt );
! 6826: }else{
! 6827: kl=0;
! 6828: for (k=1; k<=cptcoveff; k++){ /* For each combination of covariate */
! 6829: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate value corresponding to k1 combination and kth covariate */
! 6830: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
! 6831: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
! 6832: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
! 6833: vlv= nbcode[Tvaraff[k]][lv];
! 6834: kl++;
! 6835: /* kl=6+(cpt-1)*(nlstate+1)+1+(i-1); /\* 6+(1-1)*(2+1)+1+(1-1)=7, 6+(2-1)(2+1)+1+(1-1)=10 *\/ */
! 6836: /*6+(cpt-1)*(nlstate+1)+1+(i-1)+(nlstate+1)*nlstate; 6+(1-1)*(2+1)+1+(1-1) +(2+1)*2=13 */
! 6837: /*6+1+(i-1)+(nlstate+1)*nlstate; 6+1+(1-1) +(2+1)*2=13 */
! 6838: /* '' u 6:(($1==1 && $2==0 && $3==2 && $4==0)? $9/(1.-$15) : 1/0):($5==2000? 3:2) t 'p.1' with line lc variable*/
! 6839: if(k==cptcoveff){
! 6840: fprintf(ficgp,"$%d==%d && $%d==%d)? $%d : 1/0) t 'Observed prevalence in state %d' w l lt 2",kl+1, Tvaraff[k],kl+1+1,nbcode[Tvaraff[k]][lv], \
! 6841: 2+cptcoveff*2+3*(cpt-1), cpt ); /* 4 or 6 ?*/
! 6842: }else{
! 6843: fprintf(ficgp,"$%d==%d && $%d==%d && ",kl+1, Tvaraff[k],kl+1+1,nbcode[Tvaraff[k]][lv]);
! 6844: kl++;
! 6845: }
! 6846: } /* end covariate */
! 6847: } /* end if no covariate */
! 6848:
1.238 brouard 6849: if(backcast==1){ /* We need to get the corresponding values of the covariates involved in this combination k1 */
6850: /* fprintf(ficgp,",\"%s\" every :::%d::%d u 1:($%d) t\"Backward stable prevalence\" w l lt 3",subdirf2(fileresu,"PLB_"),k1-1,k1-1,1+cpt); */
1.242 brouard 6851: fprintf(ficgp,",\"%s\" u 1:((",subdirf2(fileresu,"PLB_")); /* Age is in 1, nres in 2 to be fixed */
1.238 brouard 6852: if(cptcoveff ==0){
1.245 brouard 6853: fprintf(ficgp,"$%d)) t 'Backward prevalence in state %d' with line lt 3", 2+(cpt-1), cpt );
1.238 brouard 6854: }else{
6855: kl=0;
6856: for (k=1; k<=cptcoveff; k++){ /* For each combination of covariate */
6857: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate value corresponding to k1 combination and kth covariate */
6858: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6859: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6860: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6861: vlv= nbcode[Tvaraff[k]][lv];
1.223 brouard 6862: kl++;
1.238 brouard 6863: /* kl=6+(cpt-1)*(nlstate+1)+1+(i-1); /\* 6+(1-1)*(2+1)+1+(1-1)=7, 6+(2-1)(2+1)+1+(1-1)=10 *\/ */
6864: /*6+(cpt-1)*(nlstate+1)+1+(i-1)+(nlstate+1)*nlstate; 6+(1-1)*(2+1)+1+(1-1) +(2+1)*2=13 */
6865: /*6+1+(i-1)+(nlstate+1)*nlstate; 6+1+(1-1) +(2+1)*2=13 */
6866: /* '' u 6:(($1==1 && $2==0 && $3==2 && $4==0)? $9/(1.-$15) : 1/0):($5==2000? 3:2) t 'p.1' with line lc variable*/
6867: if(k==cptcoveff){
1.245 brouard 6868: fprintf(ficgp,"$%d==%d && $%d==%d)? $%d : 1/0) t 'Backward prevalence in state %d' w l lt 3",kl+1, Tvaraff[k],kl+1+1,nbcode[Tvaraff[k]][lv], \
1.242 brouard 6869: 2+cptcoveff*2+(cpt-1), cpt ); /* 4 or 6 ?*/
1.238 brouard 6870: }else{
6871: fprintf(ficgp,"$%d==%d && $%d==%d && ",kl+1, Tvaraff[k],kl+1+1,nbcode[Tvaraff[k]][lv]);
6872: kl++;
6873: }
6874: } /* end covariate */
6875: } /* end if no covariate */
6876: } /* end if backcast */
1.264 brouard 6877: fprintf(ficgp,"\nset out ;unset label;\n");
1.238 brouard 6878: } /* nres */
1.201 brouard 6879: } /* k1 */
6880: } /* cpt */
1.235 brouard 6881:
6882:
1.126 brouard 6883: /*2 eme*/
1.238 brouard 6884: for (k1=1; k1<= m ; k1 ++){
6885: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
1.253 brouard 6886: if(m != 1 && TKresult[nres]!= k1)
1.238 brouard 6887: continue;
6888: fprintf(ficgp,"\n# 2nd: Total life expectancy with CI: 't' files ");
1.264 brouard 6889: strcpy(gplotlabel,"(");
1.238 brouard 6890: for (k=1; k<=cptcoveff; k++){ /* For each covariate and each value */
1.225 brouard 6891: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate number corresponding to k1 combination */
1.223 brouard 6892: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6893: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6894: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6895: vlv= nbcode[Tvaraff[k]][lv];
6896: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
1.264 brouard 6897: sprintf(gplotlabel+strlen(gplotlabel)," V%d=%d ",Tvaraff[k],vlv);
1.211 brouard 6898: }
1.237 brouard 6899: /* for(k=1; k <= ncovds; k++){ */
1.236 brouard 6900: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
1.238 brouard 6901: printf(" V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
1.236 brouard 6902: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
1.264 brouard 6903: sprintf(gplotlabel+strlen(gplotlabel)," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
1.238 brouard 6904: }
1.264 brouard 6905: strcpy(gplotlabel+strlen(gplotlabel),")");
1.211 brouard 6906: fprintf(ficgp,"\n#\n");
1.223 brouard 6907: if(invalidvarcomb[k1]){
6908: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6909: continue;
6910: }
1.219 brouard 6911:
1.241 brouard 6912: fprintf(ficgp,"\nset out \"%s_%d-%d.svg\" \n",subdirf2(optionfilefiname,"E_"),k1,nres);
1.238 brouard 6913: for(vpopbased=0; vpopbased <= popbased; vpopbased++){ /* Done for vpopbased=0 and vpopbased=1 if popbased==1*/
1.264 brouard 6914: fprintf(ficgp,"\nset label \"popbased %d %s\" at graph 0.98,0.5 center rotate font \"Helvetica,12\"\n",vpopbased,gplotlabel);
6915: if(vpopbased==0){
1.238 brouard 6916: fprintf(ficgp,"set ylabel \"Years\" \nset ter svg size 640, 480\nplot [%.f:%.f] ",ageminpar,fage);
1.264 brouard 6917: }else
1.238 brouard 6918: fprintf(ficgp,"\nreplot ");
6919: for (i=1; i<= nlstate+1 ; i ++) {
6920: k=2*i;
1.261 brouard 6921: fprintf(ficgp,"\"%s\" every :::%d::%d u 1:($2==%d && $4!=0 ?$4 : 1/0) \"%%lf %%lf %%lf",subdirf2(fileresu,"T_"),nres-1,nres-1, vpopbased);
1.238 brouard 6922: for (j=1; j<= nlstate+1 ; j ++) {
6923: if (j==i) fprintf(ficgp," %%lf (%%lf)");
6924: else fprintf(ficgp," %%*lf (%%*lf)");
6925: }
6926: if (i== 1) fprintf(ficgp,"\" t\"TLE\" w l lt %d, \\\n",i);
6927: else fprintf(ficgp,"\" t\"LE in state (%d)\" w l lt %d, \\\n",i-1,i+1);
1.261 brouard 6928: fprintf(ficgp,"\"%s\" every :::%d::%d u 1:($2==%d && $4!=0 ? $4-$5*2 : 1/0) \"%%lf %%lf %%lf",subdirf2(fileresu,"T_"),nres-1,nres-1,vpopbased);
1.238 brouard 6929: for (j=1; j<= nlstate+1 ; j ++) {
6930: if (j==i) fprintf(ficgp," %%lf (%%lf)");
6931: else fprintf(ficgp," %%*lf (%%*lf)");
6932: }
6933: fprintf(ficgp,"\" t\"\" w l lt 0,");
1.261 brouard 6934: fprintf(ficgp,"\"%s\" every :::%d::%d u 1:($2==%d && $4!=0 ? $4+$5*2 : 1/0) \"%%lf %%lf %%lf",subdirf2(fileresu,"T_"),nres-1,nres-1,vpopbased);
1.238 brouard 6935: for (j=1; j<= nlstate+1 ; j ++) {
6936: if (j==i) fprintf(ficgp," %%lf (%%lf)");
6937: else fprintf(ficgp," %%*lf (%%*lf)");
6938: }
6939: if (i== (nlstate+1)) fprintf(ficgp,"\" t\"\" w l lt 0");
6940: else fprintf(ficgp,"\" t\"\" w l lt 0,\\\n");
6941: } /* state */
6942: } /* vpopbased */
1.264 brouard 6943: fprintf(ficgp,"\nset out;set out \"%s_%d-%d.svg\"; replot; set out; unset label;\n",subdirf2(optionfilefiname,"E_"),k1,nres); /* Buggy gnuplot */
1.238 brouard 6944: } /* end nres */
6945: } /* k1 end 2 eme*/
6946:
6947:
6948: /*3eme*/
6949: for (k1=1; k1<= m ; k1 ++){
6950: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
1.253 brouard 6951: if(m != 1 && TKresult[nres]!= k1)
1.238 brouard 6952: continue;
6953:
6954: for (cpt=1; cpt<= nlstate ; cpt ++) {
1.261 brouard 6955: fprintf(ficgp,"\n\n# 3d: Life expectancy with EXP_ files: combination=%d state=%d",k1, cpt);
1.264 brouard 6956: strcpy(gplotlabel,"(");
1.238 brouard 6957: for (k=1; k<=cptcoveff; k++){ /* For each covariate and each value */
6958: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate number corresponding to k1 combination */
6959: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6960: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6961: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6962: vlv= nbcode[Tvaraff[k]][lv];
6963: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
1.264 brouard 6964: sprintf(gplotlabel+strlen(gplotlabel)," V%d=%d ",Tvaraff[k],vlv);
1.238 brouard 6965: }
6966: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
6967: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
1.264 brouard 6968: sprintf(gplotlabel+strlen(gplotlabel)," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
1.238 brouard 6969: }
1.264 brouard 6970: strcpy(gplotlabel+strlen(gplotlabel),")");
1.238 brouard 6971: fprintf(ficgp,"\n#\n");
6972: if(invalidvarcomb[k1]){
6973: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6974: continue;
6975: }
6976:
6977: /* k=2+nlstate*(2*cpt-2); */
6978: k=2+(nlstate+1)*(cpt-1);
1.241 brouard 6979: fprintf(ficgp,"\nset out \"%s_%d-%d-%d.svg\" \n",subdirf2(optionfilefiname,"EXP_"),cpt,k1,nres);
1.264 brouard 6980: fprintf(ficgp,"set label \"%s\" at graph 0.98,0.5 center rotate font \"Helvetica,12\"\n",gplotlabel);
1.238 brouard 6981: fprintf(ficgp,"set ter svg size 640, 480\n\
1.261 brouard 6982: plot [%.f:%.f] \"%s\" every :::%d::%d u 1:%d t \"e%d1\" w l",ageminpar,fage,subdirf2(fileresu,"E_"),nres-1,nres-1,k,cpt);
1.238 brouard 6983: /*fprintf(ficgp,",\"e%s\" every :::%d::%d u 1:($%d-2*$%d) \"\%%lf ",fileres,k1-1,k1-1,k,k+1);
6984: for (i=1; i<= nlstate*2 ; i ++) fprintf(ficgp,"\%%lf (\%%lf) ");
6985: fprintf(ficgp,"\" t \"e%d1\" w l",cpt);
6986: fprintf(ficgp,",\"e%s\" every :::%d::%d u 1:($%d+2*$%d) \"\%%lf ",fileres,k1-1,k1-1,k,k+1);
6987: for (i=1; i<= nlstate*2 ; i ++) fprintf(ficgp,"\%%lf (\%%lf) ");
6988: fprintf(ficgp,"\" t \"e%d1\" w l",cpt);
1.219 brouard 6989:
1.238 brouard 6990: */
6991: for (i=1; i< nlstate ; i ++) {
1.261 brouard 6992: fprintf(ficgp," ,\"%s\" every :::%d::%d u 1:%d t \"e%d%d\" w l",subdirf2(fileresu,"E_"),nres-1,nres-1,k+i,cpt,i+1);
1.238 brouard 6993: /* fprintf(ficgp," ,\"%s\" every :::%d::%d u 1:%d t \"e%d%d\" w l",subdirf2(fileres,"e"),k1-1,k1-1,k+2*i,cpt,i+1);*/
1.219 brouard 6994:
1.238 brouard 6995: }
1.261 brouard 6996: fprintf(ficgp," ,\"%s\" every :::%d::%d u 1:%d t \"e%d.\" w l",subdirf2(fileresu,"E_"),nres-1,nres-1,k+nlstate,cpt);
1.238 brouard 6997: }
1.264 brouard 6998: fprintf(ficgp,"\nunset label;\n");
1.238 brouard 6999: } /* end nres */
7000: } /* end kl 3eme */
1.126 brouard 7001:
1.223 brouard 7002: /* 4eme */
1.201 brouard 7003: /* Survival functions (period) from state i in state j by initial state i */
1.238 brouard 7004: for (k1=1; k1<=m; k1++){ /* For each covariate and each value */
7005: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
1.253 brouard 7006: if(m != 1 && TKresult[nres]!= k1)
1.223 brouard 7007: continue;
1.238 brouard 7008: for (cpt=1; cpt<=nlstate ; cpt ++) { /* For each life state cpt*/
1.264 brouard 7009: strcpy(gplotlabel,"(");
1.238 brouard 7010: fprintf(ficgp,"\n#\n#\n# Survival functions in state j : 'LIJ_' files, cov=%d state=%d",k1, cpt);
7011: for (k=1; k<=cptcoveff; k++){ /* For each covariate and each value */
7012: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate number corresponding to k1 combination */
7013: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
7014: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
7015: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
7016: vlv= nbcode[Tvaraff[k]][lv];
7017: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
1.264 brouard 7018: sprintf(gplotlabel+strlen(gplotlabel)," V%d=%d ",Tvaraff[k],vlv);
1.238 brouard 7019: }
7020: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
7021: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
1.264 brouard 7022: sprintf(gplotlabel+strlen(gplotlabel)," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
1.238 brouard 7023: }
1.264 brouard 7024: strcpy(gplotlabel+strlen(gplotlabel),")");
1.238 brouard 7025: fprintf(ficgp,"\n#\n");
7026: if(invalidvarcomb[k1]){
7027: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
7028: continue;
1.223 brouard 7029: }
1.238 brouard 7030:
1.241 brouard 7031: fprintf(ficgp,"\nset out \"%s_%d-%d-%d.svg\" \n",subdirf2(optionfilefiname,"LIJ_"),cpt,k1,nres);
1.264 brouard 7032: fprintf(ficgp,"set label \"Alive state %d %s\" at graph 0.98,0.5 center rotate font \"Helvetica,12\"\n",cpt,gplotlabel);
1.238 brouard 7033: fprintf(ficgp,"set xlabel \"Age\" \nset ylabel \"Probability to be alive\" \n\
7034: set ter svg size 640, 480\nunset log y\nplot [%.f:%.f] ", ageminpar, agemaxpar);
7035: k=3;
7036: for (i=1; i<= nlstate ; i ++){
7037: if(i==1){
7038: fprintf(ficgp,"\"%s\"",subdirf2(fileresu,"PIJ_"));
7039: }else{
7040: fprintf(ficgp,", '' ");
7041: }
7042: l=(nlstate+ndeath)*(i-1)+1;
7043: fprintf(ficgp," u ($1==%d ? ($3):1/0):($%d/($%d",k1,k+l+(cpt-1),k+l);
7044: for (j=2; j<= nlstate+ndeath ; j ++)
7045: fprintf(ficgp,"+$%d",k+l+j-1);
7046: fprintf(ficgp,")) t \"l(%d,%d)\" w l",i,cpt);
7047: } /* nlstate */
1.264 brouard 7048: fprintf(ficgp,"\nset out; unset label;\n");
1.238 brouard 7049: } /* end cpt state*/
7050: } /* end nres */
7051: } /* end covariate k1 */
7052:
1.220 brouard 7053: /* 5eme */
1.201 brouard 7054: /* Survival functions (period) from state i in state j by final state j */
1.238 brouard 7055: for (k1=1; k1<= m ; k1++){ /* For each covariate combination if any */
7056: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
1.253 brouard 7057: if(m != 1 && TKresult[nres]!= k1)
1.227 brouard 7058: continue;
1.238 brouard 7059: for (cpt=1; cpt<=nlstate ; cpt ++) { /* For each inital state */
1.264 brouard 7060: strcpy(gplotlabel,"(");
1.238 brouard 7061: fprintf(ficgp,"\n#\n#\n# Survival functions in state j and all livestates from state i by final state j: 'lij' files, cov=%d state=%d",k1, cpt);
7062: for (k=1; k<=cptcoveff; k++){ /* For each covariate and each value */
7063: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate number corresponding to k1 combination */
7064: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
7065: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
7066: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
7067: vlv= nbcode[Tvaraff[k]][lv];
7068: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
1.264 brouard 7069: sprintf(gplotlabel+strlen(gplotlabel)," V%d=%d ",Tvaraff[k],vlv);
1.238 brouard 7070: }
7071: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
7072: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
1.264 brouard 7073: sprintf(gplotlabel+strlen(gplotlabel)," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
1.238 brouard 7074: }
1.264 brouard 7075: strcpy(gplotlabel+strlen(gplotlabel),")");
1.238 brouard 7076: fprintf(ficgp,"\n#\n");
7077: if(invalidvarcomb[k1]){
7078: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
7079: continue;
7080: }
1.227 brouard 7081:
1.241 brouard 7082: fprintf(ficgp,"\nset out \"%s_%d-%d-%d.svg\" \n",subdirf2(optionfilefiname,"LIJT_"),cpt,k1,nres);
1.264 brouard 7083: fprintf(ficgp,"set label \"Alive state %d %s\" at graph 0.98,0.5 center rotate font \"Helvetica,12\"\n",cpt,gplotlabel);
1.238 brouard 7084: fprintf(ficgp,"set xlabel \"Age\" \nset ylabel \"Probability to be alive\" \n\
7085: set ter svg size 640, 480\nunset log y\nplot [%.f:%.f] ", ageminpar, agemaxpar);
7086: k=3;
7087: for (j=1; j<= nlstate ; j ++){ /* Lived in state j */
7088: if(j==1)
7089: fprintf(ficgp,"\"%s\"",subdirf2(fileresu,"PIJ_"));
7090: else
7091: fprintf(ficgp,", '' ");
7092: l=(nlstate+ndeath)*(cpt-1) +j;
7093: fprintf(ficgp," u (($1==%d && (floor($2)%%5 == 0)) ? ($3):1/0):($%d",k1,k+l);
7094: /* for (i=2; i<= nlstate+ndeath ; i ++) */
7095: /* fprintf(ficgp,"+$%d",k+l+i-1); */
7096: fprintf(ficgp,") t \"l(%d,%d)\" w l",cpt,j);
7097: } /* nlstate */
7098: fprintf(ficgp,", '' ");
7099: fprintf(ficgp," u (($1==%d && (floor($2)%%5 == 0)) ? ($3):1/0):(",k1);
7100: for (j=1; j<= nlstate ; j ++){ /* Lived in state j */
7101: l=(nlstate+ndeath)*(cpt-1) +j;
7102: if(j < nlstate)
7103: fprintf(ficgp,"$%d +",k+l);
7104: else
7105: fprintf(ficgp,"$%d) t\"l(%d,.)\" w l",k+l,cpt);
7106: }
1.264 brouard 7107: fprintf(ficgp,"\nset out; unset label;\n");
1.238 brouard 7108: } /* end cpt state*/
7109: } /* end covariate */
7110: } /* end nres */
1.227 brouard 7111:
1.220 brouard 7112: /* 6eme */
1.202 brouard 7113: /* CV preval stable (period) for each covariate */
1.237 brouard 7114: for (k1=1; k1<= m ; k1 ++) /* For each covariate combination if any */
7115: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
1.253 brouard 7116: if(m != 1 && TKresult[nres]!= k1)
1.237 brouard 7117: continue;
1.255 brouard 7118: for (cpt=1; cpt<=nlstate ; cpt ++) { /* For each life state of arrival */
1.264 brouard 7119: strcpy(gplotlabel,"(");
1.211 brouard 7120: fprintf(ficgp,"\n#\n#\n#CV preval stable (period): 'pij' files, covariatecombination#=%d state=%d",k1, cpt);
1.225 brouard 7121: for (k=1; k<=cptcoveff; k++){ /* For each covariate and each value */
1.227 brouard 7122: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate number corresponding to k1 combination */
7123: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
7124: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
7125: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
7126: vlv= nbcode[Tvaraff[k]][lv];
7127: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
1.264 brouard 7128: sprintf(gplotlabel+strlen(gplotlabel)," V%d=%d ",Tvaraff[k],vlv);
1.211 brouard 7129: }
1.237 brouard 7130: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
7131: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
1.264 brouard 7132: sprintf(gplotlabel+strlen(gplotlabel)," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
1.237 brouard 7133: }
1.264 brouard 7134: strcpy(gplotlabel+strlen(gplotlabel),")");
1.211 brouard 7135: fprintf(ficgp,"\n#\n");
1.223 brouard 7136: if(invalidvarcomb[k1]){
1.227 brouard 7137: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
7138: continue;
1.223 brouard 7139: }
1.227 brouard 7140:
1.241 brouard 7141: fprintf(ficgp,"\nset out \"%s_%d-%d-%d.svg\" \n",subdirf2(optionfilefiname,"P_"),cpt,k1,nres);
1.264 brouard 7142: fprintf(ficgp,"set label \"Alive state %d %s\" at graph 0.98,0.5 center rotate font \"Helvetica,12\"\n",cpt,gplotlabel);
1.126 brouard 7143: fprintf(ficgp,"set xlabel \"Age\" \nset ylabel \"Probability\" \n\
1.238 brouard 7144: set ter svg size 640, 480\nunset log y\nplot [%.f:%.f] ", ageminpar, agemaxpar);
1.211 brouard 7145: k=3; /* Offset */
1.255 brouard 7146: for (i=1; i<= nlstate ; i ++){ /* State of origin */
1.227 brouard 7147: if(i==1)
7148: fprintf(ficgp,"\"%s\"",subdirf2(fileresu,"PIJ_"));
7149: else
7150: fprintf(ficgp,", '' ");
1.255 brouard 7151: l=(nlstate+ndeath)*(i-1)+1; /* 1, 1+ nlstate+ndeath, 1+2*(nlstate+ndeath) */
1.227 brouard 7152: fprintf(ficgp," u ($1==%d ? ($3):1/0):($%d/($%d",k1,k+l+(cpt-1),k+l);
7153: for (j=2; j<= nlstate ; j ++)
7154: fprintf(ficgp,"+$%d",k+l+j-1);
7155: fprintf(ficgp,")) t \"prev(%d,%d)\" w l",i,cpt);
1.153 brouard 7156: } /* nlstate */
1.264 brouard 7157: fprintf(ficgp,"\nset out; unset label;\n");
1.153 brouard 7158: } /* end cpt state*/
7159: } /* end covariate */
1.227 brouard 7160:
7161:
1.220 brouard 7162: /* 7eme */
1.218 brouard 7163: if(backcast == 1){
1.217 brouard 7164: /* CV back preval stable (period) for each covariate */
1.237 brouard 7165: for (k1=1; k1<= m ; k1 ++) /* For each covariate combination if any */
7166: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
1.253 brouard 7167: if(m != 1 && TKresult[nres]!= k1)
1.237 brouard 7168: continue;
1.255 brouard 7169: for (cpt=1; cpt<=nlstate ; cpt ++) { /* For each life ending state */
1.264 brouard 7170: strcpy(gplotlabel,"(");
7171: fprintf(ficgp,"\n#\n#\n#CV Back preval stable (period): 'pijb' files, covariatecombination#=%d state=%d",k1, cpt);
1.227 brouard 7172: for (k=1; k<=cptcoveff; k++){ /* For each covariate and each value */
7173: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate number corresponding to k1 combination */
7174: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
7175: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
1.223 brouard 7176: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
1.227 brouard 7177: vlv= nbcode[Tvaraff[k]][lv];
7178: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
1.264 brouard 7179: sprintf(gplotlabel+strlen(gplotlabel)," V%d=%d ",Tvaraff[k],vlv);
1.227 brouard 7180: }
1.237 brouard 7181: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
7182: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
1.264 brouard 7183: sprintf(gplotlabel+strlen(gplotlabel)," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
1.237 brouard 7184: }
1.264 brouard 7185: strcpy(gplotlabel+strlen(gplotlabel),")");
1.227 brouard 7186: fprintf(ficgp,"\n#\n");
7187: if(invalidvarcomb[k1]){
7188: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
7189: continue;
7190: }
7191:
1.241 brouard 7192: fprintf(ficgp,"\nset out \"%s_%d-%d-%d.svg\" \n",subdirf2(optionfilefiname,"PB_"),cpt,k1,nres);
1.264 brouard 7193: fprintf(ficgp,"set label \"Ending alive state %d %s\" at graph 0.98,0.5 center rotate font \"Helvetica,12\"\n",cpt,gplotlabel);
1.227 brouard 7194: fprintf(ficgp,"set xlabel \"Age\" \nset ylabel \"Probability\" \n\
1.238 brouard 7195: set ter svg size 640, 480\nunset log y\nplot [%.f:%.f] ", ageminpar, agemaxpar);
1.227 brouard 7196: k=3; /* Offset */
1.255 brouard 7197: for (i=1; i<= nlstate ; i ++){ /* State of origin */
1.227 brouard 7198: if(i==1)
7199: fprintf(ficgp,"\"%s\"",subdirf2(fileresu,"PIJB_"));
7200: else
7201: fprintf(ficgp,", '' ");
7202: /* l=(nlstate+ndeath)*(i-1)+1; */
1.255 brouard 7203: l=(nlstate+ndeath)*(cpt-1)+1; /* fixed for i; cpt=1 1, cpt=2 1+ nlstate+ndeath, 1+2*(nlstate+ndeath) */
1.227 brouard 7204: /* fprintf(ficgp," u ($1==%d ? ($3):1/0):($%d/($%d",k1,k+l+(cpt-1),k+l); /\* a vérifier *\/ */
7205: /* fprintf(ficgp," u ($1==%d ? ($3):1/0):($%d/($%d",k1,k+l+(cpt-1),k+l+(cpt-1)+i-1); /\* a vérifier *\/ */
1.255 brouard 7206: fprintf(ficgp," u ($1==%d ? ($3):1/0):($%d",k1,k+l+i-1); /* To be verified */
1.227 brouard 7207: /* for (j=2; j<= nlstate ; j ++) */
7208: /* fprintf(ficgp,"+$%d",k+l+j-1); */
7209: /* /\* fprintf(ficgp,"+$%d",k+l+j-1); *\/ */
7210: fprintf(ficgp,") t \"bprev(%d,%d)\" w l",i,cpt);
7211: } /* nlstate */
1.264 brouard 7212: fprintf(ficgp,"\nset out; unset label;\n");
1.218 brouard 7213: } /* end cpt state*/
7214: } /* end covariate */
7215: } /* End if backcast */
7216:
1.223 brouard 7217: /* 8eme */
1.218 brouard 7218: if(prevfcast==1){
7219: /* Projection from cross-sectional to stable (period) for each covariate */
7220:
1.237 brouard 7221: for (k1=1; k1<= m ; k1 ++) /* For each covariate combination if any */
7222: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
1.253 brouard 7223: if(m != 1 && TKresult[nres]!= k1)
1.237 brouard 7224: continue;
1.211 brouard 7225: for (cpt=1; cpt<=nlstate ; cpt ++) { /* For each life state */
1.264 brouard 7226: strcpy(gplotlabel,"(");
1.227 brouard 7227: fprintf(ficgp,"\n#\n#\n#Projection of prevalence to stable (period): 'PROJ_' files, covariatecombination#=%d state=%d",k1, cpt);
7228: for (k=1; k<=cptcoveff; k++){ /* For each correspondig covariate value */
7229: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate value corresponding to k1 combination and kth covariate */
7230: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
7231: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
7232: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
7233: vlv= nbcode[Tvaraff[k]][lv];
7234: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
1.264 brouard 7235: sprintf(gplotlabel+strlen(gplotlabel)," V%d=%d ",Tvaraff[k],vlv);
1.227 brouard 7236: }
1.237 brouard 7237: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
7238: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
1.264 brouard 7239: sprintf(gplotlabel+strlen(gplotlabel)," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
1.237 brouard 7240: }
1.264 brouard 7241: strcpy(gplotlabel+strlen(gplotlabel),")");
1.227 brouard 7242: fprintf(ficgp,"\n#\n");
7243: if(invalidvarcomb[k1]){
7244: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
7245: continue;
7246: }
7247:
7248: fprintf(ficgp,"# hpijx=probability over h years, hp.jx is weighted by observed prev\n ");
1.241 brouard 7249: fprintf(ficgp,"\nset out \"%s_%d-%d-%d.svg\" \n",subdirf2(optionfilefiname,"PROJ_"),cpt,k1,nres);
1.264 brouard 7250: fprintf(ficgp,"set label \"Alive state %d %s\" at graph 0.98,0.5 center rotate font \"Helvetica,12\"\n",cpt,gplotlabel);
1.227 brouard 7251: fprintf(ficgp,"set xlabel \"Age\" \nset ylabel \"Prevalence\" \n\
1.238 brouard 7252: set ter svg size 640, 480\nunset log y\nplot [%.f:%.f] ", ageminpar, agemaxpar);
1.227 brouard 7253: for (i=1; i<= nlstate+1 ; i ++){ /* nlstate +1 p11 p21 p.1 */
7254: /*# V1 = 1 V2 = 0 yearproj age p11 p21 p.1 p12 p22 p.2 p13 p23 p.3*/
7255: /*# 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 */
7256: /*# yearproj age p11 p21 p.1 p12 p22 p.2 p13 p23 p.3*/
7257: /*# 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 */
7258: if(i==1){
7259: fprintf(ficgp,"\"%s\"",subdirf2(fileresu,"F_"));
7260: }else{
7261: fprintf(ficgp,",\\\n '' ");
7262: }
7263: if(cptcoveff ==0){ /* No covariate */
7264: ioffset=2; /* Age is in 2 */
7265: /*# yearproj age p11 p21 p31 p.1 p12 p22 p32 p.2 p13 p23 p33 p.3 p14 p24 p34 p.4*/
7266: /*# 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 */
7267: /*# V1 = 1 yearproj age p11 p21 p31 p.1 p12 p22 p32 p.2 p13 p23 p33 p.3 p14 p24 p34 p.4*/
7268: /*# 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 */
7269: fprintf(ficgp," u %d:(", ioffset);
7270: if(i==nlstate+1)
7271: fprintf(ficgp," $%d/(1.-$%d)) t 'pw.%d' with line ", \
7272: ioffset+(cpt-1)*(nlstate+1)+1+(i-1), ioffset+1+(i-1)+(nlstate+1)*nlstate,cpt );
7273: else
7274: fprintf(ficgp," $%d/(1.-$%d)) t 'p%d%d' with line ", \
7275: ioffset+(cpt-1)*(nlstate+1)+1+(i-1), ioffset+1+(i-1)+(nlstate+1)*nlstate,i,cpt );
7276: }else{ /* more than 2 covariates */
7277: if(cptcoveff ==1){
7278: ioffset=4; /* Age is in 4 */
7279: }else{
7280: ioffset=6; /* Age is in 6 */
7281: /*# V1 = 1 V2 = 0 yearproj age p11 p21 p.1 p12 p22 p.2 p13 p23 p.3*/
7282: /*# 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 */
7283: }
7284: fprintf(ficgp," u %d:(",ioffset);
7285: kl=0;
7286: strcpy(gplotcondition,"(");
7287: for (k=1; k<=cptcoveff; k++){ /* For each covariate writing the chain of conditions */
7288: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate value corresponding to combination k1 and covariate k */
7289: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
7290: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
7291: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
7292: vlv= nbcode[Tvaraff[k]][lv]; /* Value of the modality of Tvaraff[k] */
7293: kl++;
7294: sprintf(gplotcondition+strlen(gplotcondition),"$%d==%d && $%d==%d " ,kl,Tvaraff[k], kl+1, nbcode[Tvaraff[k]][lv]);
7295: kl++;
7296: if(k <cptcoveff && cptcoveff>1)
7297: sprintf(gplotcondition+strlen(gplotcondition)," && ");
7298: }
7299: strcpy(gplotcondition+strlen(gplotcondition),")");
7300: /* kl=6+(cpt-1)*(nlstate+1)+1+(i-1); /\* 6+(1-1)*(2+1)+1+(1-1)=7, 6+(2-1)(2+1)+1+(1-1)=10 *\/ */
7301: /*6+(cpt-1)*(nlstate+1)+1+(i-1)+(nlstate+1)*nlstate; 6+(1-1)*(2+1)+1+(1-1) +(2+1)*2=13 */
7302: /*6+1+(i-1)+(nlstate+1)*nlstate; 6+1+(1-1) +(2+1)*2=13 */
7303: /* '' u 6:(($1==1 && $2==0 && $3==2 && $4==0)? $9/(1.-$15) : 1/0):($5==2000? 3:2) t 'p.1' with line lc variable*/
7304: if(i==nlstate+1){
7305: fprintf(ficgp,"%s ? $%d/(1.-$%d) : 1/0) t 'p.%d' with line ", gplotcondition, \
7306: ioffset+(cpt-1)*(nlstate+1)+1+(i-1), ioffset+1+(i-1)+(nlstate+1)*nlstate,cpt );
7307: }else{
7308: fprintf(ficgp,"%s ? $%d/(1.-$%d) : 1/0) t 'p%d%d' with line ", gplotcondition, \
7309: ioffset+(cpt-1)*(nlstate+1)+1+(i-1), ioffset +1+(i-1)+(nlstate+1)*nlstate,i,cpt );
7310: }
7311: } /* end if covariate */
7312: } /* nlstate */
1.264 brouard 7313: fprintf(ficgp,"\nset out; unset label;\n");
1.223 brouard 7314: } /* end cpt state*/
7315: } /* end covariate */
7316: } /* End if prevfcast */
1.227 brouard 7317:
7318:
1.238 brouard 7319: /* 9eme writing MLE parameters */
7320: fprintf(ficgp,"\n##############\n#9eme MLE estimated parameters\n#############\n");
1.126 brouard 7321: for(i=1,jk=1; i <=nlstate; i++){
1.187 brouard 7322: fprintf(ficgp,"# initial state %d\n",i);
1.126 brouard 7323: for(k=1; k <=(nlstate+ndeath); k++){
7324: if (k != i) {
1.227 brouard 7325: fprintf(ficgp,"# current state %d\n",k);
7326: for(j=1; j <=ncovmodel; j++){
7327: fprintf(ficgp,"p%d=%f; ",jk,p[jk]);
7328: jk++;
7329: }
7330: fprintf(ficgp,"\n");
1.126 brouard 7331: }
7332: }
1.223 brouard 7333: }
1.187 brouard 7334: fprintf(ficgp,"##############\n#\n");
1.227 brouard 7335:
1.145 brouard 7336: /*goto avoid;*/
1.238 brouard 7337: /* 10eme Graphics of probabilities or incidences using written MLE parameters */
7338: fprintf(ficgp,"\n##############\n#10eme Graphics of probabilities or incidences\n#############\n");
1.187 brouard 7339: fprintf(ficgp,"# logi(p12/p11)=a12+b12*age+c12age*age+d12*V1+e12*V1*age\n");
7340: fprintf(ficgp,"# logi(p12/p11)=p1 +p2*age +p3*age*age+ p4*V1+ p5*V1*age\n");
7341: fprintf(ficgp,"# logi(p13/p11)=a13+b13*age+c13age*age+d13*V1+e13*V1*age\n");
7342: fprintf(ficgp,"# logi(p13/p11)=p6 +p7*age +p8*age*age+ p9*V1+ p10*V1*age\n");
7343: fprintf(ficgp,"# p12+p13+p14+p11=1=p11(1+exp(a12+b12*age+c12age*age+d12*V1+e12*V1*age)\n");
7344: fprintf(ficgp,"# +exp(a13+b13*age+c13age*age+d13*V1+e13*V1*age)+...)\n");
7345: fprintf(ficgp,"# p11=1/(1+exp(a12+b12*age+c12age*age+d12*V1+e12*V1*age)\n");
7346: fprintf(ficgp,"# +exp(a13+b13*age+c13age*age+d13*V1+e13*V1*age)+...)\n");
7347: fprintf(ficgp,"# p12=exp(a12+b12*age+c12age*age+d12*V1+e12*V1*age)/\n");
7348: fprintf(ficgp,"# (1+exp(a12+b12*age+c12age*age+d12*V1+e12*V1*age)\n");
7349: fprintf(ficgp,"# +exp(a13+b13*age+c13age*age+d13*V1+e13*V1*age))\n");
7350: fprintf(ficgp,"# +exp(a14+b14*age+c14age*age+d14*V1+e14*V1*age)+...)\n");
7351: fprintf(ficgp,"#\n");
1.223 brouard 7352: for(ng=1; ng<=3;ng++){ /* Number of graphics: first is logit, 2nd is probabilities, third is incidences per year*/
1.238 brouard 7353: fprintf(ficgp,"#Number of graphics: first is logit, 2nd is probabilities, third is incidences per year\n");
1.237 brouard 7354: fprintf(ficgp,"#model=%s \n",model);
1.238 brouard 7355: fprintf(ficgp,"# Type of graphic ng=%d\n",ng);
1.264 brouard 7356: fprintf(ficgp,"# k1=1 to 2^%d=%d\n",cptcoveff,m);/* to be checked */
7357: for(k1=1; k1 <=m; k1++) /* For each combination of covariate */
1.237 brouard 7358: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
1.264 brouard 7359: if(m != 1 && TKresult[nres]!= k1)
1.237 brouard 7360: continue;
1.264 brouard 7361: fprintf(ficgp,"\n\n# Combination of dummy k1=%d which is ",k1);
7362: strcpy(gplotlabel,"(");
7363: sprintf(gplotlabel+strlen(gplotlabel)," Dummy combination %d ",k1);
7364: for (k=1; k<=cptcoveff; k++){ /* For each correspondig covariate value */
7365: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate value corresponding to k1 combination and kth covariate */
7366: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
7367: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
7368: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
7369: vlv= nbcode[Tvaraff[k]][lv];
7370: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
7371: sprintf(gplotlabel+strlen(gplotlabel)," V%d=%d ",Tvaraff[k],vlv);
7372: }
1.237 brouard 7373: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
7374: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
1.264 brouard 7375: sprintf(gplotlabel+strlen(gplotlabel)," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
1.237 brouard 7376: }
1.264 brouard 7377: strcpy(gplotlabel+strlen(gplotlabel),")");
1.237 brouard 7378: fprintf(ficgp,"\n#\n");
1.264 brouard 7379: fprintf(ficgp,"\nset out \"%s_%d-%d-%d.svg\" ",subdirf2(optionfilefiname,"PE_"),k1,ng,nres);
7380: fprintf(ficgp,"\nset label \"%s\" at graph 0.98,0.5 center rotate font \"Helvetica,12\"\n",gplotlabel);
1.223 brouard 7381: fprintf(ficgp,"\nset ter svg size 640, 480 ");
7382: if (ng==1){
7383: fprintf(ficgp,"\nset ylabel \"Value of the logit of the model\"\n"); /* exp(a12+b12*x) could be nice */
7384: fprintf(ficgp,"\nunset log y");
7385: }else if (ng==2){
7386: fprintf(ficgp,"\nset ylabel \"Probability\"\n");
7387: fprintf(ficgp,"\nset log y");
7388: }else if (ng==3){
7389: fprintf(ficgp,"\nset ylabel \"Quasi-incidence per year\"\n");
7390: fprintf(ficgp,"\nset log y");
7391: }else
7392: fprintf(ficgp,"\nunset title ");
7393: fprintf(ficgp,"\nplot [%.f:%.f] ",ageminpar,agemaxpar);
7394: i=1;
7395: for(k2=1; k2<=nlstate; k2++) {
7396: k3=i;
7397: for(k=1; k<=(nlstate+ndeath); k++) {
7398: if (k != k2){
7399: switch( ng) {
7400: case 1:
7401: if(nagesqr==0)
7402: fprintf(ficgp," p%d+p%d*x",i,i+1);
7403: else /* nagesqr =1 */
7404: fprintf(ficgp," p%d+p%d*x+p%d*x*x",i,i+1,i+1+nagesqr);
7405: break;
7406: case 2: /* ng=2 */
7407: if(nagesqr==0)
7408: fprintf(ficgp," exp(p%d+p%d*x",i,i+1);
7409: else /* nagesqr =1 */
7410: fprintf(ficgp," exp(p%d+p%d*x+p%d*x*x",i,i+1,i+1+nagesqr);
7411: break;
7412: case 3:
7413: if(nagesqr==0)
7414: fprintf(ficgp," %f*exp(p%d+p%d*x",YEARM/stepm,i,i+1);
7415: else /* nagesqr =1 */
7416: fprintf(ficgp," %f*exp(p%d+p%d*x+p%d*x*x",YEARM/stepm,i,i+1,i+1+nagesqr);
7417: break;
7418: }
7419: ij=1;/* To be checked else nbcode[0][0] wrong */
1.237 brouard 7420: ijp=1; /* product no age */
7421: /* for(j=3; j <=ncovmodel-nagesqr; j++) { */
7422: for(j=1; j <=cptcovt; j++) { /* For each covariate of the simplified model */
1.223 brouard 7423: /* printf("Tage[%d]=%d, j=%d\n", ij, Tage[ij], j); */
1.237 brouard 7424: if(j==Tage[ij]) { /* Product by age */
7425: if(ij <=cptcovage) { /* V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1, 2 V5 and V1 */
1.238 brouard 7426: if(DummyV[j]==0){
1.237 brouard 7427: fprintf(ficgp,"+p%d*%d*x",i+j+2+nagesqr-1,Tinvresult[nres][Tvar[j]]);;
7428: }else{ /* quantitative */
7429: fprintf(ficgp,"+p%d*%f*x",i+j+2+nagesqr-1,Tqinvresult[nres][Tvar[j]]); /* Tqinvresult in decoderesult */
1.264 brouard 7430: /* fprintf(ficgp,"+p%d*%d*x",i+j+nagesqr-1,nbcode[Tvar[j-2]][codtabm(k1,Tvar[j-2])]); */
1.237 brouard 7431: }
7432: ij++;
7433: }
7434: }else if(j==Tprod[ijp]) { /* */
7435: /* printf("Tprod[%d]=%d, j=%d\n", ij, Tprod[ijp], j); */
7436: if(ijp <=cptcovprod) { /* Product */
1.238 brouard 7437: if(DummyV[Tvard[ijp][1]]==0){/* Vn is dummy */
7438: if(DummyV[Tvard[ijp][2]]==0){/* Vn and Vm are dummy */
1.264 brouard 7439: /* fprintf(ficgp,"+p%d*%d*%d",i+j+2+nagesqr-1,nbcode[Tvard[ijp][1]][codtabm(k1,j)],nbcode[Tvard[ijp][2]][codtabm(k1,j)]); */
1.237 brouard 7440: fprintf(ficgp,"+p%d*%d*%d",i+j+2+nagesqr-1,Tinvresult[nres][Tvard[ijp][1]],Tinvresult[nres][Tvard[ijp][2]]);
7441: }else{ /* Vn is dummy and Vm is quanti */
1.264 brouard 7442: /* fprintf(ficgp,"+p%d*%d*%f",i+j+2+nagesqr-1,nbcode[Tvard[ijp][1]][codtabm(k1,j)],Tqinvresult[nres][Tvard[ijp][2]]); */
1.237 brouard 7443: fprintf(ficgp,"+p%d*%d*%f",i+j+2+nagesqr-1,Tinvresult[nres][Tvard[ijp][1]],Tqinvresult[nres][Tvard[ijp][2]]);
7444: }
7445: }else{ /* Vn*Vm Vn is quanti */
1.238 brouard 7446: if(DummyV[Tvard[ijp][2]]==0){
1.237 brouard 7447: fprintf(ficgp,"+p%d*%d*%f",i+j+2+nagesqr-1,Tinvresult[nres][Tvard[ijp][2]],Tqinvresult[nres][Tvard[ijp][1]]);
7448: }else{ /* Both quanti */
7449: fprintf(ficgp,"+p%d*%f*%f",i+j+2+nagesqr-1,Tqinvresult[nres][Tvard[ijp][1]],Tqinvresult[nres][Tvard[ijp][2]]);
7450: }
7451: }
1.238 brouard 7452: ijp++;
1.237 brouard 7453: }
7454: } else{ /* simple covariate */
1.264 brouard 7455: /* fprintf(ficgp,"+p%d*%d",i+j+2+nagesqr-1,nbcode[Tvar[j]][codtabm(k1,j)]); /\* Valgrind bug nbcode *\/ */
1.237 brouard 7456: if(Dummy[j]==0){
7457: fprintf(ficgp,"+p%d*%d",i+j+2+nagesqr-1,Tinvresult[nres][Tvar[j]]); /* */
7458: }else{ /* quantitative */
7459: fprintf(ficgp,"+p%d*%f",i+j+2+nagesqr-1,Tqinvresult[nres][Tvar[j]]); /* */
1.264 brouard 7460: /* fprintf(ficgp,"+p%d*%d*x",i+j+nagesqr-1,nbcode[Tvar[j-2]][codtabm(k1,Tvar[j-2])]); */
1.223 brouard 7461: }
1.237 brouard 7462: } /* end simple */
7463: } /* end j */
1.223 brouard 7464: }else{
7465: i=i-ncovmodel;
7466: if(ng !=1 ) /* For logit formula of log p11 is more difficult to get */
7467: fprintf(ficgp," (1.");
7468: }
1.227 brouard 7469:
1.223 brouard 7470: if(ng != 1){
7471: fprintf(ficgp,")/(1");
1.227 brouard 7472:
1.264 brouard 7473: for(cpt=1; cpt <=nlstate; cpt++){
1.223 brouard 7474: if(nagesqr==0)
1.264 brouard 7475: fprintf(ficgp,"+exp(p%d+p%d*x",k3+(cpt-1)*ncovmodel,k3+(cpt-1)*ncovmodel+1);
1.223 brouard 7476: else /* nagesqr =1 */
1.264 brouard 7477: fprintf(ficgp,"+exp(p%d+p%d*x+p%d*x*x",k3+(cpt-1)*ncovmodel,k3+(cpt-1)*ncovmodel+1,k3+(cpt-1)*ncovmodel+1+nagesqr);
1.217 brouard 7478:
1.223 brouard 7479: ij=1;
7480: for(j=3; j <=ncovmodel-nagesqr; j++){
1.237 brouard 7481: if((j-2)==Tage[ij]) { /* Bug valgrind */
7482: if(ij <=cptcovage) { /* Bug valgrind */
1.264 brouard 7483: fprintf(ficgp,"+p%d*%d*x",k3+(cpt-1)*ncovmodel+1+j-2+nagesqr,nbcode[Tvar[j-2]][codtabm(k1,j-2)]);
7484: /* fprintf(ficgp,"+p%d*%d*x",k3+(cpt-1)*ncovmodel+1+j-2+nagesqr,nbcode[Tvar[j-2]][codtabm(k1,Tvar[j-2])]); */
1.223 brouard 7485: ij++;
7486: }
7487: }
7488: else
1.264 brouard 7489: fprintf(ficgp,"+p%d*%d",k3+(cpt-1)*ncovmodel+1+j-2+nagesqr,nbcode[Tvar[j-2]][codtabm(k1,j-2)]);/* Valgrind bug nbcode */
1.223 brouard 7490: }
7491: fprintf(ficgp,")");
7492: }
7493: fprintf(ficgp,")");
7494: if(ng ==2)
7495: fprintf(ficgp," t \"p%d%d\" ", k2,k);
7496: else /* ng= 3 */
7497: fprintf(ficgp," t \"i%d%d\" ", k2,k);
7498: }else{ /* end ng <> 1 */
7499: if( k !=k2) /* logit p11 is hard to draw */
7500: fprintf(ficgp," t \"logit(p%d%d)\" ", k2,k);
7501: }
7502: if ((k+k2)!= (nlstate*2+ndeath) && ng != 1)
7503: fprintf(ficgp,",");
7504: if (ng == 1 && k!=k2 && (k+k2)!= (nlstate*2+ndeath))
7505: fprintf(ficgp,",");
7506: i=i+ncovmodel;
7507: } /* end k */
7508: } /* end k2 */
1.264 brouard 7509: fprintf(ficgp,"\n set out; unset label;\n");
7510: } /* end k1 */
1.223 brouard 7511: } /* end ng */
7512: /* avoid: */
7513: fflush(ficgp);
1.126 brouard 7514: } /* end gnuplot */
7515:
7516:
7517: /*************** Moving average **************/
1.219 brouard 7518: /* int movingaverage(double ***probs, double bage, double fage, double ***mobaverage, int mobilav, double bageout, double fageout){ */
1.222 brouard 7519: int movingaverage(double ***probs, double bage, double fage, double ***mobaverage, int mobilav){
1.218 brouard 7520:
1.222 brouard 7521: int i, cpt, cptcod;
7522: int modcovmax =1;
7523: int mobilavrange, mob;
7524: int iage=0;
7525:
7526: double sum=0.;
7527: double age;
7528: double *sumnewp, *sumnewm;
7529: double *agemingood, *agemaxgood; /* Currently identical for all covariates */
7530:
7531:
1.225 brouard 7532: /* modcovmax=2*cptcoveff;/\* Max number of modalities. We suppose */
1.222 brouard 7533: /* a covariate has 2 modalities, should be equal to ncovcombmax *\/ */
7534:
7535: sumnewp = vector(1,ncovcombmax);
7536: sumnewm = vector(1,ncovcombmax);
7537: agemingood = vector(1,ncovcombmax);
7538: agemaxgood = vector(1,ncovcombmax);
7539:
7540: for (cptcod=1;cptcod<=ncovcombmax;cptcod++){
7541: sumnewm[cptcod]=0.;
7542: sumnewp[cptcod]=0.;
7543: agemingood[cptcod]=0;
7544: agemaxgood[cptcod]=0;
7545: }
7546: if (cptcovn<1) ncovcombmax=1; /* At least 1 pass */
7547:
7548: if(mobilav==1||mobilav ==3 ||mobilav==5 ||mobilav== 7){
7549: if(mobilav==1) mobilavrange=5; /* default */
7550: else mobilavrange=mobilav;
7551: for (age=bage; age<=fage; age++)
7552: for (i=1; i<=nlstate;i++)
7553: for (cptcod=1;cptcod<=ncovcombmax;cptcod++)
7554: mobaverage[(int)age][i][cptcod]=probs[(int)age][i][cptcod];
7555: /* We keep the original values on the extreme ages bage, fage and for
7556: fage+1 and bage-1 we use a 3 terms moving average; for fage+2 bage+2
7557: we use a 5 terms etc. until the borders are no more concerned.
7558: */
7559: for (mob=3;mob <=mobilavrange;mob=mob+2){
7560: for (age=bage+(mob-1)/2; age<=fage-(mob-1)/2; age++){
7561: for (i=1; i<=nlstate;i++){
7562: for (cptcod=1;cptcod<=ncovcombmax;cptcod++){
7563: mobaverage[(int)age][i][cptcod] =probs[(int)age][i][cptcod];
7564: for (cpt=1;cpt<=(mob-1)/2;cpt++){
7565: mobaverage[(int)age][i][cptcod] +=probs[(int)age-cpt][i][cptcod];
7566: mobaverage[(int)age][i][cptcod] +=probs[(int)age+cpt][i][cptcod];
7567: }
7568: mobaverage[(int)age][i][cptcod]=mobaverage[(int)age][i][cptcod]/mob;
7569: }
7570: }
7571: }/* end age */
7572: }/* end mob */
7573: }else
7574: return -1;
7575: for (cptcod=1;cptcod<=ncovcombmax;cptcod++){
7576: /* for (age=bage+(mob-1)/2; age<=fage-(mob-1)/2; age++){ */
7577: if(invalidvarcomb[cptcod]){
7578: printf("\nCombination (%d) ignored because no cases \n",cptcod);
7579: continue;
7580: }
1.219 brouard 7581:
1.222 brouard 7582: agemingood[cptcod]=fage-(mob-1)/2;
7583: for (age=fage-(mob-1)/2; age>=bage; age--){/* From oldest to youngest, finding the youngest wrong */
7584: sumnewm[cptcod]=0.;
7585: for (i=1; i<=nlstate;i++){
7586: sumnewm[cptcod]+=mobaverage[(int)age][i][cptcod];
7587: }
7588: if(fabs(sumnewm[cptcod] - 1.) <= 1.e-3) { /* good */
7589: agemingood[cptcod]=age;
7590: }else{ /* bad */
7591: for (i=1; i<=nlstate;i++){
7592: mobaverage[(int)age][i][cptcod]=mobaverage[(int)agemingood[cptcod]][i][cptcod];
7593: } /* i */
7594: } /* end bad */
7595: }/* age */
7596: sum=0.;
7597: for (i=1; i<=nlstate;i++){
7598: sum+=mobaverage[(int)agemingood[cptcod]][i][cptcod];
7599: }
7600: if(fabs(sum - 1.) > 1.e-3) { /* bad */
7601: printf("For this combination of covariate cptcod=%d, we can't get a smoothed prevalence which sums to one at any descending age!\n",cptcod);
7602: /* for (i=1; i<=nlstate;i++){ */
7603: /* mobaverage[(int)age][i][cptcod]=mobaverage[(int)agemingood[cptcod]][i][cptcod]; */
7604: /* } /\* i *\/ */
7605: } /* end bad */
7606: /* else{ /\* We found some ages summing to one, we will smooth the oldest *\/ */
7607: /* From youngest, finding the oldest wrong */
7608: agemaxgood[cptcod]=bage+(mob-1)/2;
7609: for (age=bage+(mob-1)/2; age<=fage; age++){
7610: sumnewm[cptcod]=0.;
7611: for (i=1; i<=nlstate;i++){
7612: sumnewm[cptcod]+=mobaverage[(int)age][i][cptcod];
7613: }
7614: if(fabs(sumnewm[cptcod] - 1.) <= 1.e-3) { /* good */
7615: agemaxgood[cptcod]=age;
7616: }else{ /* bad */
7617: for (i=1; i<=nlstate;i++){
7618: mobaverage[(int)age][i][cptcod]=mobaverage[(int)agemaxgood[cptcod]][i][cptcod];
7619: } /* i */
7620: } /* end bad */
7621: }/* age */
7622: sum=0.;
7623: for (i=1; i<=nlstate;i++){
7624: sum+=mobaverage[(int)agemaxgood[cptcod]][i][cptcod];
7625: }
7626: if(fabs(sum - 1.) > 1.e-3) { /* bad */
7627: printf("For this combination of covariate cptcod=%d, we can't get a smoothed prevalence which sums to one at any ascending age!\n",cptcod);
7628: /* for (i=1; i<=nlstate;i++){ */
7629: /* mobaverage[(int)age][i][cptcod]=mobaverage[(int)agemingood[cptcod]][i][cptcod]; */
7630: /* } /\* i *\/ */
7631: } /* end bad */
7632:
7633: for (age=bage; age<=fage; age++){
1.235 brouard 7634: /* printf("%d %d ", cptcod, (int)age); */
1.222 brouard 7635: sumnewp[cptcod]=0.;
7636: sumnewm[cptcod]=0.;
7637: for (i=1; i<=nlstate;i++){
7638: sumnewp[cptcod]+=probs[(int)age][i][cptcod];
7639: sumnewm[cptcod]+=mobaverage[(int)age][i][cptcod];
7640: /* printf("%.4f %.4f ",probs[(int)age][i][cptcod], mobaverage[(int)age][i][cptcod]); */
7641: }
7642: /* printf("%.4f %.4f \n",sumnewp[cptcod], sumnewm[cptcod]); */
7643: }
7644: /* printf("\n"); */
7645: /* } */
7646: /* brutal averaging */
7647: for (i=1; i<=nlstate;i++){
7648: for (age=1; age<=bage; age++){
7649: mobaverage[(int)age][i][cptcod]=mobaverage[(int)agemingood[cptcod]][i][cptcod];
7650: /* printf("age=%d i=%d cptcod=%d mobaverage=%.4f \n",(int)age,i, cptcod, mobaverage[(int)age][i][cptcod]); */
7651: }
7652: for (age=fage; age<=AGESUP; age++){
7653: mobaverage[(int)age][i][cptcod]=mobaverage[(int)agemaxgood[cptcod]][i][cptcod];
7654: /* printf("age=%d i=%d cptcod=%d mobaverage=%.4f \n",(int)age,i, cptcod, mobaverage[(int)age][i][cptcod]); */
7655: }
7656: } /* end i status */
7657: for (i=nlstate+1; i<=nlstate+ndeath;i++){
7658: for (age=1; age<=AGESUP; age++){
7659: /*printf("i=%d, age=%d, cptcod=%d\n",i, (int)age, cptcod);*/
7660: mobaverage[(int)age][i][cptcod]=0.;
7661: }
7662: }
7663: }/* end cptcod */
7664: free_vector(sumnewm,1, ncovcombmax);
7665: free_vector(sumnewp,1, ncovcombmax);
7666: free_vector(agemaxgood,1, ncovcombmax);
7667: free_vector(agemingood,1, ncovcombmax);
7668: return 0;
7669: }/* End movingaverage */
1.218 brouard 7670:
1.126 brouard 7671:
7672: /************** Forecasting ******************/
1.235 brouard 7673: void prevforecast(char fileres[], double anproj1, double mproj1, double jproj1, double ageminpar, double agemax, double dateprev1, double dateprev2, int mobilav, double bage, double fage, int firstpass, int lastpass, double anproj2, double p[], int cptcoveff){
1.126 brouard 7674: /* proj1, year, month, day of starting projection
7675: agemin, agemax range of age
7676: dateprev1 dateprev2 range of dates during which prevalence is computed
7677: anproj2 year of en of projection (same day and month as proj1).
7678: */
1.235 brouard 7679: int yearp, stepsize, hstepm, nhstepm, j, k, cptcod, i, h, i1, k4, nres=0;
1.126 brouard 7680: double agec; /* generic age */
7681: double agelim, ppij, yp,yp1,yp2,jprojmean,mprojmean,anprojmean;
7682: double *popeffectif,*popcount;
7683: double ***p3mat;
1.218 brouard 7684: /* double ***mobaverage; */
1.126 brouard 7685: char fileresf[FILENAMELENGTH];
7686:
7687: agelim=AGESUP;
1.211 brouard 7688: /* Compute observed prevalence between dateprev1 and dateprev2 by counting the number of people
7689: in each health status at the date of interview (if between dateprev1 and dateprev2).
7690: We still use firstpass and lastpass as another selection.
7691: */
1.214 brouard 7692: /* freqsummary(fileres, agemin, agemax, s, agev, nlstate, imx,Tvaraff,nbcode, ncodemax,mint,anint,strstart,\ */
7693: /* firstpass, lastpass, stepm, weightopt, model); */
1.126 brouard 7694:
1.201 brouard 7695: strcpy(fileresf,"F_");
7696: strcat(fileresf,fileresu);
1.126 brouard 7697: if((ficresf=fopen(fileresf,"w"))==NULL) {
7698: printf("Problem with forecast resultfile: %s\n", fileresf);
7699: fprintf(ficlog,"Problem with forecast resultfile: %s\n", fileresf);
7700: }
1.235 brouard 7701: printf("\nComputing forecasting: result on file '%s', please wait... \n", fileresf);
7702: fprintf(ficlog,"\nComputing forecasting: result on file '%s', please wait... \n", fileresf);
1.126 brouard 7703:
1.225 brouard 7704: if (cptcoveff==0) ncodemax[cptcoveff]=1;
1.126 brouard 7705:
7706:
7707: stepsize=(int) (stepm+YEARM-1)/YEARM;
7708: if (stepm<=12) stepsize=1;
7709: if(estepm < stepm){
7710: printf ("Problem %d lower than %d\n",estepm, stepm);
7711: }
7712: else hstepm=estepm;
7713:
7714: hstepm=hstepm/stepm;
7715: yp1=modf(dateintmean,&yp);/* extracts integral of datemean in yp and
7716: fractional in yp1 */
7717: anprojmean=yp;
7718: yp2=modf((yp1*12),&yp);
7719: mprojmean=yp;
7720: yp1=modf((yp2*30.5),&yp);
7721: jprojmean=yp;
7722: if(jprojmean==0) jprojmean=1;
7723: if(mprojmean==0) jprojmean=1;
7724:
1.227 brouard 7725: i1=pow(2,cptcoveff);
1.126 brouard 7726: if (cptcovn < 1){i1=1;}
7727:
7728: fprintf(ficresf,"# Mean day of interviews %.lf/%.lf/%.lf (%.2f) between %.2f and %.2f \n",jprojmean,mprojmean,anprojmean,dateintmean,dateprev1,dateprev2);
7729:
7730: fprintf(ficresf,"#****** Routine prevforecast **\n");
1.227 brouard 7731:
1.126 brouard 7732: /* if (h==(int)(YEARM*yearp)){ */
1.235 brouard 7733: for(nres=1; nres <= nresult; nres++) /* For each resultline */
7734: for(k=1; k<=i1;k++){
1.253 brouard 7735: if(i1 != 1 && TKresult[nres]!= k)
1.235 brouard 7736: continue;
1.227 brouard 7737: if(invalidvarcomb[k]){
7738: printf("\nCombination (%d) projection ignored because no cases \n",k);
7739: continue;
7740: }
7741: fprintf(ficresf,"\n#****** hpijx=probability over h years, hp.jx is weighted by observed prev \n#");
7742: for(j=1;j<=cptcoveff;j++) {
7743: fprintf(ficresf," V%d (=) %d",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
7744: }
1.235 brouard 7745: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
1.238 brouard 7746: fprintf(ficresf," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
1.235 brouard 7747: }
1.227 brouard 7748: fprintf(ficresf," yearproj age");
7749: for(j=1; j<=nlstate+ndeath;j++){
7750: for(i=1; i<=nlstate;i++)
7751: fprintf(ficresf," p%d%d",i,j);
7752: fprintf(ficresf," wp.%d",j);
7753: }
7754: for (yearp=0; yearp<=(anproj2-anproj1);yearp +=stepsize) {
7755: fprintf(ficresf,"\n");
7756: fprintf(ficresf,"\n# Forecasting at date %.lf/%.lf/%.lf ",jproj1,mproj1,anproj1+yearp);
7757: for (agec=fage; agec>=(ageminpar-1); agec--){
7758: nhstepm=(int) rint((agelim-agec)*YEARM/stepm);
7759: nhstepm = nhstepm/hstepm;
7760: p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
7761: oldm=oldms;savm=savms;
1.235 brouard 7762: hpxij(p3mat,nhstepm,agec,hstepm,p,nlstate,stepm,oldm,savm, k,nres);
1.227 brouard 7763:
7764: for (h=0; h<=nhstepm; h++){
7765: if (h*hstepm/YEARM*stepm ==yearp) {
7766: fprintf(ficresf,"\n");
7767: for(j=1;j<=cptcoveff;j++)
7768: fprintf(ficresf,"%d %d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
7769: fprintf(ficresf,"%.f %.f ",anproj1+yearp,agec+h*hstepm/YEARM*stepm);
7770: }
7771: for(j=1; j<=nlstate+ndeath;j++) {
7772: ppij=0.;
7773: for(i=1; i<=nlstate;i++) {
7774: if (mobilav==1)
7775: ppij=ppij+p3mat[i][j][h]*mobaverage[(int)agec][i][k];
7776: else {
7777: ppij=ppij+p3mat[i][j][h]*probs[(int)(agec)][i][k];
7778: }
7779: if (h*hstepm/YEARM*stepm== yearp) {
7780: fprintf(ficresf," %.3f", p3mat[i][j][h]);
7781: }
7782: } /* end i */
7783: if (h*hstepm/YEARM*stepm==yearp) {
7784: fprintf(ficresf," %.3f", ppij);
7785: }
7786: }/* end j */
7787: } /* end h */
7788: free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
7789: } /* end agec */
7790: } /* end yearp */
7791: } /* end k */
1.219 brouard 7792:
1.126 brouard 7793: fclose(ficresf);
1.215 brouard 7794: printf("End of Computing forecasting \n");
7795: fprintf(ficlog,"End of Computing forecasting\n");
7796:
1.126 brouard 7797: }
7798:
1.218 brouard 7799: /* /\************** Back Forecasting ******************\/ */
1.225 brouard 7800: /* void prevbackforecast(char fileres[], double anback1, double mback1, double jback1, double ageminpar, double agemax, double dateprev1, double dateprev2, int mobilav, double bage, double fage, int firstpass, int lastpass, double anback2, double p[], int cptcoveff){ */
1.218 brouard 7801: /* /\* back1, year, month, day of starting backection */
7802: /* agemin, agemax range of age */
7803: /* dateprev1 dateprev2 range of dates during which prevalence is computed */
7804: /* anback2 year of en of backection (same day and month as back1). */
7805: /* *\/ */
7806: /* int yearp, stepsize, hstepm, nhstepm, j, k, cptcod, i, h, i1; */
7807: /* double agec; /\* generic age *\/ */
7808: /* double agelim, ppij, yp,yp1,yp2,jprojmean,mprojmean,anprojmean; */
7809: /* double *popeffectif,*popcount; */
7810: /* double ***p3mat; */
7811: /* /\* double ***mobaverage; *\/ */
7812: /* char fileresfb[FILENAMELENGTH]; */
7813:
7814: /* agelim=AGESUP; */
7815: /* /\* Compute observed prevalence between dateprev1 and dateprev2 by counting the number of people */
7816: /* in each health status at the date of interview (if between dateprev1 and dateprev2). */
7817: /* We still use firstpass and lastpass as another selection. */
7818: /* *\/ */
7819: /* /\* freqsummary(fileres, agemin, agemax, s, agev, nlstate, imx,Tvaraff,nbcode, ncodemax,mint,anint,strstart,\ *\/ */
7820: /* /\* firstpass, lastpass, stepm, weightopt, model); *\/ */
7821: /* prevalence(probs, ageminpar, agemax, s, agev, nlstate, imx, Tvar, nbcode, ncodemax, mint, anint, dateprev1, dateprev2, firstpass, lastpass); */
7822:
7823: /* strcpy(fileresfb,"FB_"); */
7824: /* strcat(fileresfb,fileresu); */
7825: /* if((ficresfb=fopen(fileresfb,"w"))==NULL) { */
7826: /* printf("Problem with back forecast resultfile: %s\n", fileresfb); */
7827: /* fprintf(ficlog,"Problem with back forecast resultfile: %s\n", fileresfb); */
7828: /* } */
7829: /* printf("Computing back forecasting: result on file '%s', please wait... \n", fileresfb); */
7830: /* fprintf(ficlog,"Computing back forecasting: result on file '%s', please wait... \n", fileresfb); */
7831:
1.225 brouard 7832: /* if (cptcoveff==0) ncodemax[cptcoveff]=1; */
1.218 brouard 7833:
7834: /* /\* if (mobilav!=0) { *\/ */
7835: /* /\* mobaverage= ma3x(1, AGESUP,1,NCOVMAX, 1,NCOVMAX); *\/ */
7836: /* /\* if (movingaverage(probs, ageminpar, fage, mobaverage,mobilav)!=0){ *\/ */
7837: /* /\* fprintf(ficlog," Error in movingaverage mobilav=%d\n",mobilav); *\/ */
7838: /* /\* printf(" Error in movingaverage mobilav=%d\n",mobilav); *\/ */
7839: /* /\* } *\/ */
7840: /* /\* } *\/ */
7841:
7842: /* stepsize=(int) (stepm+YEARM-1)/YEARM; */
7843: /* if (stepm<=12) stepsize=1; */
7844: /* if(estepm < stepm){ */
7845: /* printf ("Problem %d lower than %d\n",estepm, stepm); */
7846: /* } */
7847: /* else hstepm=estepm; */
7848:
7849: /* hstepm=hstepm/stepm; */
7850: /* yp1=modf(dateintmean,&yp);/\* extracts integral of datemean in yp and */
7851: /* fractional in yp1 *\/ */
7852: /* anprojmean=yp; */
7853: /* yp2=modf((yp1*12),&yp); */
7854: /* mprojmean=yp; */
7855: /* yp1=modf((yp2*30.5),&yp); */
7856: /* jprojmean=yp; */
7857: /* if(jprojmean==0) jprojmean=1; */
7858: /* if(mprojmean==0) jprojmean=1; */
7859:
1.225 brouard 7860: /* i1=cptcoveff; */
1.218 brouard 7861: /* if (cptcovn < 1){i1=1;} */
1.217 brouard 7862:
1.218 brouard 7863: /* fprintf(ficresfb,"# Mean day of interviews %.lf/%.lf/%.lf (%.2f) between %.2f and %.2f \n",jprojmean,mprojmean,anprojmean,dateintmean,dateprev1,dateprev2); */
1.217 brouard 7864:
1.218 brouard 7865: /* fprintf(ficresfb,"#****** Routine prevbackforecast **\n"); */
7866:
7867: /* /\* if (h==(int)(YEARM*yearp)){ *\/ */
7868: /* for(cptcov=1, k=0;cptcov<=i1;cptcov++){ */
1.225 brouard 7869: /* for(cptcod=1;cptcod<=ncodemax[cptcoveff];cptcod++){ */
1.218 brouard 7870: /* k=k+1; */
7871: /* fprintf(ficresfb,"\n#****** hbijx=probability over h years, hp.jx is weighted by observed prev \n#"); */
1.225 brouard 7872: /* for(j=1;j<=cptcoveff;j++) { */
1.218 brouard 7873: /* fprintf(ficresfb," V%d (=) %d",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]); */
7874: /* } */
7875: /* fprintf(ficresfb," yearbproj age"); */
7876: /* for(j=1; j<=nlstate+ndeath;j++){ */
7877: /* for(i=1; i<=nlstate;i++) */
7878: /* fprintf(ficresfb," p%d%d",i,j); */
7879: /* fprintf(ficresfb," p.%d",j); */
7880: /* } */
7881: /* for (yearp=0; yearp>=(anback2-anback1);yearp -=stepsize) { */
7882: /* /\* for (yearp=0; yearp<=(anproj2-anproj1);yearp +=stepsize) { *\/ */
7883: /* fprintf(ficresfb,"\n"); */
7884: /* fprintf(ficresfb,"\n# Back Forecasting at date %.lf/%.lf/%.lf ",jback1,mback1,anback1+yearp); */
7885: /* for (agec=fage; agec>=(ageminpar-1); agec--){ */
7886: /* nhstepm=(int) rint((agelim-agec)*YEARM/stepm); */
7887: /* nhstepm = nhstepm/hstepm; */
7888: /* p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm); */
7889: /* oldm=oldms;savm=savms; */
7890: /* hbxij(p3mat,nhstepm,agec,hstepm,p,prevacurrent,nlstate,stepm,oldm,savm,oldm,savm, dnewm, doldm, dsavm, k); */
7891: /* for (h=0; h<=nhstepm; h++){ */
7892: /* if (h*hstepm/YEARM*stepm ==yearp) { */
7893: /* fprintf(ficresfb,"\n"); */
1.225 brouard 7894: /* for(j=1;j<=cptcoveff;j++) */
1.218 brouard 7895: /* fprintf(ficresfb,"%d %d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]); */
7896: /* fprintf(ficresfb,"%.f %.f ",anback1+yearp,agec+h*hstepm/YEARM*stepm); */
7897: /* } */
7898: /* for(j=1; j<=nlstate+ndeath;j++) { */
7899: /* ppij=0.; */
7900: /* for(i=1; i<=nlstate;i++) { */
7901: /* if (mobilav==1) */
7902: /* ppij=ppij+p3mat[i][j][h]*mobaverage[(int)agec][i][cptcod]; */
7903: /* else { */
7904: /* ppij=ppij+p3mat[i][j][h]*probs[(int)(agec)][i][cptcod]; */
7905: /* } */
7906: /* if (h*hstepm/YEARM*stepm== yearp) { */
7907: /* fprintf(ficresfb," %.3f", p3mat[i][j][h]); */
7908: /* } */
7909: /* } /\* end i *\/ */
7910: /* if (h*hstepm/YEARM*stepm==yearp) { */
7911: /* fprintf(ficresfb," %.3f", ppij); */
7912: /* } */
7913: /* }/\* end j *\/ */
7914: /* } /\* end h *\/ */
7915: /* free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm); */
7916: /* } /\* end agec *\/ */
7917: /* } /\* end yearp *\/ */
7918: /* } /\* end cptcod *\/ */
7919: /* } /\* end cptcov *\/ */
7920:
7921: /* /\* if (mobilav!=0) free_ma3x(mobaverage,1, AGESUP,1,NCOVMAX, 1,NCOVMAX); *\/ */
7922:
7923: /* fclose(ficresfb); */
7924: /* printf("End of Computing Back forecasting \n"); */
7925: /* fprintf(ficlog,"End of Computing Back forecasting\n"); */
1.217 brouard 7926:
1.218 brouard 7927: /* } */
1.217 brouard 7928:
1.126 brouard 7929: /************** Forecasting *****not tested NB*************/
1.227 brouard 7930: /* void populforecast(char fileres[], double anpyram,double mpyram,double jpyram,double ageminpar, double agemax,double dateprev1, double dateprev2s, int mobilav, double agedeb, double fage, int popforecast, char popfile[], double anpyram1,double p[], int i2){ */
1.126 brouard 7931:
1.227 brouard 7932: /* int cpt, stepsize, hstepm, nhstepm, j,k,c, cptcod, i,h; */
7933: /* int *popage; */
7934: /* double calagedatem, agelim, kk1, kk2; */
7935: /* double *popeffectif,*popcount; */
7936: /* double ***p3mat,***tabpop,***tabpopprev; */
7937: /* /\* double ***mobaverage; *\/ */
7938: /* char filerespop[FILENAMELENGTH]; */
1.126 brouard 7939:
1.227 brouard 7940: /* tabpop= ma3x(1, AGESUP,1,NCOVMAX, 1,NCOVMAX); */
7941: /* tabpopprev= ma3x(1, AGESUP,1,NCOVMAX, 1,NCOVMAX); */
7942: /* agelim=AGESUP; */
7943: /* calagedatem=(anpyram+mpyram/12.+jpyram/365.-dateintmean)*YEARM; */
1.126 brouard 7944:
1.227 brouard 7945: /* prevalence(probs, ageminpar, agemax, s, agev, nlstate, imx, Tvar, nbcode, ncodemax, mint, anint, dateprev1, dateprev2, firstpass, lastpass); */
1.126 brouard 7946:
7947:
1.227 brouard 7948: /* strcpy(filerespop,"POP_"); */
7949: /* strcat(filerespop,fileresu); */
7950: /* if((ficrespop=fopen(filerespop,"w"))==NULL) { */
7951: /* printf("Problem with forecast resultfile: %s\n", filerespop); */
7952: /* fprintf(ficlog,"Problem with forecast resultfile: %s\n", filerespop); */
7953: /* } */
7954: /* printf("Computing forecasting: result on file '%s' \n", filerespop); */
7955: /* fprintf(ficlog,"Computing forecasting: result on file '%s' \n", filerespop); */
1.126 brouard 7956:
1.227 brouard 7957: /* if (cptcoveff==0) ncodemax[cptcoveff]=1; */
1.126 brouard 7958:
1.227 brouard 7959: /* /\* if (mobilav!=0) { *\/ */
7960: /* /\* mobaverage= ma3x(1, AGESUP,1,NCOVMAX, 1,NCOVMAX); *\/ */
7961: /* /\* if (movingaverage(probs, ageminpar, fage, mobaverage,mobilav)!=0){ *\/ */
7962: /* /\* fprintf(ficlog," Error in movingaverage mobilav=%d\n",mobilav); *\/ */
7963: /* /\* printf(" Error in movingaverage mobilav=%d\n",mobilav); *\/ */
7964: /* /\* } *\/ */
7965: /* /\* } *\/ */
1.126 brouard 7966:
1.227 brouard 7967: /* stepsize=(int) (stepm+YEARM-1)/YEARM; */
7968: /* if (stepm<=12) stepsize=1; */
1.126 brouard 7969:
1.227 brouard 7970: /* agelim=AGESUP; */
1.126 brouard 7971:
1.227 brouard 7972: /* hstepm=1; */
7973: /* hstepm=hstepm/stepm; */
1.218 brouard 7974:
1.227 brouard 7975: /* if (popforecast==1) { */
7976: /* if((ficpop=fopen(popfile,"r"))==NULL) { */
7977: /* printf("Problem with population file : %s\n",popfile);exit(0); */
7978: /* fprintf(ficlog,"Problem with population file : %s\n",popfile);exit(0); */
7979: /* } */
7980: /* popage=ivector(0,AGESUP); */
7981: /* popeffectif=vector(0,AGESUP); */
7982: /* popcount=vector(0,AGESUP); */
1.126 brouard 7983:
1.227 brouard 7984: /* i=1; */
7985: /* while ((c=fscanf(ficpop,"%d %lf\n",&popage[i],&popcount[i])) != EOF) i=i+1; */
1.218 brouard 7986:
1.227 brouard 7987: /* imx=i; */
7988: /* for (i=1; i<imx;i++) popeffectif[popage[i]]=popcount[i]; */
7989: /* } */
1.218 brouard 7990:
1.227 brouard 7991: /* for(cptcov=1,k=0;cptcov<=i2;cptcov++){ */
7992: /* for(cptcod=1;cptcod<=ncodemax[cptcoveff];cptcod++){ */
7993: /* k=k+1; */
7994: /* fprintf(ficrespop,"\n#******"); */
7995: /* for(j=1;j<=cptcoveff;j++) { */
7996: /* fprintf(ficrespop," V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]); */
7997: /* } */
7998: /* fprintf(ficrespop,"******\n"); */
7999: /* fprintf(ficrespop,"# Age"); */
8000: /* for(j=1; j<=nlstate+ndeath;j++) fprintf(ficrespop," P.%d",j); */
8001: /* if (popforecast==1) fprintf(ficrespop," [Population]"); */
1.126 brouard 8002:
1.227 brouard 8003: /* for (cpt=0; cpt<=0;cpt++) { */
8004: /* fprintf(ficrespop,"\n\n# Forecasting at date %.lf/%.lf/%.lf ",jpyram,mpyram,anpyram+cpt); */
1.126 brouard 8005:
1.227 brouard 8006: /* for (agedeb=(fage-((int)calagedatem %12/12.)); agedeb>=(ageminpar-((int)calagedatem %12)/12.); agedeb--){ */
8007: /* nhstepm=(int) rint((agelim-agedeb)*YEARM/stepm); */
8008: /* nhstepm = nhstepm/hstepm; */
1.126 brouard 8009:
1.227 brouard 8010: /* p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm); */
8011: /* oldm=oldms;savm=savms; */
8012: /* hpxij(p3mat,nhstepm,agedeb,hstepm,p,nlstate,stepm,oldm,savm, k); */
1.218 brouard 8013:
1.227 brouard 8014: /* for (h=0; h<=nhstepm; h++){ */
8015: /* if (h==(int) (calagedatem+YEARM*cpt)) { */
8016: /* fprintf(ficrespop,"\n %3.f ",agedeb+h*hstepm/YEARM*stepm); */
8017: /* } */
8018: /* for(j=1; j<=nlstate+ndeath;j++) { */
8019: /* kk1=0.;kk2=0; */
8020: /* for(i=1; i<=nlstate;i++) { */
8021: /* if (mobilav==1) */
8022: /* kk1=kk1+p3mat[i][j][h]*mobaverage[(int)agedeb+1][i][cptcod]; */
8023: /* else { */
8024: /* kk1=kk1+p3mat[i][j][h]*probs[(int)(agedeb+1)][i][cptcod]; */
8025: /* } */
8026: /* } */
8027: /* if (h==(int)(calagedatem+12*cpt)){ */
8028: /* tabpop[(int)(agedeb)][j][cptcod]=kk1; */
8029: /* /\*fprintf(ficrespop," %.3f", kk1); */
8030: /* if (popforecast==1) fprintf(ficrespop," [%.f]", kk1*popeffectif[(int)agedeb+1]);*\/ */
8031: /* } */
8032: /* } */
8033: /* for(i=1; i<=nlstate;i++){ */
8034: /* kk1=0.; */
8035: /* for(j=1; j<=nlstate;j++){ */
8036: /* kk1= kk1+tabpop[(int)(agedeb)][j][cptcod]; */
8037: /* } */
8038: /* tabpopprev[(int)(agedeb)][i][cptcod]=tabpop[(int)(agedeb)][i][cptcod]/kk1*popeffectif[(int)(agedeb+(calagedatem+12*cpt)*hstepm/YEARM*stepm-1)]; */
8039: /* } */
1.218 brouard 8040:
1.227 brouard 8041: /* if (h==(int)(calagedatem+12*cpt)) */
8042: /* for(j=1; j<=nlstate;j++) */
8043: /* fprintf(ficrespop," %15.2f",tabpopprev[(int)(agedeb+1)][j][cptcod]); */
8044: /* } */
8045: /* free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm); */
8046: /* } */
8047: /* } */
1.218 brouard 8048:
1.227 brouard 8049: /* /\******\/ */
1.218 brouard 8050:
1.227 brouard 8051: /* for (cpt=1; cpt<=(anpyram1-anpyram);cpt++) { */
8052: /* fprintf(ficrespop,"\n\n# Forecasting at date %.lf/%.lf/%.lf ",jpyram,mpyram,anpyram+cpt); */
8053: /* for (agedeb=(fage-((int)calagedatem %12/12.)); agedeb>=(ageminpar-((int)calagedatem %12)/12.); agedeb--){ */
8054: /* nhstepm=(int) rint((agelim-agedeb)*YEARM/stepm); */
8055: /* nhstepm = nhstepm/hstepm; */
1.126 brouard 8056:
1.227 brouard 8057: /* p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm); */
8058: /* oldm=oldms;savm=savms; */
8059: /* hpxij(p3mat,nhstepm,agedeb,hstepm,p,nlstate,stepm,oldm,savm, k); */
8060: /* for (h=0; h<=nhstepm; h++){ */
8061: /* if (h==(int) (calagedatem+YEARM*cpt)) { */
8062: /* fprintf(ficresf,"\n %3.f ",agedeb+h*hstepm/YEARM*stepm); */
8063: /* } */
8064: /* for(j=1; j<=nlstate+ndeath;j++) { */
8065: /* kk1=0.;kk2=0; */
8066: /* for(i=1; i<=nlstate;i++) { */
8067: /* kk1=kk1+p3mat[i][j][h]*tabpopprev[(int)agedeb+1][i][cptcod]; */
8068: /* } */
8069: /* if (h==(int)(calagedatem+12*cpt)) fprintf(ficresf," %15.2f", kk1); */
8070: /* } */
8071: /* } */
8072: /* free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm); */
8073: /* } */
8074: /* } */
8075: /* } */
8076: /* } */
1.218 brouard 8077:
1.227 brouard 8078: /* /\* if (mobilav!=0) free_ma3x(mobaverage,1, AGESUP,1,NCOVMAX, 1,NCOVMAX); *\/ */
1.218 brouard 8079:
1.227 brouard 8080: /* if (popforecast==1) { */
8081: /* free_ivector(popage,0,AGESUP); */
8082: /* free_vector(popeffectif,0,AGESUP); */
8083: /* free_vector(popcount,0,AGESUP); */
8084: /* } */
8085: /* free_ma3x(tabpop,1, AGESUP,1,NCOVMAX, 1,NCOVMAX); */
8086: /* free_ma3x(tabpopprev,1, AGESUP,1,NCOVMAX, 1,NCOVMAX); */
8087: /* fclose(ficrespop); */
8088: /* } /\* End of popforecast *\/ */
1.218 brouard 8089:
1.126 brouard 8090: int fileappend(FILE *fichier, char *optionfich)
8091: {
8092: if((fichier=fopen(optionfich,"a"))==NULL) {
8093: printf("Problem with file: %s\n", optionfich);
8094: fprintf(ficlog,"Problem with file: %s\n", optionfich);
8095: return (0);
8096: }
8097: fflush(fichier);
8098: return (1);
8099: }
8100:
8101:
8102: /**************** function prwizard **********************/
8103: void prwizard(int ncovmodel, int nlstate, int ndeath, char model[], FILE *ficparo)
8104: {
8105:
8106: /* Wizard to print covariance matrix template */
8107:
1.164 brouard 8108: char ca[32], cb[32];
8109: int i,j, k, li, lj, lk, ll, jj, npar, itimes;
1.126 brouard 8110: int numlinepar;
8111:
8112: printf("# Parameters nlstate*nlstate*ncov a12*1 + b12 * age + ...\n");
8113: fprintf(ficparo,"# Parameters nlstate*nlstate*ncov a12*1 + b12 * age + ...\n");
8114: for(i=1; i <=nlstate; i++){
8115: jj=0;
8116: for(j=1; j <=nlstate+ndeath; j++){
8117: if(j==i) continue;
8118: jj++;
8119: /*ca[0]= k+'a'-1;ca[1]='\0';*/
8120: printf("%1d%1d",i,j);
8121: fprintf(ficparo,"%1d%1d",i,j);
8122: for(k=1; k<=ncovmodel;k++){
8123: /* printf(" %lf",param[i][j][k]); */
8124: /* fprintf(ficparo," %lf",param[i][j][k]); */
8125: printf(" 0.");
8126: fprintf(ficparo," 0.");
8127: }
8128: printf("\n");
8129: fprintf(ficparo,"\n");
8130: }
8131: }
8132: printf("# Scales (for hessian or gradient estimation)\n");
8133: fprintf(ficparo,"# Scales (for hessian or gradient estimation)\n");
8134: npar= (nlstate+ndeath-1)*nlstate*ncovmodel; /* Number of parameters*/
8135: for(i=1; i <=nlstate; i++){
8136: jj=0;
8137: for(j=1; j <=nlstate+ndeath; j++){
8138: if(j==i) continue;
8139: jj++;
8140: fprintf(ficparo,"%1d%1d",i,j);
8141: printf("%1d%1d",i,j);
8142: fflush(stdout);
8143: for(k=1; k<=ncovmodel;k++){
8144: /* printf(" %le",delti3[i][j][k]); */
8145: /* fprintf(ficparo," %le",delti3[i][j][k]); */
8146: printf(" 0.");
8147: fprintf(ficparo," 0.");
8148: }
8149: numlinepar++;
8150: printf("\n");
8151: fprintf(ficparo,"\n");
8152: }
8153: }
8154: printf("# Covariance matrix\n");
8155: /* # 121 Var(a12)\n\ */
8156: /* # 122 Cov(b12,a12) Var(b12)\n\ */
8157: /* # 131 Cov(a13,a12) Cov(a13,b12, Var(a13)\n\ */
8158: /* # 132 Cov(b13,a12) Cov(b13,b12, Cov(b13,a13) Var(b13)\n\ */
8159: /* # 212 Cov(a21,a12) Cov(a21,b12, Cov(a21,a13) Cov(a21,b13) Var(a21)\n\ */
8160: /* # 212 Cov(b21,a12) Cov(b21,b12, Cov(b21,a13) Cov(b21,b13) Cov(b21,a21) Var(b21)\n\ */
8161: /* # 232 Cov(a23,a12) Cov(a23,b12, Cov(a23,a13) Cov(a23,b13) Cov(a23,a21) Cov(a23,b21) Var(a23)\n\ */
8162: /* # 232 Cov(b23,a12) Cov(b23,b12) ... Var (b23)\n" */
8163: fflush(stdout);
8164: fprintf(ficparo,"# Covariance matrix\n");
8165: /* # 121 Var(a12)\n\ */
8166: /* # 122 Cov(b12,a12) Var(b12)\n\ */
8167: /* # ...\n\ */
8168: /* # 232 Cov(b23,a12) Cov(b23,b12) ... Var (b23)\n" */
8169:
8170: for(itimes=1;itimes<=2;itimes++){
8171: jj=0;
8172: for(i=1; i <=nlstate; i++){
8173: for(j=1; j <=nlstate+ndeath; j++){
8174: if(j==i) continue;
8175: for(k=1; k<=ncovmodel;k++){
8176: jj++;
8177: ca[0]= k+'a'-1;ca[1]='\0';
8178: if(itimes==1){
8179: printf("#%1d%1d%d",i,j,k);
8180: fprintf(ficparo,"#%1d%1d%d",i,j,k);
8181: }else{
8182: printf("%1d%1d%d",i,j,k);
8183: fprintf(ficparo,"%1d%1d%d",i,j,k);
8184: /* printf(" %.5le",matcov[i][j]); */
8185: }
8186: ll=0;
8187: for(li=1;li <=nlstate; li++){
8188: for(lj=1;lj <=nlstate+ndeath; lj++){
8189: if(lj==li) continue;
8190: for(lk=1;lk<=ncovmodel;lk++){
8191: ll++;
8192: if(ll<=jj){
8193: cb[0]= lk +'a'-1;cb[1]='\0';
8194: if(ll<jj){
8195: if(itimes==1){
8196: printf(" Cov(%s%1d%1d,%s%1d%1d)",ca,i,j,cb, li,lj);
8197: fprintf(ficparo," Cov(%s%1d%1d,%s%1d%1d)",ca,i,j,cb, li,lj);
8198: }else{
8199: printf(" 0.");
8200: fprintf(ficparo," 0.");
8201: }
8202: }else{
8203: if(itimes==1){
8204: printf(" Var(%s%1d%1d)",ca,i,j);
8205: fprintf(ficparo," Var(%s%1d%1d)",ca,i,j);
8206: }else{
8207: printf(" 0.");
8208: fprintf(ficparo," 0.");
8209: }
8210: }
8211: }
8212: } /* end lk */
8213: } /* end lj */
8214: } /* end li */
8215: printf("\n");
8216: fprintf(ficparo,"\n");
8217: numlinepar++;
8218: } /* end k*/
8219: } /*end j */
8220: } /* end i */
8221: } /* end itimes */
8222:
8223: } /* end of prwizard */
8224: /******************* Gompertz Likelihood ******************************/
8225: double gompertz(double x[])
8226: {
8227: double A,B,L=0.0,sump=0.,num=0.;
8228: int i,n=0; /* n is the size of the sample */
8229:
1.220 brouard 8230: for (i=1;i<=imx ; i++) {
1.126 brouard 8231: sump=sump+weight[i];
8232: /* sump=sump+1;*/
8233: num=num+1;
8234: }
8235:
8236:
8237: /* for (i=0; i<=imx; i++)
8238: if (wav[i]>0) printf("i=%d ageex=%lf agecens=%lf agedc=%lf cens=%d %d\n" ,i,ageexmed[i],agecens[i],agedc[i],cens[i],wav[i]);*/
8239:
8240: for (i=1;i<=imx ; i++)
8241: {
8242: if (cens[i] == 1 && wav[i]>1)
8243: A=-x[1]/(x[2])*(exp(x[2]*(agecens[i]-agegomp))-exp(x[2]*(ageexmed[i]-agegomp)));
8244:
8245: if (cens[i] == 0 && wav[i]>1)
8246: A=-x[1]/(x[2])*(exp(x[2]*(agedc[i]-agegomp))-exp(x[2]*(ageexmed[i]-agegomp)))
8247: +log(x[1]/YEARM)+x[2]*(agedc[i]-agegomp)+log(YEARM);
8248:
8249: /*if (wav[i] > 1 && agecens[i] > 15) {*/ /* ??? */
8250: if (wav[i] > 1 ) { /* ??? */
8251: L=L+A*weight[i];
8252: /* printf("\ni=%d A=%f L=%lf x[1]=%lf x[2]=%lf ageex=%lf agecens=%lf cens=%d agedc=%lf weight=%lf\n",i,A,L,x[1],x[2],ageexmed[i]*12,agecens[i]*12,cens[i],agedc[i]*12,weight[i]);*/
8253: }
8254: }
8255:
8256: /*printf("x1=%2.9f x2=%2.9f x3=%2.9f L=%f\n",x[1],x[2],x[3],L);*/
8257:
8258: return -2*L*num/sump;
8259: }
8260:
1.136 brouard 8261: #ifdef GSL
8262: /******************* Gompertz_f Likelihood ******************************/
8263: double gompertz_f(const gsl_vector *v, void *params)
8264: {
8265: double A,B,LL=0.0,sump=0.,num=0.;
8266: double *x= (double *) v->data;
8267: int i,n=0; /* n is the size of the sample */
8268:
8269: for (i=0;i<=imx-1 ; i++) {
8270: sump=sump+weight[i];
8271: /* sump=sump+1;*/
8272: num=num+1;
8273: }
8274:
8275:
8276: /* for (i=0; i<=imx; i++)
8277: if (wav[i]>0) printf("i=%d ageex=%lf agecens=%lf agedc=%lf cens=%d %d\n" ,i,ageexmed[i],agecens[i],agedc[i],cens[i],wav[i]);*/
8278: printf("x[0]=%lf x[1]=%lf\n",x[0],x[1]);
8279: for (i=1;i<=imx ; i++)
8280: {
8281: if (cens[i] == 1 && wav[i]>1)
8282: A=-x[0]/(x[1])*(exp(x[1]*(agecens[i]-agegomp))-exp(x[1]*(ageexmed[i]-agegomp)));
8283:
8284: if (cens[i] == 0 && wav[i]>1)
8285: A=-x[0]/(x[1])*(exp(x[1]*(agedc[i]-agegomp))-exp(x[1]*(ageexmed[i]-agegomp)))
8286: +log(x[0]/YEARM)+x[1]*(agedc[i]-agegomp)+log(YEARM);
8287:
8288: /*if (wav[i] > 1 && agecens[i] > 15) {*/ /* ??? */
8289: if (wav[i] > 1 ) { /* ??? */
8290: LL=LL+A*weight[i];
8291: /* printf("\ni=%d A=%f L=%lf x[1]=%lf x[2]=%lf ageex=%lf agecens=%lf cens=%d agedc=%lf weight=%lf\n",i,A,L,x[1],x[2],ageexmed[i]*12,agecens[i]*12,cens[i],agedc[i]*12,weight[i]);*/
8292: }
8293: }
8294:
8295: /*printf("x1=%2.9f x2=%2.9f x3=%2.9f L=%f\n",x[1],x[2],x[3],L);*/
8296: printf("x[0]=%lf x[1]=%lf -2*LL*num/sump=%lf\n",x[0],x[1],-2*LL*num/sump);
8297:
8298: return -2*LL*num/sump;
8299: }
8300: #endif
8301:
1.126 brouard 8302: /******************* Printing html file ***********/
1.201 brouard 8303: void printinghtmlmort(char fileresu[], char title[], char datafile[], int firstpass, \
1.126 brouard 8304: int lastpass, int stepm, int weightopt, char model[],\
8305: int imx, double p[],double **matcov,double agemortsup){
8306: int i,k;
8307:
8308: fprintf(fichtm,"<ul><li><h4>Result files </h4>\n Force of mortality. Parameters of the Gompertz fit (with confidence interval in brackets):<br>");
8309: fprintf(fichtm," mu(age) =%lf*exp(%lf*(age-%d)) per year<br><br>",p[1],p[2],agegomp);
8310: for (i=1;i<=2;i++)
8311: fprintf(fichtm," p[%d] = %lf [%f ; %f]<br>\n",i,p[i],p[i]-2*sqrt(matcov[i][i]),p[i]+2*sqrt(matcov[i][i]));
1.199 brouard 8312: fprintf(fichtm,"<br><br><img src=\"graphmort.svg\">");
1.126 brouard 8313: fprintf(fichtm,"</ul>");
8314:
8315: fprintf(fichtm,"<ul><li><h4>Life table</h4>\n <br>");
8316:
8317: fprintf(fichtm,"\nAge l<inf>x</inf> q<inf>x</inf> d(x,x+1) L<inf>x</inf> T<inf>x</inf> e<infx</inf><br>");
8318:
8319: for (k=agegomp;k<(agemortsup-2);k++)
8320: fprintf(fichtm,"%d %.0lf %lf %.0lf %.0lf %.0lf %lf<br>\n",k,lsurv[k],p[1]*exp(p[2]*(k-agegomp)),(p[1]*exp(p[2]*(k-agegomp)))*lsurv[k],lpop[k],tpop[k],tpop[k]/lsurv[k]);
8321:
8322:
8323: fflush(fichtm);
8324: }
8325:
8326: /******************* Gnuplot file **************/
1.201 brouard 8327: void printinggnuplotmort(char fileresu[], char optionfilefiname[], double ageminpar, double agemaxpar, double fage , char pathc[], double p[]){
1.126 brouard 8328:
8329: char dirfileres[132],optfileres[132];
1.164 brouard 8330:
1.126 brouard 8331: int ng;
8332:
8333:
8334: /*#ifdef windows */
8335: fprintf(ficgp,"cd \"%s\" \n",pathc);
8336: /*#endif */
8337:
8338:
8339: strcpy(dirfileres,optionfilefiname);
8340: strcpy(optfileres,"vpl");
1.199 brouard 8341: fprintf(ficgp,"set out \"graphmort.svg\"\n ");
1.126 brouard 8342: fprintf(ficgp,"set xlabel \"Age\"\n set ylabel \"Force of mortality (per year)\" \n ");
1.199 brouard 8343: fprintf(ficgp, "set ter svg size 640, 480\n set log y\n");
1.145 brouard 8344: /* fprintf(ficgp, "set size 0.65,0.65\n"); */
1.126 brouard 8345: fprintf(ficgp,"plot [%d:100] %lf*exp(%lf*(x-%d))",agegomp,p[1],p[2],agegomp);
8346:
8347: }
8348:
1.136 brouard 8349: int readdata(char datafile[], int firstobs, int lastobs, int *imax)
8350: {
1.126 brouard 8351:
1.136 brouard 8352: /*-------- data file ----------*/
8353: FILE *fic;
8354: char dummy[]=" ";
1.240 brouard 8355: int i=0, j=0, n=0, iv=0, v;
1.223 brouard 8356: int lstra;
1.136 brouard 8357: int linei, month, year,iout;
8358: char line[MAXLINE], linetmp[MAXLINE];
1.164 brouard 8359: char stra[MAXLINE], strb[MAXLINE];
1.136 brouard 8360: char *stratrunc;
1.223 brouard 8361:
1.240 brouard 8362: DummyV=ivector(1,NCOVMAX); /* 1 to 3 */
8363: FixedV=ivector(1,NCOVMAX); /* 1 to 3 */
1.126 brouard 8364:
1.240 brouard 8365: for(v=1; v <=ncovcol;v++){
8366: DummyV[v]=0;
8367: FixedV[v]=0;
8368: }
8369: for(v=ncovcol+1; v <=ncovcol+nqv;v++){
8370: DummyV[v]=1;
8371: FixedV[v]=0;
8372: }
8373: for(v=ncovcol+nqv+1; v <=ncovcol+nqv+ntv;v++){
8374: DummyV[v]=0;
8375: FixedV[v]=1;
8376: }
8377: for(v=ncovcol+nqv+ntv+1; v <=ncovcol+nqv+ntv+nqtv;v++){
8378: DummyV[v]=1;
8379: FixedV[v]=1;
8380: }
8381: for(v=1; v <=ncovcol+nqv+ntv+nqtv;v++){
8382: printf("Covariate type in the data: V%d, DummyV(V%d)=%d, FixedV(V%d)=%d\n",v,v,DummyV[v],v,FixedV[v]);
8383: fprintf(ficlog,"Covariate type in the data: V%d, DummyV(V%d)=%d, FixedV(V%d)=%d\n",v,v,DummyV[v],v,FixedV[v]);
8384: }
1.126 brouard 8385:
1.136 brouard 8386: if((fic=fopen(datafile,"r"))==NULL) {
1.218 brouard 8387: printf("Problem while opening datafile: %s with errno='%s'\n", datafile,strerror(errno));fflush(stdout);
8388: fprintf(ficlog,"Problem while opening datafile: %s with errno='%s'\n", datafile,strerror(errno));fflush(ficlog);return 1;
1.136 brouard 8389: }
1.126 brouard 8390:
1.136 brouard 8391: i=1;
8392: linei=0;
8393: while ((fgets(line, MAXLINE, fic) != NULL) &&((i >= firstobs) && (i <=lastobs))) {
8394: linei=linei+1;
8395: for(j=strlen(line); j>=0;j--){ /* Untabifies line */
8396: if(line[j] == '\t')
8397: line[j] = ' ';
8398: }
8399: for(j=strlen(line)-1; (line[j]==' ')||(line[j]==10)||(line[j]==13);j--){
8400: ;
8401: };
8402: line[j+1]=0; /* Trims blanks at end of line */
8403: if(line[0]=='#'){
8404: fprintf(ficlog,"Comment line\n%s\n",line);
8405: printf("Comment line\n%s\n",line);
8406: continue;
8407: }
8408: trimbb(linetmp,line); /* Trims multiple blanks in line */
1.164 brouard 8409: strcpy(line, linetmp);
1.223 brouard 8410:
8411: /* Loops on waves */
8412: for (j=maxwav;j>=1;j--){
8413: for (iv=nqtv;iv>=1;iv--){ /* Loop on time varying quantitative variables */
1.238 brouard 8414: cutv(stra, strb, line, ' ');
8415: if(strb[0]=='.') { /* Missing value */
8416: lval=-1;
8417: cotqvar[j][iv][i]=-1; /* 0.0/0.0 */
8418: cotvar[j][ntv+iv][i]=-1; /* For performance reasons */
8419: if(isalpha(strb[1])) { /* .m or .d Really Missing value */
8420: printf("Error reading data around '%s' at line number %d for individual %d, '%s'\nShould be the %d th quantitative value out of %d measured at wave %d. If missing, you should remove this individual or impute a value. Exiting.\n", strb, linei,i,line,iv, nqtv, j);
8421: fprintf(ficlog,"Error reading data around '%s' at line number %d for individual %d, '%s'\nShould be the %d th quantitative value out of %d measured at wave %d. If missing, you should remove this individual or impute a value. Exiting.\n", strb, linei,i,line,iv, nqtv, j);fflush(ficlog);
8422: return 1;
8423: }
8424: }else{
8425: errno=0;
8426: /* what_kind_of_number(strb); */
8427: dval=strtod(strb,&endptr);
8428: /* if( strb[0]=='\0' || (*endptr != '\0')){ */
8429: /* if(strb != endptr && *endptr == '\0') */
8430: /* dval=dlval; */
8431: /* if (errno == ERANGE && (lval == LONG_MAX || lval == LONG_MIN)) */
8432: if( strb[0]=='\0' || (*endptr != '\0')){
8433: printf("Error reading data around '%s' at line number %d for individual %d, '%s'\nShould be the %d th quantitative value out of %d measured at wave %d. Setting maxwav=%d might be wrong. Exiting.\n", strb, linei,i,line,iv, nqtv, j,maxwav);
8434: fprintf(ficlog,"Error reading data around '%s' at line number %d for individual %d, '%s'\nShould be the %d th quantitative value out of %d measured at wave %d. Setting maxwav=%d might be wrong. Exiting.\n", strb, linei,i,line, iv, nqtv, j,maxwav);fflush(ficlog);
8435: return 1;
8436: }
8437: cotqvar[j][iv][i]=dval;
8438: cotvar[j][ntv+iv][i]=dval;
8439: }
8440: strcpy(line,stra);
1.223 brouard 8441: }/* end loop ntqv */
1.225 brouard 8442:
1.223 brouard 8443: for (iv=ntv;iv>=1;iv--){ /* Loop on time varying dummies */
1.238 brouard 8444: cutv(stra, strb, line, ' ');
8445: if(strb[0]=='.') { /* Missing value */
8446: lval=-1;
8447: }else{
8448: errno=0;
8449: lval=strtol(strb,&endptr,10);
8450: /* if (errno == ERANGE && (lval == LONG_MAX || lval == LONG_MIN))*/
8451: if( strb[0]=='\0' || (*endptr != '\0')){
8452: printf("Error reading data around '%s' at line number %d for individual %d, '%s'\nShould be the %d th dummy covariate out of %d measured at wave %d. Setting maxwav=%d might be wrong. Exiting.\n", strb, linei,i,line,iv, ntv, j,maxwav);
8453: fprintf(ficlog,"Error reading data around '%s' at line number %d for individual %d, '%s'\nShould be the %d dummy covariate out of %d measured wave %d. Setting maxwav=%d might be wrong. Exiting.\n", strb, linei,i,line,iv, ntv,j,maxwav);fflush(ficlog);
8454: return 1;
8455: }
8456: }
8457: if(lval <-1 || lval >1){
8458: printf("Error reading data around '%ld' at line number %d for individual %d, '%s'\n \
1.223 brouard 8459: Should be a value of %d(nth) covariate (0 should be the value for the reference and 1\n \
8460: for the alternative. IMaCh does not build design variables automatically, do it yourself.\n \
1.238 brouard 8461: For example, for multinomial values like 1, 2 and 3,\n \
8462: build V1=0 V2=0 for the reference value (1),\n \
8463: V1=1 V2=0 for (2) \n \
1.223 brouard 8464: and V1=0 V2=1 for (3). V1=1 V2=1 should not exist and the corresponding\n \
1.238 brouard 8465: output of IMaCh is often meaningless.\n \
1.223 brouard 8466: Exiting.\n",lval,linei, i,line,j);
1.238 brouard 8467: fprintf(ficlog,"Error reading data around '%ld' at line number %d for individual %d, '%s'\n \
1.223 brouard 8468: Should be a value of %d(nth) covariate (0 should be the value for the reference and 1\n \
8469: for the alternative. IMaCh does not build design variables automatically, do it yourself.\n \
1.238 brouard 8470: For example, for multinomial values like 1, 2 and 3,\n \
8471: build V1=0 V2=0 for the reference value (1),\n \
8472: V1=1 V2=0 for (2) \n \
1.223 brouard 8473: and V1=0 V2=1 for (3). V1=1 V2=1 should not exist and the corresponding\n \
1.238 brouard 8474: output of IMaCh is often meaningless.\n \
1.223 brouard 8475: Exiting.\n",lval,linei, i,line,j);fflush(ficlog);
1.238 brouard 8476: return 1;
8477: }
8478: cotvar[j][iv][i]=(double)(lval);
8479: strcpy(line,stra);
1.223 brouard 8480: }/* end loop ntv */
1.225 brouard 8481:
1.223 brouard 8482: /* Statuses at wave */
1.137 brouard 8483: cutv(stra, strb, line, ' ');
1.223 brouard 8484: if(strb[0]=='.') { /* Missing value */
1.238 brouard 8485: lval=-1;
1.136 brouard 8486: }else{
1.238 brouard 8487: errno=0;
8488: lval=strtol(strb,&endptr,10);
8489: /* if (errno == ERANGE && (lval == LONG_MAX || lval == LONG_MIN))*/
8490: if( strb[0]=='\0' || (*endptr != '\0')){
8491: printf("Error reading data around '%s' at line number %d for individual %d, '%s'\nShould be a status of wave %d. Setting maxwav=%d might be wrong. Exiting.\n", strb, linei,i,line,j,maxwav);
8492: fprintf(ficlog,"Error reading data around '%s' at line number %d for individual %d, '%s'\nShould be a status of wave %d. Setting maxwav=%d might be wrong. Exiting.\n", strb, linei,i,line,j,maxwav);fflush(ficlog);
8493: return 1;
8494: }
1.136 brouard 8495: }
1.225 brouard 8496:
1.136 brouard 8497: s[j][i]=lval;
1.225 brouard 8498:
1.223 brouard 8499: /* Date of Interview */
1.136 brouard 8500: strcpy(line,stra);
8501: cutv(stra, strb,line,' ');
1.169 brouard 8502: if( (iout=sscanf(strb,"%d/%d",&month, &year)) != 0){
1.136 brouard 8503: }
1.169 brouard 8504: else if( (iout=sscanf(strb,"%s.",dummy)) != 0){
1.225 brouard 8505: month=99;
8506: year=9999;
1.136 brouard 8507: }else{
1.225 brouard 8508: printf("Error reading data around '%s' at line number %d for individual %d, '%s'\nShould be a date of interview (mm/yyyy or .) at wave %d. Exiting.\n",strb, linei,i, line,j);
8509: fprintf(ficlog,"Error reading data around '%s' at line number %d for individual %d, '%s'\nShould be a date of interview (mm/yyyy or .) at wave %d. Exiting.\n",strb, linei,i, line,j);fflush(ficlog);
8510: return 1;
1.136 brouard 8511: }
8512: anint[j][i]= (double) year;
8513: mint[j][i]= (double)month;
8514: strcpy(line,stra);
1.223 brouard 8515: } /* End loop on waves */
1.225 brouard 8516:
1.223 brouard 8517: /* Date of death */
1.136 brouard 8518: cutv(stra, strb,line,' ');
1.169 brouard 8519: if( (iout=sscanf(strb,"%d/%d",&month, &year)) != 0){
1.136 brouard 8520: }
1.169 brouard 8521: else if( (iout=sscanf(strb,"%s.",dummy)) != 0){
1.136 brouard 8522: month=99;
8523: year=9999;
8524: }else{
1.141 brouard 8525: printf("Error reading data around '%s' at line number %d for individual %d, '%s'\nShould be a date of death (mm/yyyy or .). Exiting.\n",strb, linei,i,line);
1.225 brouard 8526: fprintf(ficlog,"Error reading data around '%s' at line number %d for individual %d, '%s'\nShould be a date of death (mm/yyyy or .). Exiting.\n",strb, linei,i,line);fflush(ficlog);
8527: return 1;
1.136 brouard 8528: }
8529: andc[i]=(double) year;
8530: moisdc[i]=(double) month;
8531: strcpy(line,stra);
8532:
1.223 brouard 8533: /* Date of birth */
1.136 brouard 8534: cutv(stra, strb,line,' ');
1.169 brouard 8535: if( (iout=sscanf(strb,"%d/%d",&month, &year)) != 0){
1.136 brouard 8536: }
1.169 brouard 8537: else if( (iout=sscanf(strb,"%s.", dummy)) != 0){
1.136 brouard 8538: month=99;
8539: year=9999;
8540: }else{
1.141 brouard 8541: printf("Error reading data around '%s' at line number %d for individual %d, '%s'\nShould be a date of birth (mm/yyyy or .). Exiting.\n",strb, linei,i,line);
8542: fprintf(ficlog,"Error reading data around '%s' at line number %d for individual %d, '%s'\nShould be a date of birth (mm/yyyy or .). Exiting.\n",strb, linei,i,line);fflush(ficlog);
1.225 brouard 8543: return 1;
1.136 brouard 8544: }
8545: if (year==9999) {
1.141 brouard 8546: printf("Error reading data around '%s' at line number %d for individual %d, '%s'\nShould be a date of birth (mm/yyyy) but at least the year of birth should be given. Exiting.\n",strb, linei,i,line);
8547: fprintf(ficlog,"Error reading data around '%s' at line number %d for individual %d, '%s'\nShould be a date of birth (mm/yyyy) but at least the year of birth should be given. Exiting.\n",strb, linei,i,line);fflush(ficlog);
1.225 brouard 8548: return 1;
8549:
1.136 brouard 8550: }
8551: annais[i]=(double)(year);
8552: moisnais[i]=(double)(month);
8553: strcpy(line,stra);
1.225 brouard 8554:
1.223 brouard 8555: /* Sample weight */
1.136 brouard 8556: cutv(stra, strb,line,' ');
8557: errno=0;
8558: dval=strtod(strb,&endptr);
8559: if( strb[0]=='\0' || (*endptr != '\0')){
1.141 brouard 8560: printf("Error reading data around '%f' at line number %d, \"%s\" for individual %d\nShould be a weight. Exiting.\n",dval, i,line,linei);
8561: fprintf(ficlog,"Error reading data around '%f' at line number %d, \"%s\" for individual %d\nShould be a weight. Exiting.\n",dval, i,line,linei);
1.136 brouard 8562: fflush(ficlog);
8563: return 1;
8564: }
8565: weight[i]=dval;
8566: strcpy(line,stra);
1.225 brouard 8567:
1.223 brouard 8568: for (iv=nqv;iv>=1;iv--){ /* Loop on fixed quantitative variables */
8569: cutv(stra, strb, line, ' ');
8570: if(strb[0]=='.') { /* Missing value */
1.225 brouard 8571: lval=-1;
1.223 brouard 8572: }else{
1.225 brouard 8573: errno=0;
8574: /* what_kind_of_number(strb); */
8575: dval=strtod(strb,&endptr);
8576: /* if(strb != endptr && *endptr == '\0') */
8577: /* dval=dlval; */
8578: /* if (errno == ERANGE && (lval == LONG_MAX || lval == LONG_MIN)) */
8579: if( strb[0]=='\0' || (*endptr != '\0')){
8580: printf("Error reading data around '%s' at line number %d for individual %d, '%s'\nShould be the %d th quantitative value (out of %d) constant for all waves. Setting maxwav=%d might be wrong. Exiting.\n", strb, linei,i,line, iv, nqv, maxwav);
8581: fprintf(ficlog,"Error reading data around '%s' at line number %d for individual %d, '%s'\nShould be the %d th quantitative value (out of %d) constant for all waves. Setting maxwav=%d might be wrong. Exiting.\n", strb, linei,i,line, iv, nqv, maxwav);fflush(ficlog);
8582: return 1;
8583: }
8584: coqvar[iv][i]=dval;
1.226 brouard 8585: covar[ncovcol+iv][i]=dval; /* including qvar in standard covar for performance reasons */
1.223 brouard 8586: }
8587: strcpy(line,stra);
8588: }/* end loop nqv */
1.136 brouard 8589:
1.223 brouard 8590: /* Covariate values */
1.136 brouard 8591: for (j=ncovcol;j>=1;j--){
8592: cutv(stra, strb,line,' ');
1.223 brouard 8593: if(strb[0]=='.') { /* Missing covariate value */
1.225 brouard 8594: lval=-1;
1.136 brouard 8595: }else{
1.225 brouard 8596: errno=0;
8597: lval=strtol(strb,&endptr,10);
8598: if( strb[0]=='\0' || (*endptr != '\0')){
8599: printf("Error reading data around '%ld' at line number %d for individual %d, '%s'\nShould be a covariate value (=0 for the reference or 1 for alternative). Exiting.\n",lval, linei,i, line);
8600: fprintf(ficlog,"Error reading data around '%ld' at line number %d for individual %d, '%s'\nShould be a covariate value (=0 for the reference or 1 for alternative). Exiting.\n",lval, linei,i, line);fflush(ficlog);
8601: return 1;
8602: }
1.136 brouard 8603: }
8604: if(lval <-1 || lval >1){
1.225 brouard 8605: printf("Error reading data around '%ld' at line number %d for individual %d, '%s'\n \
1.136 brouard 8606: Should be a value of %d(nth) covariate (0 should be the value for the reference and 1\n \
8607: for the alternative. IMaCh does not build design variables automatically, do it yourself.\n \
1.225 brouard 8608: For example, for multinomial values like 1, 2 and 3,\n \
8609: build V1=0 V2=0 for the reference value (1),\n \
8610: V1=1 V2=0 for (2) \n \
1.136 brouard 8611: and V1=0 V2=1 for (3). V1=1 V2=1 should not exist and the corresponding\n \
1.225 brouard 8612: output of IMaCh is often meaningless.\n \
1.136 brouard 8613: Exiting.\n",lval,linei, i,line,j);
1.225 brouard 8614: fprintf(ficlog,"Error reading data around '%ld' at line number %d for individual %d, '%s'\n \
1.136 brouard 8615: Should be a value of %d(nth) covariate (0 should be the value for the reference and 1\n \
8616: for the alternative. IMaCh does not build design variables automatically, do it yourself.\n \
1.225 brouard 8617: For example, for multinomial values like 1, 2 and 3,\n \
8618: build V1=0 V2=0 for the reference value (1),\n \
8619: V1=1 V2=0 for (2) \n \
1.136 brouard 8620: and V1=0 V2=1 for (3). V1=1 V2=1 should not exist and the corresponding\n \
1.225 brouard 8621: output of IMaCh is often meaningless.\n \
1.136 brouard 8622: Exiting.\n",lval,linei, i,line,j);fflush(ficlog);
1.225 brouard 8623: return 1;
1.136 brouard 8624: }
8625: covar[j][i]=(double)(lval);
8626: strcpy(line,stra);
8627: }
8628: lstra=strlen(stra);
1.225 brouard 8629:
1.136 brouard 8630: if(lstra > 9){ /* More than 2**32 or max of what printf can write with %ld */
8631: stratrunc = &(stra[lstra-9]);
8632: num[i]=atol(stratrunc);
8633: }
8634: else
8635: num[i]=atol(stra);
8636: /*if((s[2][i]==2) && (s[3][i]==-1)&&(s[4][i]==9)){
8637: printf("%ld %.lf %.lf %.lf %.lf/%.lf %.lf/%.lf %.lf/%.lf %d %.lf/%.lf %d %.lf/%.lf %d %.lf/%.lf %d\n",num[i],(covar[1][i]), (covar[2][i]),weight[i], (moisnais[i]), (annais[i]), (moisdc[i]), (andc[i]), (mint[1][i]), (anint[1][i]), (s[1][i]), (mint[2][i]), (anint[2][i]), (s[2][i]), (mint[3][i]), (anint[3][i]), (s[3][i]), (mint[4][i]), (anint[4][i]), (s[4][i])); ij=ij+1;}*/
8638:
8639: i=i+1;
8640: } /* End loop reading data */
1.225 brouard 8641:
1.136 brouard 8642: *imax=i-1; /* Number of individuals */
8643: fclose(fic);
1.225 brouard 8644:
1.136 brouard 8645: return (0);
1.164 brouard 8646: /* endread: */
1.225 brouard 8647: printf("Exiting readdata: ");
8648: fclose(fic);
8649: return (1);
1.223 brouard 8650: }
1.126 brouard 8651:
1.234 brouard 8652: void removefirstspace(char **stri){/*, char stro[]) {*/
1.230 brouard 8653: char *p1 = *stri, *p2 = *stri;
1.235 brouard 8654: while (*p2 == ' ')
1.234 brouard 8655: p2++;
8656: /* while ((*p1++ = *p2++) !=0) */
8657: /* ; */
8658: /* do */
8659: /* while (*p2 == ' ') */
8660: /* p2++; */
8661: /* while (*p1++ == *p2++); */
8662: *stri=p2;
1.145 brouard 8663: }
8664:
1.235 brouard 8665: int decoderesult ( char resultline[], int nres)
1.230 brouard 8666: /**< This routine decode one result line and returns the combination # of dummy covariates only **/
8667: {
1.235 brouard 8668: int j=0, k=0, k1=0, k2=0, k3=0, k4=0, match=0, k2q=0, k3q=0, k4q=0;
1.230 brouard 8669: char resultsav[MAXLINE];
1.234 brouard 8670: int resultmodel[MAXLINE];
8671: int modelresult[MAXLINE];
1.230 brouard 8672: char stra[80], strb[80], strc[80], strd[80],stre[80];
8673:
1.234 brouard 8674: removefirstspace(&resultline);
1.233 brouard 8675: printf("decoderesult:%s\n",resultline);
1.230 brouard 8676:
8677: if (strstr(resultline,"v") !=0){
8678: printf("Error. 'v' must be in upper case 'V' result: %s ",resultline);
8679: fprintf(ficlog,"Error. 'v' must be in upper case result: %s ",resultline);fflush(ficlog);
8680: return 1;
8681: }
8682: trimbb(resultsav, resultline);
8683: if (strlen(resultsav) >1){
8684: j=nbocc(resultsav,'='); /**< j=Number of covariate values'=' */
8685: }
1.253 brouard 8686: if(j == 0){ /* Resultline but no = */
8687: TKresult[nres]=0; /* Combination for the nresult and the model */
8688: return (0);
8689: }
8690:
1.234 brouard 8691: if( j != cptcovs ){ /* Be careful if a variable is in a product but not single */
8692: printf("ERROR: the number of variable in the resultline, %d, differs from the number of variable used in the model line, %d.\n",j, cptcovs);
8693: fprintf(ficlog,"ERROR: the number of variable in the resultline, %d, differs from the number of variable used in the model line, %d.\n",j, cptcovs);
8694: }
8695: for(k=1; k<=j;k++){ /* Loop on any covariate of the result line */
8696: if(nbocc(resultsav,'=') >1){
8697: cutl(stra,strb,resultsav,' '); /* keeps in strb after the first ' '
8698: resultsav= V4=1 V5=25.1 V3=0 strb=V3=0 stra= V4=1 V5=25.1 */
8699: cutl(strc,strd,strb,'='); /* strb:V4=1 strc=1 strd=V4 */
8700: }else
8701: cutl(strc,strd,resultsav,'=');
1.230 brouard 8702: Tvalsel[k]=atof(strc); /* 1 */
1.234 brouard 8703:
1.230 brouard 8704: cutl(strc,stre,strd,'V'); /* strd='V4' strc=4 stre='V' */;
8705: Tvarsel[k]=atoi(strc);
8706: /* Typevarsel[k]=1; /\* 1 for age product *\/ */
8707: /* cptcovsel++; */
8708: if (nbocc(stra,'=') >0)
8709: strcpy(resultsav,stra); /* and analyzes it */
8710: }
1.235 brouard 8711: /* Checking for missing or useless values in comparison of current model needs */
1.236 brouard 8712: for(k1=1; k1<= cptcovt ;k1++){ /* model line V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
8713: if(Typevar[k1]==0){ /* Single covariate in model */
1.234 brouard 8714: match=0;
1.236 brouard 8715: for(k2=1; k2 <=j;k2++){/* result line V4=1 V5=24.1 V3=1 V2=8 V1=0 */
1.237 brouard 8716: if(Tvar[k1]==Tvarsel[k2]) {/* Tvar[1]=5 == Tvarsel[2]=5 */
1.236 brouard 8717: modelresult[k2]=k1;/* modelresult[2]=1 modelresult[1]=2 modelresult[3]=3 modelresult[6]=4 modelresult[9]=5 */
1.234 brouard 8718: match=1;
8719: break;
8720: }
8721: }
8722: if(match == 0){
8723: printf("Error in result line: %d value missing; result: %s, model=%s\n",k1, resultline, model);
8724: }
8725: }
8726: }
1.235 brouard 8727: /* Checking for missing or useless values in comparison of current model needs */
8728: for(k2=1; k2 <=j;k2++){ /* result line V4=1 V5=24.1 V3=1 V2=8 V1=0 */
1.234 brouard 8729: match=0;
1.235 brouard 8730: for(k1=1; k1<= cptcovt ;k1++){ /* model line V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
8731: if(Typevar[k1]==0){ /* Single */
1.237 brouard 8732: if(Tvar[k1]==Tvarsel[k2]) { /* Tvar[2]=4 == Tvarsel[1]=4 */
1.235 brouard 8733: resultmodel[k1]=k2; /* resultmodel[2]=1 resultmodel[1]=2 resultmodel[3]=3 resultmodel[6]=4 resultmodel[9]=5 */
1.234 brouard 8734: ++match;
8735: }
8736: }
8737: }
8738: if(match == 0){
8739: printf("Error in result line: %d value missing; result: %s, model=%s\n",k1, resultline, model);
8740: }else if(match > 1){
8741: printf("Error in result line: %d doubled; result: %s, model=%s\n",k2, resultline, model);
8742: }
8743: }
1.235 brouard 8744:
1.234 brouard 8745: /* We need to deduce which combination number is chosen and save quantitative values */
1.235 brouard 8746: /* model line V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
8747: /* result line V4=1 V5=25.1 V3=0 V2=8 V1=1 */
8748: /* should give a combination of dummy V4=1, V3=0, V1=1 => V4*2**(0) + V3*2**(1) + V1*2**(2) = 5 + (1offset) = 6*/
8749: /* result line V4=1 V5=24.1 V3=1 V2=8 V1=0 */
8750: /* should give a combination of dummy V4=1, V3=1, V1=0 => V4*2**(0) + V3*2**(1) + V1*2**(2) = 3 + (1offset) = 4*/
8751: /* 1 0 0 0 */
8752: /* 2 1 0 0 */
8753: /* 3 0 1 0 */
8754: /* 4 1 1 0 */ /* V4=1, V3=1, V1=0 */
8755: /* 5 0 0 1 */
8756: /* 6 1 0 1 */ /* V4=1, V3=0, V1=1 */
8757: /* 7 0 1 1 */
8758: /* 8 1 1 1 */
1.237 brouard 8759: /* V(Tvresult)=Tresult V4=1 V3=0 V1=1 Tresult[nres=1][2]=0 */
8760: /* V(Tvqresult)=Tqresult V5=25.1 V2=8 Tqresult[nres=1][1]=25.1 */
8761: /* V5*age V5 known which value for nres? */
8762: /* Tqinvresult[2]=8 Tqinvresult[1]=25.1 */
1.235 brouard 8763: for(k1=1, k=0, k4=0, k4q=0; k1 <=cptcovt;k1++){ /* model line */
8764: if( Dummy[k1]==0 && Typevar[k1]==0 ){ /* Single dummy */
1.237 brouard 8765: k3= resultmodel[k1]; /* resultmodel[2(V4)] = 1=k3 */
1.235 brouard 8766: k2=(int)Tvarsel[k3]; /* Tvarsel[resultmodel[2]]= Tvarsel[1] = 4=k2 */
8767: k+=Tvalsel[k3]*pow(2,k4); /* Tvalsel[1]=1 */
1.237 brouard 8768: Tresult[nres][k4+1]=Tvalsel[k3];/* Tresult[nres][1]=1(V4=1) Tresult[nres][2]=0(V3=0) */
8769: Tvresult[nres][k4+1]=(int)Tvarsel[k3];/* Tvresult[nres][1]=4 Tvresult[nres][3]=1 */
8770: Tinvresult[nres][(int)Tvarsel[k3]]=Tvalsel[k3]; /* Tinvresult[nres][4]=1 */
1.235 brouard 8771: printf("Decoderesult Dummy k=%d, V(k2=V%d)= Tvalsel[%d]=%d, 2**(%d)\n",k, k2, k3, (int)Tvalsel[k3], k4);
8772: k4++;;
8773: } else if( Dummy[k1]==1 && Typevar[k1]==0 ){ /* Single quantitative */
8774: k3q= resultmodel[k1]; /* resultmodel[2] = 1=k3 */
8775: k2q=(int)Tvarsel[k3q]; /* Tvarsel[resultmodel[2]]= Tvarsel[1] = 4=k2 */
1.237 brouard 8776: Tqresult[nres][k4q+1]=Tvalsel[k3q]; /* Tqresult[nres][1]=25.1 */
8777: Tvqresult[nres][k4q+1]=(int)Tvarsel[k3q]; /* Tvqresult[nres][1]=5 */
8778: Tqinvresult[nres][(int)Tvarsel[k3q]]=Tvalsel[k3q]; /* Tqinvresult[nres][5]=25.1 */
1.235 brouard 8779: printf("Decoderesult Quantitative nres=%d, V(k2q=V%d)= Tvalsel[%d]=%d, Tvarsel[%d]=%f\n",nres, k2q, k3q, Tvarsel[k3q], k3q, Tvalsel[k3q]);
8780: k4q++;;
8781: }
8782: }
1.234 brouard 8783:
1.235 brouard 8784: TKresult[nres]=++k; /* Combination for the nresult and the model */
1.230 brouard 8785: return (0);
8786: }
1.235 brouard 8787:
1.230 brouard 8788: int decodemodel( char model[], int lastobs)
8789: /**< This routine decodes the model and returns:
1.224 brouard 8790: * Model V1+V2+V3+V8+V7*V8+V5*V6+V8*age+V3*age+age*age
8791: * - nagesqr = 1 if age*age in the model, otherwise 0.
8792: * - cptcovt total number of covariates of the model nbocc(+)+1 = 8 excepting constant and age and age*age
8793: * - cptcovn or number of covariates k of the models excluding age*products =6 and age*age
8794: * - cptcovage number of covariates with age*products =2
8795: * - cptcovs number of simple covariates
8796: * - Tvar[k] is the id of the kth covariate Tvar[1]@12 {1, 2, 3, 8, 10, 11, 8, 3, 7, 8, 5, 6}, thus Tvar[5=V7*V8]=10
8797: * which is a new column after the 9 (ncovcol) variables.
8798: * - if k is a product Vn*Vm covar[k][i] is filled with correct values for each individual
8799: * - Tprod[l] gives the kth covariates of the product Vn*Vm l=1 to cptcovprod-cptcovage
8800: * Tprod[1]@2 {5, 6}: position of first product V7*V8 is 5, and second V5*V6 is 6.
8801: * - Tvard[k] p Tvard[1][1]@4 {7, 8, 5, 6} for V7*V8 and V5*V6 .
8802: */
1.136 brouard 8803: {
1.238 brouard 8804: int i, j, k, ks, v;
1.227 brouard 8805: int j1, k1, k2, k3, k4;
1.136 brouard 8806: char modelsav[80];
1.145 brouard 8807: char stra[80], strb[80], strc[80], strd[80],stre[80];
1.187 brouard 8808: char *strpt;
1.136 brouard 8809:
1.145 brouard 8810: /*removespace(model);*/
1.136 brouard 8811: if (strlen(model) >1){ /* If there is at least 1 covariate */
1.145 brouard 8812: j=0, j1=0, k1=0, k2=-1, ks=0, cptcovn=0;
1.137 brouard 8813: if (strstr(model,"AGE") !=0){
1.192 brouard 8814: printf("Error. AGE must be in lower case 'age' model=1+age+%s. ",model);
8815: fprintf(ficlog,"Error. AGE must be in lower case model=1+age+%s. ",model);fflush(ficlog);
1.136 brouard 8816: return 1;
8817: }
1.141 brouard 8818: if (strstr(model,"v") !=0){
8819: printf("Error. 'v' must be in upper case 'V' model=%s ",model);
8820: fprintf(ficlog,"Error. 'v' must be in upper case model=%s ",model);fflush(ficlog);
8821: return 1;
8822: }
1.187 brouard 8823: strcpy(modelsav,model);
8824: if ((strpt=strstr(model,"age*age")) !=0){
8825: printf(" strpt=%s, model=%s\n",strpt, model);
8826: if(strpt != model){
1.234 brouard 8827: printf("Error in model: 'model=%s'; 'age*age' should in first place before other covariates\n \
1.192 brouard 8828: 'model=1+age+age*age+V1.' or 'model=1+age+age*age+V1+V1*age.', please swap as well as \n \
1.187 brouard 8829: corresponding column of parameters.\n",model);
1.234 brouard 8830: fprintf(ficlog,"Error in model: 'model=%s'; 'age*age' should in first place before other covariates\n \
1.192 brouard 8831: 'model=1+age+age*age+V1.' or 'model=1+age+age*age+V1+V1*age.', please swap as well as \n \
1.187 brouard 8832: corresponding column of parameters.\n",model); fflush(ficlog);
1.234 brouard 8833: return 1;
1.225 brouard 8834: }
1.187 brouard 8835: nagesqr=1;
8836: if (strstr(model,"+age*age") !=0)
1.234 brouard 8837: substrchaine(modelsav, model, "+age*age");
1.187 brouard 8838: else if (strstr(model,"age*age+") !=0)
1.234 brouard 8839: substrchaine(modelsav, model, "age*age+");
1.187 brouard 8840: else
1.234 brouard 8841: substrchaine(modelsav, model, "age*age");
1.187 brouard 8842: }else
8843: nagesqr=0;
8844: if (strlen(modelsav) >1){
8845: j=nbocc(modelsav,'+'); /**< j=Number of '+' */
8846: j1=nbocc(modelsav,'*'); /**< j1=Number of '*' */
1.224 brouard 8847: cptcovs=j+1-j1; /**< Number of simple covariates V1+V1*age+V3 +V3*V4+age*age=> V1 + V3 =5-3=2 */
1.187 brouard 8848: cptcovt= j+1; /* Number of total covariates in the model, not including
1.225 brouard 8849: * cst, age and age*age
8850: * V1+V1*age+ V3 + V3*V4+age*age=> 3+1=4*/
8851: /* including age products which are counted in cptcovage.
8852: * but the covariates which are products must be treated
8853: * separately: ncovn=4- 2=2 (V1+V3). */
1.187 brouard 8854: cptcovprod=j1; /**< Number of products V1*V2 +v3*age = 2 */
8855: cptcovprodnoage=0; /**< Number of covariate products without age: V3*V4 =1 */
1.225 brouard 8856:
8857:
1.187 brouard 8858: /* Design
8859: * V1 V2 V3 V4 V5 V6 V7 V8 V9 Weight
8860: * < ncovcol=8 >
8861: * Model V2 + V1 + V3*age + V3 + V5*V6 + V7*V8 + V8*age + V8
8862: * k= 1 2 3 4 5 6 7 8
8863: * cptcovn number of covariates (not including constant and age ) = # of + plus 1 = 7+1=8
8864: * covar[k,i], value of kth covariate if not including age for individual i:
1.224 brouard 8865: * covar[1][i]= (V1), covar[4][i]=(V4), covar[8][i]=(V8)
8866: * Tvar[k] # of the kth covariate: Tvar[1]=2 Tvar[2]=1 Tvar[4]=3 Tvar[8]=8
1.187 brouard 8867: * if multiplied by age: V3*age Tvar[3=V3*age]=3 (V3) Tvar[7]=8 and
8868: * Tage[++cptcovage]=k
8869: * if products, new covar are created after ncovcol with k1
8870: * Tvar[k]=ncovcol+k1; # of the kth covariate product: Tvar[5]=ncovcol+1=10 Tvar[6]=ncovcol+1=11
8871: * Tprod[k1]=k; Tprod[1]=5 Tprod[2]= 6; gives the position of the k1th product
8872: * Tvard[k1][1]=m Tvard[k1][2]=m; Tvard[1][1]=5 (V5) Tvard[1][2]=6 Tvard[2][1]=7 (V7) Tvard[2][2]=8
8873: * Tvar[cptcovn+k2]=Tvard[k1][1];Tvar[cptcovn+k2+1]=Tvard[k1][2];
8874: * Tvar[8+1]=5;Tvar[8+2]=6;Tvar[8+3]=7;Tvar[8+4]=8 inverted
8875: * V1 V2 V3 V4 V5 V6 V7 V8 V9 V10 V11
8876: * < ncovcol=8 >
8877: * Model V2 + V1 + V3*age + V3 + V5*V6 + V7*V8 + V8*age + V8 d1 d1 d2 d2
8878: * k= 1 2 3 4 5 6 7 8 9 10 11 12
8879: * Tvar[k]= 2 1 3 3 10 11 8 8 5 6 7 8
8880: * p Tvar[1]@12={2, 1, 3, 3, 11, 10, 8, 8, 7, 8, 5, 6}
8881: * p Tprod[1]@2={ 6, 5}
8882: *p Tvard[1][1]@4= {7, 8, 5, 6}
8883: * covar[k][i]= V2 V1 ? V3 V5*V6? V7*V8? ? V8
8884: * cov[Tage[kk]+2]=covar[Tvar[Tage[kk]]][i]*cov[2];
8885: *How to reorganize?
8886: * Model V1 + V2 + V3 + V8 + V5*V6 + V7*V8 + V3*age + V8*age
8887: * Tvars {2, 1, 3, 3, 11, 10, 8, 8, 7, 8, 5, 6}
8888: * {2, 1, 4, 8, 5, 6, 3, 7}
8889: * Struct []
8890: */
1.225 brouard 8891:
1.187 brouard 8892: /* This loop fills the array Tvar from the string 'model'.*/
8893: /* j is the number of + signs in the model V1+V2+V3 j=2 i=3 to 1 */
8894: /* modelsav=V2+V1+V4+age*V3 strb=age*V3 stra=V2+V1+V4 */
8895: /* k=4 (age*V3) Tvar[k=4]= 3 (from V3) Tage[cptcovage=1]=4 */
8896: /* k=3 V4 Tvar[k=3]= 4 (from V4) */
8897: /* k=2 V1 Tvar[k=2]= 1 (from V1) */
8898: /* k=1 Tvar[1]=2 (from V2) */
8899: /* k=5 Tvar[5] */
8900: /* for (k=1; k<=cptcovn;k++) { */
1.198 brouard 8901: /* cov[2+k]=nbcode[Tvar[k]][codtabm(ij,Tvar[k])]; */
1.187 brouard 8902: /* } */
1.198 brouard 8903: /* for (k=1; k<=cptcovage;k++) cov[2+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,Tvar[Tage[k])]]*cov[2]; */
1.187 brouard 8904: /*
8905: * Treating invertedly V2+V1+V3*age+V2*V4 is as if written V2*V4 +V3*age + V1 + V2 */
1.227 brouard 8906: for(k=cptcovt; k>=1;k--){ /**< Number of covariates not including constant and age, neither age*age*/
8907: Tvar[k]=0; Tprod[k]=0; Tposprod[k]=0;
8908: }
1.187 brouard 8909: cptcovage=0;
8910: for(k=1; k<=cptcovt;k++){ /* Loop on total covariates of the model */
1.234 brouard 8911: cutl(stra,strb,modelsav,'+'); /* keeps in strb after the first '+'
1.225 brouard 8912: modelsav==V2+V1+V4+V3*age strb=V3*age stra=V2+V1+V4 */
1.234 brouard 8913: if (nbocc(modelsav,'+')==0) strcpy(strb,modelsav); /* and analyzes it */
8914: /* printf("i=%d a=%s b=%s sav=%s\n",i, stra,strb,modelsav);*/
8915: /*scanf("%d",i);*/
8916: if (strchr(strb,'*')) { /**< Model includes a product V2+V1+V4+V3*age strb=V3*age */
8917: cutl(strc,strd,strb,'*'); /**< strd*strc Vm*Vn: strb=V3*age(input) strc=age strd=V3 ; V3*V2 strc=V2, strd=V3 */
8918: if (strcmp(strc,"age")==0) { /**< Model includes age: Vn*age */
8919: /* covar is not filled and then is empty */
8920: cptcovprod--;
8921: cutl(stre,strb,strd,'V'); /* strd=V3(input): stre="3" */
8922: Tvar[k]=atoi(stre); /* V2+V1+V4+V3*age Tvar[4]=3 ; V1+V2*age Tvar[2]=2; V1+V1*age Tvar[2]=1 */
8923: Typevar[k]=1; /* 1 for age product */
8924: cptcovage++; /* Sums the number of covariates which include age as a product */
8925: Tage[cptcovage]=k; /* Tvar[4]=3, Tage[1] = 4 or V1+V1*age Tvar[2]=1, Tage[1]=2 */
8926: /*printf("stre=%s ", stre);*/
8927: } else if (strcmp(strd,"age")==0) { /* or age*Vn */
8928: cptcovprod--;
8929: cutl(stre,strb,strc,'V');
8930: Tvar[k]=atoi(stre);
8931: Typevar[k]=1; /* 1 for age product */
8932: cptcovage++;
8933: Tage[cptcovage]=k;
8934: } else { /* Age is not in the model product V2+V1+V1*V4+V3*age+V3*V2 strb=V3*V2*/
8935: /* loops on k1=1 (V3*V2) and k1=2 V4*V3 */
8936: cptcovn++;
8937: cptcovprodnoage++;k1++;
8938: cutl(stre,strb,strc,'V'); /* strc= Vn, stre is n; strb=V3*V2 stre=3 strc=*/
8939: Tvar[k]=ncovcol+nqv+ntv+nqtv+k1; /* For model-covariate k tells which data-covariate to use but
8940: because this model-covariate is a construction we invent a new column
8941: which is after existing variables ncovcol+nqv+ntv+nqtv + k1
8942: If already ncovcol=4 and model=V2+V1+V1*V4+age*V3+V3*V2
8943: Tvar[3=V1*V4]=4+1 Tvar[5=V3*V2]=4 + 2= 6, etc */
8944: Typevar[k]=2; /* 2 for double fixed dummy covariates */
8945: cutl(strc,strb,strd,'V'); /* strd was Vm, strc is m */
8946: Tprod[k1]=k; /* Tprod[1]=3(=V1*V4) for V2+V1+V1*V4+age*V3+V3*V2 */
8947: Tposprod[k]=k1; /* Tpsprod[3]=1, Tposprod[2]=5 */
8948: Tvard[k1][1] =atoi(strc); /* m 1 for V1*/
8949: Tvard[k1][2] =atoi(stre); /* n 4 for V4*/
8950: k2=k2+2; /* k2 is initialize to -1, We want to store the n and m in Vn*Vm at the end of Tvar */
8951: /* Tvar[cptcovt+k2]=Tvard[k1][1]; /\* Tvar[(cptcovt=4+k2=1)=5]= 1 (V1) *\/ */
8952: /* Tvar[cptcovt+k2+1]=Tvard[k1][2]; /\* Tvar[(cptcovt=4+(k2=1)+1)=6]= 4 (V4) *\/ */
1.225 brouard 8953: /*ncovcol=4 and model=V2+V1+V1*V4+age*V3+V3*V2, Tvar[3]=5, Tvar[4]=6, cptcovt=5 */
1.234 brouard 8954: /* 1 2 3 4 5 | Tvar[5+1)=1, Tvar[7]=2 */
8955: for (i=1; i<=lastobs;i++){
8956: /* Computes the new covariate which is a product of
8957: covar[n][i]* covar[m][i] and stores it at ncovol+k1 May not be defined */
8958: covar[ncovcol+k1][i]=covar[atoi(stre)][i]*covar[atoi(strc)][i];
8959: }
8960: } /* End age is not in the model */
8961: } /* End if model includes a product */
8962: else { /* no more sum */
8963: /*printf("d=%s c=%s b=%s\n", strd,strc,strb);*/
8964: /* scanf("%d",i);*/
8965: cutl(strd,strc,strb,'V');
8966: ks++; /**< Number of simple covariates dummy or quantitative, fixe or varying */
8967: cptcovn++; /** V4+V3+V5: V4 and V3 timevarying dummy covariates, V5 timevarying quantitative */
8968: Tvar[k]=atoi(strd);
8969: Typevar[k]=0; /* 0 for simple covariates */
8970: }
8971: strcpy(modelsav,stra); /* modelsav=V2+V1+V4 stra=V2+V1+V4 */
1.223 brouard 8972: /*printf("a=%s b=%s sav=%s\n", stra,strb,modelsav);
1.225 brouard 8973: scanf("%d",i);*/
1.187 brouard 8974: } /* end of loop + on total covariates */
8975: } /* end if strlen(modelsave == 0) age*age might exist */
8976: } /* end if strlen(model == 0) */
1.136 brouard 8977:
8978: /*The number n of Vn is stored in Tvar. cptcovage =number of age covariate. Tage gives the position of age. cptcovprod= number of products.
8979: If model=V1+V1*age then Tvar[1]=1 Tvar[2]=1 cptcovage=1 Tage[1]=2 cptcovprod=0*/
1.225 brouard 8980:
1.136 brouard 8981: /* printf("tvar1=%d tvar2=%d tvar3=%d cptcovage=%d Tage=%d",Tvar[1],Tvar[2],Tvar[3],cptcovage,Tage[1]);
1.225 brouard 8982: printf("cptcovprod=%d ", cptcovprod);
8983: fprintf(ficlog,"cptcovprod=%d ", cptcovprod);
8984: scanf("%d ",i);*/
8985:
8986:
1.230 brouard 8987: /* Until here, decodemodel knows only the grammar (simple, product, age*) of the model but not what kind
8988: of variable (dummy vs quantitative, fixed vs time varying) is behind. But we know the # of each. */
1.226 brouard 8989: /* ncovcol= 1, nqv=1 | ntv=2, nqtv= 1 = 5 possible variables data: 2 fixed 3, varying
8990: model= V5 + V4 +V3 + V4*V3 + V5*age + V2 + V1*V2 + V1*age + V5*age, V1 is not used saving its place
8991: k = 1 2 3 4 5 6 7 8 9
8992: Tvar[k]= 5 4 3 1+1+2+1+1=6 5 2 7 1 5
8993: Typevar[k]= 0 0 0 2 1 0 2 1 1
1.227 brouard 8994: Fixed[k] 1 1 1 1 3 0 0 or 2 2 3
8995: Dummy[k] 1 0 0 0 3 1 1 2 3
8996: Tmodelind[combination of covar]=k;
1.225 brouard 8997: */
8998: /* Dispatching between quantitative and time varying covariates */
1.226 brouard 8999: /* If Tvar[k] >ncovcol it is a product */
1.225 brouard 9000: /* Tvar[k] is the value n of Vn with n varying for 1 to nvcol, or p Vp=Vn*Vm for product */
1.226 brouard 9001: /* Computing effective variables, ie used by the model, that is from the cptcovt variables */
1.227 brouard 9002: printf("Model=%s\n\
9003: Typevar: 0 for simple covariate (dummy, quantitative, fixed or varying), 1 for age product, 2 for product \n\
9004: Fixed[k] 0=fixed (product or simple), 1 varying, 2 fixed with age product, 3 varying with age product \n\
9005: Dummy[k] 0=dummy (0 1), 1 quantitative (single or product without age), 2 dummy with age product, 3 quant with age product\n",model);
9006: fprintf(ficlog,"Model=%s\n\
9007: Typevar: 0 for simple covariate (dummy, quantitative, fixed or varying), 1 for age product, 2 for product \n\
9008: Fixed[k] 0=fixed (product or simple), 1 varying, 2 fixed with age product, 3 varying with age product \n\
9009: Dummy[k] 0=dummy (0 1), 1 quantitative (single or product without age), 2 dummy with age product, 3 quant with age product\n",model);
1.240 brouard 9010: for(k=1;k<=cptcovt; k++){ Fixed[k]=0; Dummy[k]=0;}
1.234 brouard 9011: for(k=1, ncovf=0, nsd=0, nsq=0, ncovv=0, ncova=0, ncoveff=0, nqfveff=0, ntveff=0, nqtveff=0;k<=cptcovt; k++){ /* or cptocvt */
9012: if (Tvar[k] <=ncovcol && Typevar[k]==0 ){ /* Simple fixed dummy (<=ncovcol) covariates */
1.227 brouard 9013: Fixed[k]= 0;
9014: Dummy[k]= 0;
1.225 brouard 9015: ncoveff++;
1.232 brouard 9016: ncovf++;
1.234 brouard 9017: nsd++;
9018: modell[k].maintype= FTYPE;
9019: TvarsD[nsd]=Tvar[k];
9020: TvarsDind[nsd]=k;
9021: TvarF[ncovf]=Tvar[k];
9022: TvarFind[ncovf]=k;
9023: TvarFD[ncoveff]=Tvar[k]; /* TvarFD[1]=V1 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
9024: TvarFDind[ncoveff]=k; /* TvarFDind[1]=9 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
9025: }else if( Tvar[k] <=ncovcol && Typevar[k]==2){ /* Product of fixed dummy (<=ncovcol) covariates */
9026: Fixed[k]= 0;
9027: Dummy[k]= 0;
9028: ncoveff++;
9029: ncovf++;
9030: modell[k].maintype= FTYPE;
9031: TvarF[ncovf]=Tvar[k];
9032: TvarFind[ncovf]=k;
1.230 brouard 9033: TvarFD[ncoveff]=Tvar[k]; /* TvarFD[1]=V1 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1.231 brouard 9034: TvarFDind[ncoveff]=k; /* TvarFDind[1]=9 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1.240 brouard 9035: }else if( Tvar[k] <=ncovcol+nqv && Typevar[k]==0){/* Remind that product Vn*Vm are added in k Only simple fixed quantitative variable */
1.227 brouard 9036: Fixed[k]= 0;
9037: Dummy[k]= 1;
1.230 brouard 9038: nqfveff++;
1.234 brouard 9039: modell[k].maintype= FTYPE;
9040: modell[k].subtype= FQ;
9041: nsq++;
9042: TvarsQ[nsq]=Tvar[k];
9043: TvarsQind[nsq]=k;
1.232 brouard 9044: ncovf++;
1.234 brouard 9045: TvarF[ncovf]=Tvar[k];
9046: TvarFind[ncovf]=k;
1.231 brouard 9047: TvarFQ[nqfveff]=Tvar[k]-ncovcol; /* TvarFQ[1]=V2-1=1st in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple fixed quantitative variable */
1.230 brouard 9048: TvarFQind[nqfveff]=k; /* TvarFQind[1]=6 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple fixed quantitative variable */
1.242 brouard 9049: }else if( Tvar[k] <=ncovcol+nqv+ntv && Typevar[k]==0){/* Only simple time varying dummy variables */
1.227 brouard 9050: Fixed[k]= 1;
9051: Dummy[k]= 0;
1.225 brouard 9052: ntveff++; /* Only simple time varying dummy variable */
1.234 brouard 9053: modell[k].maintype= VTYPE;
9054: modell[k].subtype= VD;
9055: nsd++;
9056: TvarsD[nsd]=Tvar[k];
9057: TvarsDind[nsd]=k;
9058: ncovv++; /* Only simple time varying variables */
9059: TvarV[ncovv]=Tvar[k];
1.242 brouard 9060: TvarVind[ncovv]=k; /* TvarVind[2]=2 TvarVind[3]=3 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Any time varying singele */
1.231 brouard 9061: TvarVD[ntveff]=Tvar[k]; /* TvarVD[1]=V4 TvarVD[2]=V3 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple time varying dummy variable */
9062: TvarVDind[ntveff]=k; /* TvarVDind[1]=2 TvarVDind[2]=3 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple time varying dummy variable */
1.228 brouard 9063: printf("Quasi Tmodelind[%d]=%d,Tvar[Tmodelind[%d]]=V%d, ncovcol=%d, nqv=%d,Tvar[k]- ncovcol-nqv=%d\n",ntveff,k,ntveff,Tvar[k], ncovcol, nqv,Tvar[k]- ncovcol-nqv);
9064: printf("Quasi TmodelInvind[%d]=%d\n",k,Tvar[k]- ncovcol-nqv);
1.231 brouard 9065: }else if( Tvar[k] <=ncovcol+nqv+ntv+nqtv && Typevar[k]==0){ /* Only simple time varying quantitative variable V5*/
1.234 brouard 9066: Fixed[k]= 1;
9067: Dummy[k]= 1;
9068: nqtveff++;
9069: modell[k].maintype= VTYPE;
9070: modell[k].subtype= VQ;
9071: ncovv++; /* Only simple time varying variables */
9072: nsq++;
9073: TvarsQ[nsq]=Tvar[k];
9074: TvarsQind[nsq]=k;
9075: TvarV[ncovv]=Tvar[k];
1.242 brouard 9076: TvarVind[ncovv]=k; /* TvarVind[1]=1 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Any time varying singele */
1.231 brouard 9077: TvarVQ[nqtveff]=Tvar[k]; /* TvarVQ[1]=V5 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple time varying quantitative variable */
9078: TvarVQind[nqtveff]=k; /* TvarVQind[1]=1 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple time varying quantitative variable */
1.234 brouard 9079: TmodelInvQind[nqtveff]=Tvar[k]- ncovcol-nqv-ntv;/* Only simple time varying quantitative variable */
9080: /* Tmodeliqind[k]=nqtveff;/\* Only simple time varying quantitative variable *\/ */
9081: printf("Quasi TmodelQind[%d]=%d,Tvar[TmodelQind[%d]]=V%d, ncovcol=%d, nqv=%d, ntv=%d,Tvar[k]- ncovcol-nqv-ntv=%d\n",nqtveff,k,nqtveff,Tvar[k], ncovcol, nqv, ntv, Tvar[k]- ncovcol-nqv-ntv);
1.228 brouard 9082: printf("Quasi TmodelInvQind[%d]=%d\n",k,Tvar[k]- ncovcol-nqv-ntv);
1.227 brouard 9083: }else if (Typevar[k] == 1) { /* product with age */
1.234 brouard 9084: ncova++;
9085: TvarA[ncova]=Tvar[k];
9086: TvarAind[ncova]=k;
1.231 brouard 9087: if (Tvar[k] <=ncovcol ){ /* Product age with fixed dummy covariatee */
1.240 brouard 9088: Fixed[k]= 2;
9089: Dummy[k]= 2;
9090: modell[k].maintype= ATYPE;
9091: modell[k].subtype= APFD;
9092: /* ncoveff++; */
1.227 brouard 9093: }else if( Tvar[k] <=ncovcol+nqv) { /* Remind that product Vn*Vm are added in k*/
1.240 brouard 9094: Fixed[k]= 2;
9095: Dummy[k]= 3;
9096: modell[k].maintype= ATYPE;
9097: modell[k].subtype= APFQ; /* Product age * fixed quantitative */
9098: /* nqfveff++; /\* Only simple fixed quantitative variable *\/ */
1.227 brouard 9099: }else if( Tvar[k] <=ncovcol+nqv+ntv ){
1.240 brouard 9100: Fixed[k]= 3;
9101: Dummy[k]= 2;
9102: modell[k].maintype= ATYPE;
9103: modell[k].subtype= APVD; /* Product age * varying dummy */
9104: /* ntveff++; /\* Only simple time varying dummy variable *\/ */
1.227 brouard 9105: }else if( Tvar[k] <=ncovcol+nqv+ntv+nqtv){
1.240 brouard 9106: Fixed[k]= 3;
9107: Dummy[k]= 3;
9108: modell[k].maintype= ATYPE;
9109: modell[k].subtype= APVQ; /* Product age * varying quantitative */
9110: /* nqtveff++;/\* Only simple time varying quantitative variable *\/ */
1.227 brouard 9111: }
9112: }else if (Typevar[k] == 2) { /* product without age */
9113: k1=Tposprod[k];
9114: if(Tvard[k1][1] <=ncovcol){
1.240 brouard 9115: if(Tvard[k1][2] <=ncovcol){
9116: Fixed[k]= 1;
9117: Dummy[k]= 0;
9118: modell[k].maintype= FTYPE;
9119: modell[k].subtype= FPDD; /* Product fixed dummy * fixed dummy */
9120: ncovf++; /* Fixed variables without age */
9121: TvarF[ncovf]=Tvar[k];
9122: TvarFind[ncovf]=k;
9123: }else if(Tvard[k1][2] <=ncovcol+nqv){
9124: Fixed[k]= 0; /* or 2 ?*/
9125: Dummy[k]= 1;
9126: modell[k].maintype= FTYPE;
9127: modell[k].subtype= FPDQ; /* Product fixed dummy * fixed quantitative */
9128: ncovf++; /* Varying variables without age */
9129: TvarF[ncovf]=Tvar[k];
9130: TvarFind[ncovf]=k;
9131: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv){
9132: Fixed[k]= 1;
9133: Dummy[k]= 0;
9134: modell[k].maintype= VTYPE;
9135: modell[k].subtype= VPDD; /* Product fixed dummy * varying dummy */
9136: ncovv++; /* Varying variables without age */
9137: TvarV[ncovv]=Tvar[k];
9138: TvarVind[ncovv]=k;
9139: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv+nqtv){
9140: Fixed[k]= 1;
9141: Dummy[k]= 1;
9142: modell[k].maintype= VTYPE;
9143: modell[k].subtype= VPDQ; /* Product fixed dummy * varying quantitative */
9144: ncovv++; /* Varying variables without age */
9145: TvarV[ncovv]=Tvar[k];
9146: TvarVind[ncovv]=k;
9147: }
1.227 brouard 9148: }else if(Tvard[k1][1] <=ncovcol+nqv){
1.240 brouard 9149: if(Tvard[k1][2] <=ncovcol){
9150: Fixed[k]= 0; /* or 2 ?*/
9151: Dummy[k]= 1;
9152: modell[k].maintype= FTYPE;
9153: modell[k].subtype= FPDQ; /* Product fixed quantitative * fixed dummy */
9154: ncovf++; /* Fixed variables without age */
9155: TvarF[ncovf]=Tvar[k];
9156: TvarFind[ncovf]=k;
9157: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv){
9158: Fixed[k]= 1;
9159: Dummy[k]= 1;
9160: modell[k].maintype= VTYPE;
9161: modell[k].subtype= VPDQ; /* Product fixed quantitative * varying dummy */
9162: ncovv++; /* Varying variables without age */
9163: TvarV[ncovv]=Tvar[k];
9164: TvarVind[ncovv]=k;
9165: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv+nqtv){
9166: Fixed[k]= 1;
9167: Dummy[k]= 1;
9168: modell[k].maintype= VTYPE;
9169: modell[k].subtype= VPQQ; /* Product fixed quantitative * varying quantitative */
9170: ncovv++; /* Varying variables without age */
9171: TvarV[ncovv]=Tvar[k];
9172: TvarVind[ncovv]=k;
9173: ncovv++; /* Varying variables without age */
9174: TvarV[ncovv]=Tvar[k];
9175: TvarVind[ncovv]=k;
9176: }
1.227 brouard 9177: }else if(Tvard[k1][1] <=ncovcol+nqv+ntv){
1.240 brouard 9178: if(Tvard[k1][2] <=ncovcol){
9179: Fixed[k]= 1;
9180: Dummy[k]= 1;
9181: modell[k].maintype= VTYPE;
9182: modell[k].subtype= VPDD; /* Product time varying dummy * fixed dummy */
9183: ncovv++; /* Varying variables without age */
9184: TvarV[ncovv]=Tvar[k];
9185: TvarVind[ncovv]=k;
9186: }else if(Tvard[k1][2] <=ncovcol+nqv){
9187: Fixed[k]= 1;
9188: Dummy[k]= 1;
9189: modell[k].maintype= VTYPE;
9190: modell[k].subtype= VPDQ; /* Product time varying dummy * fixed quantitative */
9191: ncovv++; /* Varying variables without age */
9192: TvarV[ncovv]=Tvar[k];
9193: TvarVind[ncovv]=k;
9194: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv){
9195: Fixed[k]= 1;
9196: Dummy[k]= 0;
9197: modell[k].maintype= VTYPE;
9198: modell[k].subtype= VPDD; /* Product time varying dummy * time varying dummy */
9199: ncovv++; /* Varying variables without age */
9200: TvarV[ncovv]=Tvar[k];
9201: TvarVind[ncovv]=k;
9202: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv+nqtv){
9203: Fixed[k]= 1;
9204: Dummy[k]= 1;
9205: modell[k].maintype= VTYPE;
9206: modell[k].subtype= VPDQ; /* Product time varying dummy * time varying quantitative */
9207: ncovv++; /* Varying variables without age */
9208: TvarV[ncovv]=Tvar[k];
9209: TvarVind[ncovv]=k;
9210: }
1.227 brouard 9211: }else if(Tvard[k1][1] <=ncovcol+nqv+ntv+nqtv){
1.240 brouard 9212: if(Tvard[k1][2] <=ncovcol){
9213: Fixed[k]= 1;
9214: Dummy[k]= 1;
9215: modell[k].maintype= VTYPE;
9216: modell[k].subtype= VPDQ; /* Product time varying quantitative * fixed dummy */
9217: ncovv++; /* Varying variables without age */
9218: TvarV[ncovv]=Tvar[k];
9219: TvarVind[ncovv]=k;
9220: }else if(Tvard[k1][2] <=ncovcol+nqv){
9221: Fixed[k]= 1;
9222: Dummy[k]= 1;
9223: modell[k].maintype= VTYPE;
9224: modell[k].subtype= VPQQ; /* Product time varying quantitative * fixed quantitative */
9225: ncovv++; /* Varying variables without age */
9226: TvarV[ncovv]=Tvar[k];
9227: TvarVind[ncovv]=k;
9228: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv){
9229: Fixed[k]= 1;
9230: Dummy[k]= 1;
9231: modell[k].maintype= VTYPE;
9232: modell[k].subtype= VPDQ; /* Product time varying quantitative * time varying dummy */
9233: ncovv++; /* Varying variables without age */
9234: TvarV[ncovv]=Tvar[k];
9235: TvarVind[ncovv]=k;
9236: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv+nqtv){
9237: Fixed[k]= 1;
9238: Dummy[k]= 1;
9239: modell[k].maintype= VTYPE;
9240: modell[k].subtype= VPQQ; /* Product time varying quantitative * time varying quantitative */
9241: ncovv++; /* Varying variables without age */
9242: TvarV[ncovv]=Tvar[k];
9243: TvarVind[ncovv]=k;
9244: }
1.227 brouard 9245: }else{
1.240 brouard 9246: printf("Error unknown type of covariate: Tvard[%d][1]=%d,Tvard[%d][2]=%d\n",k1,Tvard[k1][1],k1,Tvard[k1][2]);
9247: fprintf(ficlog,"Error unknown type of covariate: Tvard[%d][1]=%d,Tvard[%d][2]=%d\n",k1,Tvard[k1][1],k1,Tvard[k1][2]);
9248: } /*end k1*/
1.225 brouard 9249: }else{
1.226 brouard 9250: printf("Error, current version can't treat for performance reasons, Tvar[%d]=%d, Typevar[%d]=%d\n", k, Tvar[k], k, Typevar[k]);
9251: fprintf(ficlog,"Error, current version can't treat for performance reasons, Tvar[%d]=%d, Typevar[%d]=%d\n", k, Tvar[k], k, Typevar[k]);
1.225 brouard 9252: }
1.227 brouard 9253: printf("Decodemodel, k=%d, Tvar[%d]=V%d,Typevar=%d, Fixed=%d, Dummy=%d\n",k, k,Tvar[k],Typevar[k],Fixed[k],Dummy[k]);
1.231 brouard 9254: printf(" modell[%d].maintype=%d, modell[%d].subtype=%d\n",k,modell[k].maintype,k,modell[k].subtype);
1.227 brouard 9255: fprintf(ficlog,"Decodemodel, k=%d, Tvar[%d]=V%d,Typevar=%d, Fixed=%d, Dummy=%d\n",k, k,Tvar[k],Typevar[k],Fixed[k],Dummy[k]);
9256: }
9257: /* Searching for doublons in the model */
9258: for(k1=1; k1<= cptcovt;k1++){
9259: for(k2=1; k2 <k1;k2++){
9260: if((Typevar[k1]==Typevar[k2]) && (Fixed[Tvar[k1]]==Fixed[Tvar[k2]]) && (Dummy[Tvar[k1]]==Dummy[Tvar[k2]] )){
1.234 brouard 9261: if((Typevar[k1] == 0 || Typevar[k1] == 1)){ /* Simple or age product */
9262: if(Tvar[k1]==Tvar[k2]){
9263: printf("Error duplication in the model=%s at positions (+) %d and %d, Tvar[%d]=V%d, Tvar[%d]=V%d, Typevar=%d, Fixed=%d, Dummy=%d\n", model, k1,k2, k1, Tvar[k1], k2, Tvar[k2],Typevar[k1],Fixed[Tvar[k1]],Dummy[Tvar[k1]]);
9264: fprintf(ficlog,"Error duplication in the model=%s at positions (+) %d and %d, Tvar[%d]=V%d, Tvar[%d]=V%d, Typevar=%d, Fixed=%d, Dummy=%d\n", model, k1,k2, k1, Tvar[k1], k2, Tvar[k2],Typevar[k1],Fixed[Tvar[k1]],Dummy[Tvar[k1]]); fflush(ficlog);
9265: return(1);
9266: }
9267: }else if (Typevar[k1] ==2){
9268: k3=Tposprod[k1];
9269: k4=Tposprod[k2];
9270: if( ((Tvard[k3][1]== Tvard[k4][1])&&(Tvard[k3][2]== Tvard[k4][2])) || ((Tvard[k3][1]== Tvard[k4][2])&&(Tvard[k3][2]== Tvard[k4][1])) ){
9271: printf("Error duplication in the model=%s at positions (+) %d and %d, V%d*V%d, Typevar=%d, Fixed=%d, Dummy=%d\n",model, k1,k2, Tvard[k3][1], Tvard[k3][2],Typevar[k1],Fixed[Tvar[k1]],Dummy[Tvar[k1]]);
9272: fprintf(ficlog,"Error duplication in the model=%s at positions (+) %d and %d, V%d*V%d, Typevar=%d, Fixed=%d, Dummy=%d\n",model, k1,k2, Tvard[k3][1], Tvard[k3][2],Typevar[k1],Fixed[Tvar[k1]],Dummy[Tvar[k1]]); fflush(ficlog);
9273: return(1);
9274: }
9275: }
1.227 brouard 9276: }
9277: }
1.225 brouard 9278: }
9279: printf("ncoveff=%d, nqfveff=%d, ntveff=%d, nqtveff=%d, cptcovn=%d\n",ncoveff,nqfveff,ntveff,nqtveff,cptcovn);
9280: fprintf(ficlog,"ncoveff=%d, nqfveff=%d, ntveff=%d, nqtveff=%d, cptcovn=%d\n",ncoveff,nqfveff,ntveff,nqtveff,cptcovn);
1.234 brouard 9281: printf("ncovf=%d, ncovv=%d, ncova=%d, nsd=%d, nsq=%d\n",ncovf,ncovv,ncova,nsd,nsq);
9282: fprintf(ficlog,"ncovf=%d, ncovv=%d, ncova=%d, nsd=%d, nsq=%d\n",ncovf,ncovv,ncova,nsd, nsq);
1.137 brouard 9283: return (0); /* with covar[new additional covariate if product] and Tage if age */
1.164 brouard 9284: /*endread:*/
1.225 brouard 9285: printf("Exiting decodemodel: ");
9286: return (1);
1.136 brouard 9287: }
9288:
1.169 brouard 9289: int calandcheckages(int imx, int maxwav, double *agemin, double *agemax, int *nberr, int *nbwarn )
1.248 brouard 9290: {/* Check ages at death */
1.136 brouard 9291: int i, m;
1.218 brouard 9292: int firstone=0;
9293:
1.136 brouard 9294: for (i=1; i<=imx; i++) {
9295: for(m=2; (m<= maxwav); m++) {
9296: if (((int)mint[m][i]== 99) && (s[m][i] <= nlstate)){
9297: anint[m][i]=9999;
1.216 brouard 9298: if (s[m][i] != -2) /* Keeping initial status of unknown vital status */
9299: s[m][i]=-1;
1.136 brouard 9300: }
9301: if((int)moisdc[i]==99 && (int)andc[i]==9999 && s[m][i]>nlstate){
1.260 brouard 9302: *nberr = *nberr + 1;
1.218 brouard 9303: if(firstone == 0){
9304: firstone=1;
1.260 brouard 9305: printf("Warning (#%d)! Date of death (month %2d and year %4d) of individual %ld on line %d was unknown but status is a death state %d at wave %d. If you don't know the vital status, please enter -2. If he/she is still alive but don't know the state, please code with '-1 or '.'. Here, we do not believe in a death, skipped.\nOther similar cases in log file\n", *nberr,(int)moisdc[i],(int)andc[i],num[i],i,s[m][i],m);
1.218 brouard 9306: }
1.262 brouard 9307: fprintf(ficlog,"Warning (#%d)! Date of death (month %2d and year %4d) of individual %ld on line %d was unknown but status is a death state %d at wave %d. If you don't know the vital status, please enter -2. If he/she is still alive but don't know the state, please code with '-1 or '.'. Here, we do not believe in a death, skipped.\n", *nberr,(int)moisdc[i],(int)andc[i],num[i],i,s[m][i],m);
1.260 brouard 9308: s[m][i]=-1; /* Droping the death status */
1.136 brouard 9309: }
9310: if((int)moisdc[i]==99 && (int)andc[i]!=9999 && s[m][i]>nlstate){
1.169 brouard 9311: (*nberr)++;
1.259 brouard 9312: printf("Error (#%d)! Month of death of individual %ld on line %d was unknown (%2d) (year of death is %4d) and status is a death state %d at wave %d. Please impute an arbitrary (or not) month and rerun. Currently this transition to death will be skipped (status is set to -2).\nOther similar cases in log file\n", *nberr, num[i],i,(int)moisdc[i],(int)andc[i],s[m][i],m);
1.262 brouard 9313: fprintf(ficlog,"Error (#%d)! Month of death of individual %ld on line %d was unknown (%2d) (year of death is %4d) and status is a death state %d at wave %d. Please impute an arbitrary (or not) month and rerun. Currently this transition to death will be skipped (status is set to -2).\n", *nberr, num[i],i,(int)moisdc[i],(int)andc[i],s[m][i],m);
1.259 brouard 9314: s[m][i]=-2; /* We prefer to skip it (and to skip it in version 0.8a1 too */
1.136 brouard 9315: }
9316: }
9317: }
9318:
9319: for (i=1; i<=imx; i++) {
9320: agedc[i]=(moisdc[i]/12.+andc[i])-(moisnais[i]/12.+annais[i]);
9321: for(m=firstpass; (m<= lastpass); m++){
1.214 brouard 9322: if(s[m][i] >0 || s[m][i]==-1 || s[m][i]==-2 || s[m][i]==-4 || s[m][i]==-5){ /* What if s[m][i]=-1 */
1.136 brouard 9323: if (s[m][i] >= nlstate+1) {
1.169 brouard 9324: if(agedc[i]>0){
9325: if((int)moisdc[i]!=99 && (int)andc[i]!=9999){
1.136 brouard 9326: agev[m][i]=agedc[i];
1.214 brouard 9327: /*if(moisdc[i]==99 && andc[i]==9999) s[m][i]=-1;*/
1.169 brouard 9328: }else {
1.136 brouard 9329: if ((int)andc[i]!=9999){
9330: nbwarn++;
9331: printf("Warning negative age at death: %ld line:%d\n",num[i],i);
9332: fprintf(ficlog,"Warning negative age at death: %ld line:%d\n",num[i],i);
9333: agev[m][i]=-1;
9334: }
9335: }
1.169 brouard 9336: } /* agedc > 0 */
1.214 brouard 9337: } /* end if */
1.136 brouard 9338: else if(s[m][i] !=9){ /* Standard case, age in fractional
9339: years but with the precision of a month */
9340: agev[m][i]=(mint[m][i]/12.+1./24.+anint[m][i])-(moisnais[i]/12.+1./24.+annais[i]);
9341: if((int)mint[m][i]==99 || (int)anint[m][i]==9999)
9342: agev[m][i]=1;
9343: else if(agev[m][i] < *agemin){
9344: *agemin=agev[m][i];
9345: printf(" Min anint[%d][%d]=%.2f annais[%d]=%.2f, agemin=%.2f\n",m,i,anint[m][i], i,annais[i], *agemin);
9346: }
9347: else if(agev[m][i] >*agemax){
9348: *agemax=agev[m][i];
1.156 brouard 9349: /* printf(" Max anint[%d][%d]=%.0f annais[%d]=%.0f, agemax=%.2f\n",m,i,anint[m][i], i,annais[i], *agemax);*/
1.136 brouard 9350: }
9351: /*agev[m][i]=anint[m][i]-annais[i];*/
9352: /* agev[m][i] = age[i]+2*m;*/
1.214 brouard 9353: } /* en if 9*/
1.136 brouard 9354: else { /* =9 */
1.214 brouard 9355: /* printf("Debug num[%d]=%ld s[%d][%d]=%d\n",i,num[i], m,i, s[m][i]); */
1.136 brouard 9356: agev[m][i]=1;
9357: s[m][i]=-1;
9358: }
9359: }
1.214 brouard 9360: else if(s[m][i]==0) /*= 0 Unknown */
1.136 brouard 9361: agev[m][i]=1;
1.214 brouard 9362: else{
9363: printf("Warning, num[%d]=%ld, s[%d][%d]=%d\n", i, num[i], m, i,s[m][i]);
9364: fprintf(ficlog, "Warning, num[%d]=%ld, s[%d][%d]=%d\n", i, num[i], m, i,s[m][i]);
9365: agev[m][i]=0;
9366: }
9367: } /* End for lastpass */
9368: }
1.136 brouard 9369:
9370: for (i=1; i<=imx; i++) {
9371: for(m=firstpass; (m<=lastpass); m++){
9372: if (s[m][i] > (nlstate+ndeath)) {
1.169 brouard 9373: (*nberr)++;
1.136 brouard 9374: printf("Error: on wave %d of individual %d status %d > (nlstate+ndeath)=(%d+%d)=%d\n",m,i,s[m][i],nlstate, ndeath, nlstate+ndeath);
9375: fprintf(ficlog,"Error: on wave %d of individual %d status %d > (nlstate+ndeath)=(%d+%d)=%d\n",m,i,s[m][i],nlstate, ndeath, nlstate+ndeath);
9376: return 1;
9377: }
9378: }
9379: }
9380:
9381: /*for (i=1; i<=imx; i++){
9382: for (m=firstpass; (m<lastpass); m++){
9383: printf("%ld %d %.lf %d %d\n", num[i],(covar[1][i]),agev[m][i],s[m][i],s[m+1][i]);
9384: }
9385:
9386: }*/
9387:
9388:
1.139 brouard 9389: printf("Total number of individuals= %d, Agemin = %.2f, Agemax= %.2f\n\n", imx, *agemin, *agemax);
9390: fprintf(ficlog,"Total number of individuals= %d, Agemin = %.2f, Agemax= %.2f\n\n", imx, *agemin, *agemax);
1.136 brouard 9391:
9392: return (0);
1.164 brouard 9393: /* endread:*/
1.136 brouard 9394: printf("Exiting calandcheckages: ");
9395: return (1);
9396: }
9397:
1.172 brouard 9398: #if defined(_MSC_VER)
9399: /*printf("Visual C++ compiler: %s \n;", _MSC_FULL_VER);*/
9400: /*fprintf(ficlog, "Visual C++ compiler: %s \n;", _MSC_FULL_VER);*/
9401: //#include "stdafx.h"
9402: //#include <stdio.h>
9403: //#include <tchar.h>
9404: //#include <windows.h>
9405: //#include <iostream>
9406: typedef BOOL(WINAPI *LPFN_ISWOW64PROCESS) (HANDLE, PBOOL);
9407:
9408: LPFN_ISWOW64PROCESS fnIsWow64Process;
9409:
9410: BOOL IsWow64()
9411: {
9412: BOOL bIsWow64 = FALSE;
9413:
9414: //typedef BOOL (APIENTRY *LPFN_ISWOW64PROCESS)
9415: // (HANDLE, PBOOL);
9416:
9417: //LPFN_ISWOW64PROCESS fnIsWow64Process;
9418:
9419: HMODULE module = GetModuleHandle(_T("kernel32"));
9420: const char funcName[] = "IsWow64Process";
9421: fnIsWow64Process = (LPFN_ISWOW64PROCESS)
9422: GetProcAddress(module, funcName);
9423:
9424: if (NULL != fnIsWow64Process)
9425: {
9426: if (!fnIsWow64Process(GetCurrentProcess(),
9427: &bIsWow64))
9428: //throw std::exception("Unknown error");
9429: printf("Unknown error\n");
9430: }
9431: return bIsWow64 != FALSE;
9432: }
9433: #endif
1.177 brouard 9434:
1.191 brouard 9435: void syscompilerinfo(int logged)
1.167 brouard 9436: {
9437: /* #include "syscompilerinfo.h"*/
1.185 brouard 9438: /* command line Intel compiler 32bit windows, XP compatible:*/
9439: /* /GS /W3 /Gy
9440: /Zc:wchar_t /Zi /O2 /Fd"Release\vc120.pdb" /D "WIN32" /D "NDEBUG" /D
9441: "_CONSOLE" /D "_LIB" /D "_USING_V110_SDK71_" /D "_UNICODE" /D
9442: "UNICODE" /Qipo /Zc:forScope /Gd /Oi /MT /Fa"Release\" /EHsc /nologo
1.186 brouard 9443: /Fo"Release\" /Qprof-dir "Release\" /Fp"Release\IMaCh.pch"
9444: */
9445: /* 64 bits */
1.185 brouard 9446: /*
9447: /GS /W3 /Gy
9448: /Zc:wchar_t /Zi /O2 /Fd"x64\Release\vc120.pdb" /D "WIN32" /D "NDEBUG"
9449: /D "_CONSOLE" /D "_LIB" /D "_UNICODE" /D "UNICODE" /Qipo /Zc:forScope
9450: /Oi /MD /Fa"x64\Release\" /EHsc /nologo /Fo"x64\Release\" /Qprof-dir
9451: "x64\Release\" /Fp"x64\Release\IMaCh.pch" */
9452: /* Optimization are useless and O3 is slower than O2 */
9453: /*
9454: /GS /W3 /Gy /Zc:wchar_t /Zi /O3 /Fd"x64\Release\vc120.pdb" /D "WIN32"
9455: /D "NDEBUG" /D "_CONSOLE" /D "_LIB" /D "_UNICODE" /D "UNICODE" /Qipo
9456: /Zc:forScope /Oi /MD /Fa"x64\Release\" /EHsc /nologo /Qparallel
9457: /Fo"x64\Release\" /Qprof-dir "x64\Release\" /Fp"x64\Release\IMaCh.pch"
9458: */
1.186 brouard 9459: /* Link is */ /* /OUT:"visual studio
1.185 brouard 9460: 2013\Projects\IMaCh\Release\IMaCh.exe" /MANIFEST /NXCOMPAT
9461: /PDB:"visual studio
9462: 2013\Projects\IMaCh\Release\IMaCh.pdb" /DYNAMICBASE
9463: "kernel32.lib" "user32.lib" "gdi32.lib" "winspool.lib"
9464: "comdlg32.lib" "advapi32.lib" "shell32.lib" "ole32.lib"
9465: "oleaut32.lib" "uuid.lib" "odbc32.lib" "odbccp32.lib"
9466: /MACHINE:X86 /OPT:REF /SAFESEH /INCREMENTAL:NO
9467: /SUBSYSTEM:CONSOLE",5.01" /MANIFESTUAC:"level='asInvoker'
9468: uiAccess='false'"
9469: /ManifestFile:"Release\IMaCh.exe.intermediate.manifest" /OPT:ICF
9470: /NOLOGO /TLBID:1
9471: */
1.177 brouard 9472: #if defined __INTEL_COMPILER
1.178 brouard 9473: #if defined(__GNUC__)
9474: struct utsname sysInfo; /* For Intel on Linux and OS/X */
9475: #endif
1.177 brouard 9476: #elif defined(__GNUC__)
1.179 brouard 9477: #ifndef __APPLE__
1.174 brouard 9478: #include <gnu/libc-version.h> /* Only on gnu */
1.179 brouard 9479: #endif
1.177 brouard 9480: struct utsname sysInfo;
1.178 brouard 9481: int cross = CROSS;
9482: if (cross){
9483: printf("Cross-");
1.191 brouard 9484: if(logged) fprintf(ficlog, "Cross-");
1.178 brouard 9485: }
1.174 brouard 9486: #endif
9487:
1.171 brouard 9488: #include <stdint.h>
1.178 brouard 9489:
1.191 brouard 9490: printf("Compiled with:");if(logged)fprintf(ficlog,"Compiled with:");
1.169 brouard 9491: #if defined(__clang__)
1.191 brouard 9492: printf(" Clang/LLVM");if(logged)fprintf(ficlog," Clang/LLVM"); /* Clang/LLVM. ---------------------------------------------- */
1.169 brouard 9493: #endif
9494: #if defined(__ICC) || defined(__INTEL_COMPILER)
1.191 brouard 9495: printf(" Intel ICC/ICPC");if(logged)fprintf(ficlog," Intel ICC/ICPC");/* Intel ICC/ICPC. ------------------------------------------ */
1.169 brouard 9496: #endif
9497: #if defined(__GNUC__) || defined(__GNUG__)
1.191 brouard 9498: printf(" GNU GCC/G++");if(logged)fprintf(ficlog," GNU GCC/G++");/* GNU GCC/G++. --------------------------------------------- */
1.169 brouard 9499: #endif
9500: #if defined(__HP_cc) || defined(__HP_aCC)
1.191 brouard 9501: printf(" Hewlett-Packard C/aC++");if(logged)fprintf(fcilog," Hewlett-Packard C/aC++"); /* Hewlett-Packard C/aC++. ---------------------------------- */
1.169 brouard 9502: #endif
9503: #if defined(__IBMC__) || defined(__IBMCPP__)
1.191 brouard 9504: printf(" IBM XL C/C++"); if(logged) fprintf(ficlog," IBM XL C/C++");/* IBM XL C/C++. -------------------------------------------- */
1.169 brouard 9505: #endif
9506: #if defined(_MSC_VER)
1.191 brouard 9507: printf(" Microsoft Visual Studio");if(logged)fprintf(ficlog," Microsoft Visual Studio");/* Microsoft Visual Studio. --------------------------------- */
1.169 brouard 9508: #endif
9509: #if defined(__PGI)
1.191 brouard 9510: printf(" Portland Group PGCC/PGCPP");if(logged) fprintf(ficlog," Portland Group PGCC/PGCPP");/* Portland Group PGCC/PGCPP. ------------------------------- */
1.169 brouard 9511: #endif
9512: #if defined(__SUNPRO_C) || defined(__SUNPRO_CC)
1.191 brouard 9513: printf(" Oracle Solaris Studio");if(logged)fprintf(ficlog," Oracle Solaris Studio\n");/* Oracle Solaris Studio. ----------------------------------- */
1.167 brouard 9514: #endif
1.191 brouard 9515: printf(" for "); if (logged) fprintf(ficlog, " for ");
1.169 brouard 9516:
1.167 brouard 9517: // http://stackoverflow.com/questions/4605842/how-to-identify-platform-compiler-from-preprocessor-macros
9518: #ifdef _WIN32 // note the underscore: without it, it's not msdn official!
9519: // Windows (x64 and x86)
1.191 brouard 9520: printf("Windows (x64 and x86) ");if(logged) fprintf(ficlog,"Windows (x64 and x86) ");
1.167 brouard 9521: #elif __unix__ // all unices, not all compilers
9522: // Unix
1.191 brouard 9523: printf("Unix ");if(logged) fprintf(ficlog,"Unix ");
1.167 brouard 9524: #elif __linux__
9525: // linux
1.191 brouard 9526: printf("linux ");if(logged) fprintf(ficlog,"linux ");
1.167 brouard 9527: #elif __APPLE__
1.174 brouard 9528: // Mac OS, not sure if this is covered by __posix__ and/or __unix__ though..
1.191 brouard 9529: printf("Mac OS ");if(logged) fprintf(ficlog,"Mac OS ");
1.167 brouard 9530: #endif
9531:
9532: /* __MINGW32__ */
9533: /* __CYGWIN__ */
9534: /* __MINGW64__ */
9535: // http://msdn.microsoft.com/en-us/library/b0084kay.aspx
9536: /* _MSC_VER //the Visual C++ compiler is 17.00.51106.1, the _MSC_VER macro evaluates to 1700. Type cl /? */
9537: /* _MSC_FULL_VER //the Visual C++ compiler is 15.00.20706.01, the _MSC_FULL_VER macro evaluates to 150020706 */
9538: /* _WIN64 // Defined for applications for Win64. */
9539: /* _M_X64 // Defined for compilations that target x64 processors. */
9540: /* _DEBUG // Defined when you compile with /LDd, /MDd, and /MTd. */
1.171 brouard 9541:
1.167 brouard 9542: #if UINTPTR_MAX == 0xffffffff
1.191 brouard 9543: printf(" 32-bit"); if(logged) fprintf(ficlog," 32-bit");/* 32-bit */
1.167 brouard 9544: #elif UINTPTR_MAX == 0xffffffffffffffff
1.191 brouard 9545: printf(" 64-bit"); if(logged) fprintf(ficlog," 64-bit");/* 64-bit */
1.167 brouard 9546: #else
1.191 brouard 9547: printf(" wtf-bit"); if(logged) fprintf(ficlog," wtf-bit");/* wtf */
1.167 brouard 9548: #endif
9549:
1.169 brouard 9550: #if defined(__GNUC__)
9551: # if defined(__GNUC_PATCHLEVEL__)
9552: # define __GNUC_VERSION__ (__GNUC__ * 10000 \
9553: + __GNUC_MINOR__ * 100 \
9554: + __GNUC_PATCHLEVEL__)
9555: # else
9556: # define __GNUC_VERSION__ (__GNUC__ * 10000 \
9557: + __GNUC_MINOR__ * 100)
9558: # endif
1.174 brouard 9559: printf(" using GNU C version %d.\n", __GNUC_VERSION__);
1.191 brouard 9560: if(logged) fprintf(ficlog, " using GNU C version %d.\n", __GNUC_VERSION__);
1.176 brouard 9561:
9562: if (uname(&sysInfo) != -1) {
9563: printf("Running on: %s %s %s %s %s\n",sysInfo.sysname, sysInfo.nodename, sysInfo.release, sysInfo.version, sysInfo.machine);
1.191 brouard 9564: if(logged) fprintf(ficlog,"Running on: %s %s %s %s %s\n ",sysInfo.sysname, sysInfo.nodename, sysInfo.release, sysInfo.version, sysInfo.machine);
1.176 brouard 9565: }
9566: else
9567: perror("uname() error");
1.179 brouard 9568: //#ifndef __INTEL_COMPILER
9569: #if !defined (__INTEL_COMPILER) && !defined(__APPLE__)
1.174 brouard 9570: printf("GNU libc version: %s\n", gnu_get_libc_version());
1.191 brouard 9571: if(logged) fprintf(ficlog,"GNU libc version: %s\n", gnu_get_libc_version());
1.177 brouard 9572: #endif
1.169 brouard 9573: #endif
1.172 brouard 9574:
9575: // void main()
9576: // {
1.169 brouard 9577: #if defined(_MSC_VER)
1.174 brouard 9578: if (IsWow64()){
1.191 brouard 9579: printf("\nThe program (probably compiled for 32bit) is running under WOW64 (64bit) emulation.\n");
9580: if (logged) fprintf(ficlog, "\nThe program (probably compiled for 32bit) is running under WOW64 (64bit) emulation.\n");
1.174 brouard 9581: }
9582: else{
1.191 brouard 9583: printf("\nThe program is not running under WOW64 (i.e probably on a 64bit Windows).\n");
9584: if (logged) fprintf(ficlog, "\nThe programm is not running under WOW64 (i.e probably on a 64bit Windows).\n");
1.174 brouard 9585: }
1.172 brouard 9586: // printf("\nPress Enter to continue...");
9587: // getchar();
9588: // }
9589:
1.169 brouard 9590: #endif
9591:
1.167 brouard 9592:
1.219 brouard 9593: }
1.136 brouard 9594:
1.219 brouard 9595: int prevalence_limit(double *p, double **prlim, double ageminpar, double agemaxpar, double ftolpl, int *ncvyearp){
1.180 brouard 9596: /*--------------- Prevalence limit (period or stable prevalence) --------------*/
1.235 brouard 9597: int i, j, k, i1, k4=0, nres=0 ;
1.202 brouard 9598: /* double ftolpl = 1.e-10; */
1.180 brouard 9599: double age, agebase, agelim;
1.203 brouard 9600: double tot;
1.180 brouard 9601:
1.202 brouard 9602: strcpy(filerespl,"PL_");
9603: strcat(filerespl,fileresu);
9604: if((ficrespl=fopen(filerespl,"w"))==NULL) {
9605: printf("Problem with period (stable) prevalence resultfile: %s\n", filerespl);return 1;
9606: fprintf(ficlog,"Problem with period (stable) prevalence resultfile: %s\n", filerespl);return 1;
9607: }
1.227 brouard 9608: printf("\nComputing period (stable) prevalence: result on file '%s' \n", filerespl);
9609: fprintf(ficlog,"\nComputing period (stable) prevalence: result on file '%s' \n", filerespl);
1.202 brouard 9610: pstamp(ficrespl);
1.203 brouard 9611: fprintf(ficrespl,"# Period (stable) prevalence. Precision given by ftolpl=%g \n", ftolpl);
1.202 brouard 9612: fprintf(ficrespl,"#Age ");
9613: for(i=1; i<=nlstate;i++) fprintf(ficrespl,"%d-%d ",i,i);
9614: fprintf(ficrespl,"\n");
1.180 brouard 9615:
1.219 brouard 9616: /* prlim=matrix(1,nlstate,1,nlstate);*/ /* back in main */
1.180 brouard 9617:
1.219 brouard 9618: agebase=ageminpar;
9619: agelim=agemaxpar;
1.180 brouard 9620:
1.227 brouard 9621: /* i1=pow(2,ncoveff); */
1.234 brouard 9622: i1=pow(2,cptcoveff); /* Number of combination of dummy covariates */
1.219 brouard 9623: if (cptcovn < 1){i1=1;}
1.180 brouard 9624:
1.238 brouard 9625: for(k=1; k<=i1;k++){ /* For each combination k of dummy covariates in the model */
9626: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
1.253 brouard 9627: if(i1 != 1 && TKresult[nres]!= k)
1.238 brouard 9628: continue;
1.235 brouard 9629:
1.238 brouard 9630: /* for(cptcov=1,k=0;cptcov<=i1;cptcov++){ */
9631: /* for(cptcov=1,k=0;cptcov<=1;cptcov++){ */
9632: //for(cptcod=1;cptcod<=ncodemax[cptcov];cptcod++){
9633: /* k=k+1; */
9634: /* to clean */
9635: //printf("cptcov=%d cptcod=%d codtab=%d\n",cptcov, cptcod,codtabm(cptcod,cptcov));
9636: fprintf(ficrespl,"#******");
9637: printf("#******");
9638: fprintf(ficlog,"#******");
9639: for(j=1;j<=cptcoveff ;j++) {/* all covariates */
9640: fprintf(ficrespl," V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]); /* Here problem for varying dummy*/
9641: printf(" V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9642: fprintf(ficlog," V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9643: }
9644: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
9645: printf(" V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
9646: fprintf(ficrespl," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
9647: fprintf(ficlog," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
9648: }
9649: fprintf(ficrespl,"******\n");
9650: printf("******\n");
9651: fprintf(ficlog,"******\n");
9652: if(invalidvarcomb[k]){
9653: printf("\nCombination (%d) ignored because no case \n",k);
9654: fprintf(ficrespl,"#Combination (%d) ignored because no case \n",k);
9655: fprintf(ficlog,"\nCombination (%d) ignored because no case \n",k);
9656: continue;
9657: }
1.219 brouard 9658:
1.238 brouard 9659: fprintf(ficrespl,"#Age ");
9660: for(j=1;j<=cptcoveff;j++) {
9661: fprintf(ficrespl,"V%d %d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9662: }
9663: for(i=1; i<=nlstate;i++) fprintf(ficrespl," %d-%d ",i,i);
9664: fprintf(ficrespl,"Total Years_to_converge\n");
1.227 brouard 9665:
1.238 brouard 9666: for (age=agebase; age<=agelim; age++){
9667: /* for (age=agebase; age<=agebase; age++){ */
9668: prevalim(prlim, nlstate, p, age, oldm, savm, ftolpl, ncvyearp, k, nres);
9669: fprintf(ficrespl,"%.0f ",age );
9670: for(j=1;j<=cptcoveff;j++)
9671: fprintf(ficrespl,"%d %d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9672: tot=0.;
9673: for(i=1; i<=nlstate;i++){
9674: tot += prlim[i][i];
9675: fprintf(ficrespl," %.5f", prlim[i][i]);
9676: }
9677: fprintf(ficrespl," %.3f %d\n", tot, *ncvyearp);
9678: } /* Age */
9679: /* was end of cptcod */
9680: } /* cptcov */
9681: } /* nres */
1.219 brouard 9682: return 0;
1.180 brouard 9683: }
9684:
1.218 brouard 9685: int back_prevalence_limit(double *p, double **bprlim, double ageminpar, double agemaxpar, double ftolpl, int *ncvyearp, double dateprev1,double dateprev2, int firstpass, int lastpass, int mobilavproj){
9686: /*--------------- Back Prevalence limit (period or stable prevalence) --------------*/
9687:
9688: /* Computes the back prevalence limit for any combination of covariate values
9689: * at any age between ageminpar and agemaxpar
9690: */
1.235 brouard 9691: int i, j, k, i1, nres=0 ;
1.217 brouard 9692: /* double ftolpl = 1.e-10; */
9693: double age, agebase, agelim;
9694: double tot;
1.218 brouard 9695: /* double ***mobaverage; */
9696: /* double **dnewm, **doldm, **dsavm; /\* for use *\/ */
1.217 brouard 9697:
9698: strcpy(fileresplb,"PLB_");
9699: strcat(fileresplb,fileresu);
9700: if((ficresplb=fopen(fileresplb,"w"))==NULL) {
9701: printf("Problem with period (stable) back prevalence resultfile: %s\n", fileresplb);return 1;
9702: fprintf(ficlog,"Problem with period (stable) back prevalence resultfile: %s\n", fileresplb);return 1;
9703: }
9704: printf("Computing period (stable) back prevalence: result on file '%s' \n", fileresplb);
9705: fprintf(ficlog,"Computing period (stable) back prevalence: result on file '%s' \n", fileresplb);
9706: pstamp(ficresplb);
9707: fprintf(ficresplb,"# Period (stable) back prevalence. Precision given by ftolpl=%g \n", ftolpl);
9708: fprintf(ficresplb,"#Age ");
9709: for(i=1; i<=nlstate;i++) fprintf(ficresplb,"%d-%d ",i,i);
9710: fprintf(ficresplb,"\n");
9711:
1.218 brouard 9712:
9713: /* prlim=matrix(1,nlstate,1,nlstate);*/ /* back in main */
9714:
9715: agebase=ageminpar;
9716: agelim=agemaxpar;
9717:
9718:
1.227 brouard 9719: i1=pow(2,cptcoveff);
1.218 brouard 9720: if (cptcovn < 1){i1=1;}
1.227 brouard 9721:
1.238 brouard 9722: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
9723: for(k=1; k<=i1;k++){ /* For any combination of dummy covariates, fixed and varying */
1.253 brouard 9724: if(i1 != 1 && TKresult[nres]!= k)
1.238 brouard 9725: continue;
9726: //printf("cptcov=%d cptcod=%d codtab=%d\n",cptcov, cptcod,codtabm(cptcod,cptcov));
9727: fprintf(ficresplb,"#******");
9728: printf("#******");
9729: fprintf(ficlog,"#******");
9730: for(j=1;j<=cptcoveff ;j++) {/* all covariates */
9731: fprintf(ficresplb," V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9732: printf(" V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9733: fprintf(ficlog," V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9734: }
9735: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
9736: printf(" V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
9737: fprintf(ficresplb," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
9738: fprintf(ficlog," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
9739: }
9740: fprintf(ficresplb,"******\n");
9741: printf("******\n");
9742: fprintf(ficlog,"******\n");
9743: if(invalidvarcomb[k]){
9744: printf("\nCombination (%d) ignored because no cases \n",k);
9745: fprintf(ficresplb,"#Combination (%d) ignored because no cases \n",k);
9746: fprintf(ficlog,"\nCombination (%d) ignored because no cases \n",k);
9747: continue;
9748: }
1.218 brouard 9749:
1.238 brouard 9750: fprintf(ficresplb,"#Age ");
9751: for(j=1;j<=cptcoveff;j++) {
9752: fprintf(ficresplb,"V%d %d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9753: }
9754: for(i=1; i<=nlstate;i++) fprintf(ficresplb," %d-%d ",i,i);
9755: fprintf(ficresplb,"Total Years_to_converge\n");
1.218 brouard 9756:
9757:
1.238 brouard 9758: for (age=agebase; age<=agelim; age++){
9759: /* for (age=agebase; age<=agebase; age++){ */
9760: if(mobilavproj > 0){
9761: /* bprevalim(bprlim, mobaverage, nlstate, p, age, ageminpar, agemaxpar, oldm, savm, doldm, dsavm, ftolpl, ncvyearp, k); */
9762: /* bprevalim(bprlim, mobaverage, nlstate, p, age, oldm, savm, dnewm, doldm, dsavm, ftolpl, ncvyearp, k); */
1.242 brouard 9763: bprevalim(bprlim, mobaverage, nlstate, p, age, ftolpl, ncvyearp, k, nres);
1.238 brouard 9764: }else if (mobilavproj == 0){
9765: printf("There is no chance to get back prevalence limit if data aren't non zero and summing to 1, please try a non null mobil_average(=%d) parameter or mobil_average=-1 if you want to try at your own risk.\n",mobilavproj);
9766: fprintf(ficlog,"There is no chance to get back prevalence limit if data aren't non zero and summing to 1, please try a non null mobil_average(=%d) parameter or mobil_average=-1 if you want to try at your own risk.\n",mobilavproj);
9767: exit(1);
9768: }else{
9769: /* bprevalim(bprlim, probs, nlstate, p, age, oldm, savm, dnewm, doldm, dsavm, ftolpl, ncvyearp, k); */
1.242 brouard 9770: bprevalim(bprlim, probs, nlstate, p, age, ftolpl, ncvyearp, k,nres);
1.238 brouard 9771: }
9772: fprintf(ficresplb,"%.0f ",age );
9773: for(j=1;j<=cptcoveff;j++)
9774: fprintf(ficresplb,"%d %d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9775: tot=0.;
9776: for(i=1; i<=nlstate;i++){
9777: tot += bprlim[i][i];
9778: fprintf(ficresplb," %.5f", bprlim[i][i]);
9779: }
9780: fprintf(ficresplb," %.3f %d\n", tot, *ncvyearp);
9781: } /* Age */
9782: /* was end of cptcod */
1.255 brouard 9783: /*fprintf(ficresplb,"\n");*/ /* Seems to be necessary for gnuplot only if two result lines and no covariate. */
1.238 brouard 9784: } /* end of any combination */
9785: } /* end of nres */
1.218 brouard 9786: /* hBijx(p, bage, fage); */
9787: /* fclose(ficrespijb); */
9788:
9789: return 0;
1.217 brouard 9790: }
1.218 brouard 9791:
1.180 brouard 9792: int hPijx(double *p, int bage, int fage){
9793: /*------------- h Pij x at various ages ------------*/
9794:
9795: int stepsize;
9796: int agelim;
9797: int hstepm;
9798: int nhstepm;
1.235 brouard 9799: int h, i, i1, j, k, k4, nres=0;
1.180 brouard 9800:
9801: double agedeb;
9802: double ***p3mat;
9803:
1.201 brouard 9804: strcpy(filerespij,"PIJ_"); strcat(filerespij,fileresu);
1.180 brouard 9805: if((ficrespij=fopen(filerespij,"w"))==NULL) {
9806: printf("Problem with Pij resultfile: %s\n", filerespij); return 1;
9807: fprintf(ficlog,"Problem with Pij resultfile: %s\n", filerespij); return 1;
9808: }
9809: printf("Computing pij: result on file '%s' \n", filerespij);
9810: fprintf(ficlog,"Computing pij: result on file '%s' \n", filerespij);
9811:
9812: stepsize=(int) (stepm+YEARM-1)/YEARM;
9813: /*if (stepm<=24) stepsize=2;*/
9814:
9815: agelim=AGESUP;
9816: hstepm=stepsize*YEARM; /* Every year of age */
9817: hstepm=hstepm/stepm; /* Typically 2 years, = 2/6 months = 4 */
1.218 brouard 9818:
1.180 brouard 9819: /* hstepm=1; aff par mois*/
9820: pstamp(ficrespij);
9821: fprintf(ficrespij,"#****** h Pij x Probability to be in state j at age x+h being in i at x ");
1.227 brouard 9822: i1= pow(2,cptcoveff);
1.218 brouard 9823: /* for(cptcov=1,k=0;cptcov<=i1;cptcov++){ */
9824: /* /\*for(cptcod=1;cptcod<=ncodemax[cptcov];cptcod++){*\/ */
9825: /* k=k+1; */
1.235 brouard 9826: for(nres=1; nres <= nresult; nres++) /* For each resultline */
9827: for(k=1; k<=i1;k++){
1.253 brouard 9828: if(i1 != 1 && TKresult[nres]!= k)
1.235 brouard 9829: continue;
1.183 brouard 9830: fprintf(ficrespij,"\n#****** ");
1.227 brouard 9831: for(j=1;j<=cptcoveff;j++)
1.198 brouard 9832: fprintf(ficrespij,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
1.235 brouard 9833: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
9834: printf(" V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
9835: fprintf(ficrespij," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
9836: }
1.183 brouard 9837: fprintf(ficrespij,"******\n");
9838:
9839: for (agedeb=fage; agedeb>=bage; agedeb--){ /* If stepm=6 months */
9840: nhstepm=(int) rint((agelim-agedeb)*YEARM/stepm); /* Typically 20 years = 20*12/6=40 */
9841: nhstepm = nhstepm/hstepm; /* Typically 40/4=10 */
9842:
9843: /* nhstepm=nhstepm*YEARM; aff par mois*/
1.180 brouard 9844:
1.183 brouard 9845: p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
9846: oldm=oldms;savm=savms;
1.235 brouard 9847: hpxij(p3mat,nhstepm,agedeb,hstepm,p,nlstate,stepm,oldm,savm, k, nres);
1.183 brouard 9848: fprintf(ficrespij,"# Cov Agex agex+h hpijx with i,j=");
9849: for(i=1; i<=nlstate;i++)
9850: for(j=1; j<=nlstate+ndeath;j++)
9851: fprintf(ficrespij," %1d-%1d",i,j);
9852: fprintf(ficrespij,"\n");
9853: for (h=0; h<=nhstepm; h++){
9854: /*agedebphstep = agedeb + h*hstepm/YEARM*stepm;*/
9855: fprintf(ficrespij,"%d %3.f %3.f",k, agedeb, agedeb + h*hstepm/YEARM*stepm );
1.180 brouard 9856: for(i=1; i<=nlstate;i++)
9857: for(j=1; j<=nlstate+ndeath;j++)
1.183 brouard 9858: fprintf(ficrespij," %.5f", p3mat[i][j][h]);
1.180 brouard 9859: fprintf(ficrespij,"\n");
9860: }
1.183 brouard 9861: free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
9862: fprintf(ficrespij,"\n");
9863: }
1.180 brouard 9864: /*}*/
9865: }
1.218 brouard 9866: return 0;
1.180 brouard 9867: }
1.218 brouard 9868:
9869: int hBijx(double *p, int bage, int fage, double ***prevacurrent){
1.217 brouard 9870: /*------------- h Bij x at various ages ------------*/
9871:
9872: int stepsize;
1.218 brouard 9873: /* int agelim; */
9874: int ageminl;
1.217 brouard 9875: int hstepm;
9876: int nhstepm;
1.238 brouard 9877: int h, i, i1, j, k, nres;
1.218 brouard 9878:
1.217 brouard 9879: double agedeb;
9880: double ***p3mat;
1.218 brouard 9881:
9882: strcpy(filerespijb,"PIJB_"); strcat(filerespijb,fileresu);
9883: if((ficrespijb=fopen(filerespijb,"w"))==NULL) {
9884: printf("Problem with Pij back resultfile: %s\n", filerespijb); return 1;
9885: fprintf(ficlog,"Problem with Pij back resultfile: %s\n", filerespijb); return 1;
9886: }
9887: printf("Computing pij back: result on file '%s' \n", filerespijb);
9888: fprintf(ficlog,"Computing pij back: result on file '%s' \n", filerespijb);
9889:
9890: stepsize=(int) (stepm+YEARM-1)/YEARM;
9891: /*if (stepm<=24) stepsize=2;*/
1.217 brouard 9892:
1.218 brouard 9893: /* agelim=AGESUP; */
9894: ageminl=30;
9895: hstepm=stepsize*YEARM; /* Every year of age */
9896: hstepm=hstepm/stepm; /* Typically 2 years, = 2/6 months = 4 */
9897:
9898: /* hstepm=1; aff par mois*/
9899: pstamp(ficrespijb);
1.255 brouard 9900: fprintf(ficrespijb,"#****** h Bij x Back probability to be in state i at age x-h being in j at x: B1j+B2j+...=1 ");
1.227 brouard 9901: i1= pow(2,cptcoveff);
1.218 brouard 9902: /* for(cptcov=1,k=0;cptcov<=i1;cptcov++){ */
9903: /* /\*for(cptcod=1;cptcod<=ncodemax[cptcov];cptcod++){*\/ */
9904: /* k=k+1; */
1.238 brouard 9905: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
9906: for(k=1; k<=i1;k++){ /* For any combination of dummy covariates, fixed and varying */
1.253 brouard 9907: if(i1 != 1 && TKresult[nres]!= k)
1.238 brouard 9908: continue;
9909: fprintf(ficrespijb,"\n#****** ");
9910: for(j=1;j<=cptcoveff;j++)
9911: fprintf(ficrespijb,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9912: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
9913: fprintf(ficrespijb," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
9914: }
9915: fprintf(ficrespijb,"******\n");
1.264 brouard 9916: if(invalidvarcomb[k]){ /* Is it necessary here? */
1.238 brouard 9917: fprintf(ficrespijb,"\n#Combination (%d) ignored because no cases \n",k);
9918: continue;
9919: }
9920:
9921: /* for (agedeb=fage; agedeb>=bage; agedeb--){ /\* If stepm=6 months *\/ */
9922: for (agedeb=bage; agedeb<=fage; agedeb++){ /* If stepm=6 months and estepm=24 (2 years) */
9923: /* nhstepm=(int) rint((agelim-agedeb)*YEARM/stepm); /\* Typically 20 years = 20*12/6=40 *\/ */
9924: nhstepm=(int) rint((agedeb-ageminl)*YEARM/stepm); /* Typically 20 years = 20*12/6=40 */
9925: nhstepm = nhstepm/hstepm; /* Typically 40/4=10, because estepm=24 stepm=6 => hstepm=24/6=4 */
9926:
9927: /* nhstepm=nhstepm*YEARM; aff par mois*/
9928:
9929: p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
9930: /* oldm=oldms;savm=savms; */
9931: /* hbxij(p3mat,nhstepm,agedeb,hstepm,p,nlstate,stepm,oldm,savm, k); */
9932: hbxij(p3mat,nhstepm,agedeb,hstepm,p,prevacurrent,nlstate,stepm, k);
9933: /* hbxij(p3mat,nhstepm,agedeb,hstepm,p,prevacurrent,nlstate,stepm,oldm,savm, dnewm, doldm, dsavm, k); */
1.255 brouard 9934: fprintf(ficrespijb,"# Cov Agex agex-h hbijx with i,j=");
1.217 brouard 9935: for(i=1; i<=nlstate;i++)
9936: for(j=1; j<=nlstate+ndeath;j++)
1.238 brouard 9937: fprintf(ficrespijb," %1d-%1d",i,j);
1.217 brouard 9938: fprintf(ficrespijb,"\n");
1.238 brouard 9939: for (h=0; h<=nhstepm; h++){
9940: /*agedebphstep = agedeb + h*hstepm/YEARM*stepm;*/
9941: fprintf(ficrespijb,"%d %3.f %3.f",k, agedeb, agedeb - h*hstepm/YEARM*stepm );
9942: /* fprintf(ficrespijb,"%d %3.f %3.f",k, agedeb, agedeb + h*hstepm/YEARM*stepm ); */
9943: for(i=1; i<=nlstate;i++)
9944: for(j=1; j<=nlstate+ndeath;j++)
9945: fprintf(ficrespijb," %.5f", p3mat[i][j][h]);
9946: fprintf(ficrespijb,"\n");
9947: }
9948: free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
9949: fprintf(ficrespijb,"\n");
9950: } /* end age deb */
9951: } /* end combination */
9952: } /* end nres */
1.218 brouard 9953: return 0;
9954: } /* hBijx */
1.217 brouard 9955:
1.180 brouard 9956:
1.136 brouard 9957: /***********************************************/
9958: /**************** Main Program *****************/
9959: /***********************************************/
9960:
9961: int main(int argc, char *argv[])
9962: {
9963: #ifdef GSL
9964: const gsl_multimin_fminimizer_type *T;
9965: size_t iteri = 0, it;
9966: int rval = GSL_CONTINUE;
9967: int status = GSL_SUCCESS;
9968: double ssval;
9969: #endif
9970: int movingaverage(double ***probs, double bage,double fage, double ***mobaverage, int mobilav);
1.164 brouard 9971: int i,j, k, n=MAXN,iter=0,m,size=100, cptcod;
1.209 brouard 9972: int ncvyear=0; /* Number of years needed for the period prevalence to converge */
1.164 brouard 9973: int jj, ll, li, lj, lk;
1.136 brouard 9974: int numlinepar=0; /* Current linenumber of parameter file */
1.197 brouard 9975: int num_filled;
1.136 brouard 9976: int itimes;
9977: int NDIM=2;
9978: int vpopbased=0;
1.235 brouard 9979: int nres=0;
1.258 brouard 9980: int endishere=0;
1.136 brouard 9981:
1.164 brouard 9982: char ca[32], cb[32];
1.136 brouard 9983: /* FILE *fichtm; *//* Html File */
9984: /* FILE *ficgp;*/ /*Gnuplot File */
9985: struct stat info;
1.191 brouard 9986: double agedeb=0.;
1.194 brouard 9987:
9988: double ageminpar=AGEOVERFLOW,agemin=AGEOVERFLOW, agemaxpar=-AGEOVERFLOW, agemax=-AGEOVERFLOW;
1.219 brouard 9989: double ageminout=-AGEOVERFLOW,agemaxout=AGEOVERFLOW; /* Smaller Age range redefined after movingaverage */
1.136 brouard 9990:
1.165 brouard 9991: double fret;
1.191 brouard 9992: double dum=0.; /* Dummy variable */
1.136 brouard 9993: double ***p3mat;
1.218 brouard 9994: /* double ***mobaverage; */
1.164 brouard 9995:
9996: char line[MAXLINE];
1.197 brouard 9997: char path[MAXLINE],pathc[MAXLINE],pathcd[MAXLINE],pathtot[MAXLINE];
9998:
1.234 brouard 9999: char modeltemp[MAXLINE];
1.230 brouard 10000: char resultline[MAXLINE];
10001:
1.136 brouard 10002: char pathr[MAXLINE], pathimach[MAXLINE];
1.164 brouard 10003: char *tok, *val; /* pathtot */
1.136 brouard 10004: int firstobs=1, lastobs=10;
1.195 brouard 10005: int c, h , cpt, c2;
1.191 brouard 10006: int jl=0;
10007: int i1, j1, jk, stepsize=0;
1.194 brouard 10008: int count=0;
10009:
1.164 brouard 10010: int *tab;
1.136 brouard 10011: int mobilavproj=0 , prevfcast=0 ; /* moving average of prev, If prevfcast=1 prevalence projection */
1.217 brouard 10012: int backcast=0;
1.136 brouard 10013: int mobilav=0,popforecast=0;
1.191 brouard 10014: int hstepm=0, nhstepm=0;
1.136 brouard 10015: int agemortsup;
10016: float sumlpop=0.;
10017: double jprev1=1, mprev1=1,anprev1=2000,jprev2=1, mprev2=1,anprev2=2000;
10018: double jpyram=1, mpyram=1,anpyram=2000,jpyram1=1, mpyram1=1,anpyram1=2000;
10019:
1.191 brouard 10020: double bage=0, fage=110., age, agelim=0., agebase=0.;
1.136 brouard 10021: double ftolpl=FTOL;
10022: double **prlim;
1.217 brouard 10023: double **bprlim;
1.136 brouard 10024: double ***param; /* Matrix of parameters */
1.251 brouard 10025: double ***paramstart; /* Matrix of starting parameter values */
10026: double *p, *pstart; /* p=param[1][1] pstart is for starting values guessed by freqsummary */
1.136 brouard 10027: double **matcov; /* Matrix of covariance */
1.203 brouard 10028: double **hess; /* Hessian matrix */
1.136 brouard 10029: double ***delti3; /* Scale */
10030: double *delti; /* Scale */
10031: double ***eij, ***vareij;
10032: double **varpl; /* Variances of prevalence limits by age */
10033: double *epj, vepp;
1.164 brouard 10034:
1.136 brouard 10035: double dateprev1, dateprev2,jproj1=1,mproj1=1,anproj1=2000,jproj2=1,mproj2=1,anproj2=2000;
1.217 brouard 10036: double jback1=1,mback1=1,anback1=2000,jback2=1,mback2=1,anback2=2000;
10037:
1.136 brouard 10038: double **ximort;
1.145 brouard 10039: char *alph[]={"a","a","b","c","d","e"}, str[4]="1234";
1.136 brouard 10040: int *dcwave;
10041:
1.164 brouard 10042: char z[1]="c";
1.136 brouard 10043:
10044: /*char *strt;*/
10045: char strtend[80];
1.126 brouard 10046:
1.164 brouard 10047:
1.126 brouard 10048: /* setlocale (LC_ALL, ""); */
10049: /* bindtextdomain (PACKAGE, LOCALEDIR); */
10050: /* textdomain (PACKAGE); */
10051: /* setlocale (LC_CTYPE, ""); */
10052: /* setlocale (LC_MESSAGES, ""); */
10053:
10054: /* gettimeofday(&start_time, (struct timezone*)0); */ /* at first time */
1.157 brouard 10055: rstart_time = time(NULL);
10056: /* (void) gettimeofday(&start_time,&tzp);*/
10057: start_time = *localtime(&rstart_time);
1.126 brouard 10058: curr_time=start_time;
1.157 brouard 10059: /*tml = *localtime(&start_time.tm_sec);*/
10060: /* strcpy(strstart,asctime(&tml)); */
10061: strcpy(strstart,asctime(&start_time));
1.126 brouard 10062:
10063: /* printf("Localtime (at start)=%s",strstart); */
1.157 brouard 10064: /* tp.tm_sec = tp.tm_sec +86400; */
10065: /* tm = *localtime(&start_time.tm_sec); */
1.126 brouard 10066: /* tmg.tm_year=tmg.tm_year +dsign*dyear; */
10067: /* tmg.tm_mon=tmg.tm_mon +dsign*dmonth; */
10068: /* tmg.tm_hour=tmg.tm_hour + 1; */
1.157 brouard 10069: /* tp.tm_sec = mktime(&tmg); */
1.126 brouard 10070: /* strt=asctime(&tmg); */
10071: /* printf("Time(after) =%s",strstart); */
10072: /* (void) time (&time_value);
10073: * printf("time=%d,t-=%d\n",time_value,time_value-86400);
10074: * tm = *localtime(&time_value);
10075: * strstart=asctime(&tm);
10076: * printf("tim_value=%d,asctime=%s\n",time_value,strstart);
10077: */
10078:
10079: nberr=0; /* Number of errors and warnings */
10080: nbwarn=0;
1.184 brouard 10081: #ifdef WIN32
10082: _getcwd(pathcd, size);
10083: #else
1.126 brouard 10084: getcwd(pathcd, size);
1.184 brouard 10085: #endif
1.191 brouard 10086: syscompilerinfo(0);
1.196 brouard 10087: printf("\nIMaCh version %s, %s\n%s",version, copyright, fullversion);
1.126 brouard 10088: if(argc <=1){
10089: printf("\nEnter the parameter file name: ");
1.205 brouard 10090: if(!fgets(pathr,FILENAMELENGTH,stdin)){
10091: printf("ERROR Empty parameter file name\n");
10092: goto end;
10093: }
1.126 brouard 10094: i=strlen(pathr);
10095: if(pathr[i-1]=='\n')
10096: pathr[i-1]='\0';
1.156 brouard 10097: i=strlen(pathr);
1.205 brouard 10098: if(i >= 1 && pathr[i-1]==' ') {/* This may happen when dragging on oS/X! */
1.156 brouard 10099: pathr[i-1]='\0';
1.205 brouard 10100: }
10101: i=strlen(pathr);
10102: if( i==0 ){
10103: printf("ERROR Empty parameter file name\n");
10104: goto end;
10105: }
10106: for (tok = pathr; tok != NULL; ){
1.126 brouard 10107: printf("Pathr |%s|\n",pathr);
10108: while ((val = strsep(&tok, "\"" )) != NULL && *val == '\0');
10109: printf("val= |%s| pathr=%s\n",val,pathr);
10110: strcpy (pathtot, val);
10111: if(pathr[0] == '\0') break; /* Dirty */
10112: }
10113: }
10114: else{
10115: strcpy(pathtot,argv[1]);
10116: }
10117: /*if(getcwd(pathcd, MAXLINE)!= NULL)printf ("Error pathcd\n");*/
10118: /*cygwin_split_path(pathtot,path,optionfile);
10119: printf("pathtot=%s, path=%s, optionfile=%s\n",pathtot,path,optionfile);*/
10120: /* cutv(path,optionfile,pathtot,'\\');*/
10121:
10122: /* Split argv[0], imach program to get pathimach */
10123: printf("\nargv[0]=%s argv[1]=%s, \n",argv[0],argv[1]);
10124: split(argv[0],pathimach,optionfile,optionfilext,optionfilefiname);
10125: printf("\nargv[0]=%s pathimach=%s, \noptionfile=%s \noptionfilext=%s \noptionfilefiname=%s\n",argv[0],pathimach,optionfile,optionfilext,optionfilefiname);
10126: /* strcpy(pathimach,argv[0]); */
10127: /* Split argv[1]=pathtot, parameter file name to get path, optionfile, extension and name */
10128: split(pathtot,path,optionfile,optionfilext,optionfilefiname);
10129: printf("\npathtot=%s,\npath=%s,\noptionfile=%s \noptionfilext=%s \noptionfilefiname=%s\n",pathtot,path,optionfile,optionfilext,optionfilefiname);
1.184 brouard 10130: #ifdef WIN32
10131: _chdir(path); /* Can be a relative path */
10132: if(_getcwd(pathcd,MAXLINE) > 0) /* So pathcd is the full path */
10133: #else
1.126 brouard 10134: chdir(path); /* Can be a relative path */
1.184 brouard 10135: if (getcwd(pathcd, MAXLINE) > 0) /* So pathcd is the full path */
10136: #endif
10137: printf("Current directory %s!\n",pathcd);
1.126 brouard 10138: strcpy(command,"mkdir ");
10139: strcat(command,optionfilefiname);
10140: if((outcmd=system(command)) != 0){
1.169 brouard 10141: printf("Directory already exists (or can't create it) %s%s, err=%d\n",path,optionfilefiname,outcmd);
1.126 brouard 10142: /* fprintf(ficlog,"Problem creating directory %s%s\n",path,optionfilefiname); */
10143: /* fclose(ficlog); */
10144: /* exit(1); */
10145: }
10146: /* if((imk=mkdir(optionfilefiname))<0){ */
10147: /* perror("mkdir"); */
10148: /* } */
10149:
10150: /*-------- arguments in the command line --------*/
10151:
1.186 brouard 10152: /* Main Log file */
1.126 brouard 10153: strcat(filelog, optionfilefiname);
10154: strcat(filelog,".log"); /* */
10155: if((ficlog=fopen(filelog,"w"))==NULL) {
10156: printf("Problem with logfile %s\n",filelog);
10157: goto end;
10158: }
10159: fprintf(ficlog,"Log filename:%s\n",filelog);
1.197 brouard 10160: fprintf(ficlog,"Version %s %s",version,fullversion);
1.126 brouard 10161: fprintf(ficlog,"\nEnter the parameter file name: \n");
10162: fprintf(ficlog,"pathimach=%s\npathtot=%s\n\
10163: path=%s \n\
10164: optionfile=%s\n\
10165: optionfilext=%s\n\
1.156 brouard 10166: optionfilefiname='%s'\n",pathimach,pathtot,path,optionfile,optionfilext,optionfilefiname);
1.126 brouard 10167:
1.197 brouard 10168: syscompilerinfo(1);
1.167 brouard 10169:
1.126 brouard 10170: printf("Local time (at start):%s",strstart);
10171: fprintf(ficlog,"Local time (at start): %s",strstart);
10172: fflush(ficlog);
10173: /* (void) gettimeofday(&curr_time,&tzp); */
1.157 brouard 10174: /* printf("Elapsed time %d\n", asc_diff_time(curr_time.tm_sec-start_time.tm_sec,tmpout)); */
1.126 brouard 10175:
10176: /* */
10177: strcpy(fileres,"r");
10178: strcat(fileres, optionfilefiname);
1.201 brouard 10179: strcat(fileresu, optionfilefiname); /* Without r in front */
1.126 brouard 10180: strcat(fileres,".txt"); /* Other files have txt extension */
1.201 brouard 10181: strcat(fileresu,".txt"); /* Other files have txt extension */
1.126 brouard 10182:
1.186 brouard 10183: /* Main ---------arguments file --------*/
1.126 brouard 10184:
10185: if((ficpar=fopen(optionfile,"r"))==NULL) {
1.155 brouard 10186: printf("Problem with optionfile '%s' with errno='%s'\n",optionfile,strerror(errno));
10187: fprintf(ficlog,"Problem with optionfile '%s' with errno='%s'\n",optionfile,strerror(errno));
1.126 brouard 10188: fflush(ficlog);
1.149 brouard 10189: /* goto end; */
10190: exit(70);
1.126 brouard 10191: }
10192:
10193:
10194:
10195: strcpy(filereso,"o");
1.201 brouard 10196: strcat(filereso,fileresu);
1.126 brouard 10197: if((ficparo=fopen(filereso,"w"))==NULL) { /* opened on subdirectory */
10198: printf("Problem with Output resultfile: %s\n", filereso);
10199: fprintf(ficlog,"Problem with Output resultfile: %s\n", filereso);
10200: fflush(ficlog);
10201: goto end;
10202: }
10203:
10204: /* Reads comments: lines beginning with '#' */
10205: numlinepar=0;
1.197 brouard 10206:
10207: /* First parameter line */
10208: while(fgets(line, MAXLINE, ficpar)) {
10209: /* If line starts with a # it is a comment */
10210: if (line[0] == '#') {
10211: numlinepar++;
10212: fputs(line,stdout);
10213: fputs(line,ficparo);
10214: fputs(line,ficlog);
10215: continue;
10216: }else
10217: break;
10218: }
10219: if((num_filled=sscanf(line,"title=%s datafile=%s lastobs=%d firstpass=%d lastpass=%d\n", \
10220: title, datafile, &lastobs, &firstpass,&lastpass)) !=EOF){
10221: if (num_filled != 5) {
10222: printf("Should be 5 parameters\n");
10223: }
1.126 brouard 10224: numlinepar++;
1.197 brouard 10225: printf("title=%s datafile=%s lastobs=%d firstpass=%d lastpass=%d\n", title, datafile, lastobs, firstpass,lastpass);
10226: }
10227: /* Second parameter line */
10228: while(fgets(line, MAXLINE, ficpar)) {
10229: /* If line starts with a # it is a comment */
10230: if (line[0] == '#') {
10231: numlinepar++;
10232: fputs(line,stdout);
10233: fputs(line,ficparo);
10234: fputs(line,ficlog);
10235: continue;
10236: }else
10237: break;
10238: }
1.223 brouard 10239: if((num_filled=sscanf(line,"ftol=%lf stepm=%d ncovcol=%d nqv=%d ntv=%d nqtv=%d nlstate=%d ndeath=%d maxwav=%d mle=%d weight=%d\n", \
10240: &ftol, &stepm, &ncovcol, &nqv, &ntv, &nqtv, &nlstate, &ndeath, &maxwav, &mle, &weightopt)) !=EOF){
10241: if (num_filled != 11) {
10242: printf("Not 11 parameters, for example:ftol=1.e-8 stepm=12 ncovcol=2 nqv=1 ntv=2 nqtv=1 nlstate=2 ndeath=1 maxwav=3 mle=1 weight=1\n");
1.209 brouard 10243: printf("but line=%s\n",line);
1.197 brouard 10244: }
1.223 brouard 10245: printf("ftol=%e stepm=%d ncovcol=%d nqv=%d ntv=%d nqtv=%d nlstate=%d ndeath=%d maxwav=%d mle=%d weight=%d\n",ftol, stepm, ncovcol, nqv, ntv, nqtv, nlstate, ndeath, maxwav, mle, weightopt);
1.126 brouard 10246: }
1.203 brouard 10247: /* ftolpl=6*ftol*1.e5; /\* 6.e-3 make convergences in less than 80 loops for the prevalence limit *\/ */
1.209 brouard 10248: /*ftolpl=6.e-4; *//* 6.e-3 make convergences in less than 80 loops for the prevalence limit */
1.197 brouard 10249: /* Third parameter line */
10250: while(fgets(line, MAXLINE, ficpar)) {
10251: /* If line starts with a # it is a comment */
10252: if (line[0] == '#') {
10253: numlinepar++;
10254: fputs(line,stdout);
10255: fputs(line,ficparo);
10256: fputs(line,ficlog);
10257: continue;
10258: }else
10259: break;
10260: }
1.201 brouard 10261: if((num_filled=sscanf(line,"model=1+age%[^.\n]", model)) !=EOF){
1.263 brouard 10262: if (num_filled == 0){
10263: printf("ERROR %d: Model should be at minimum 'model=1+age.' WITHOUT space:'%s'\n",num_filled, line);
10264: fprintf(ficlog,"ERROR %d: Model should be at minimum 'model=1+age.' WITHOUT space:'%s'\n",num_filled, line);
10265: model[0]='\0';
10266: goto end;
10267: } else if (num_filled != 1){
1.197 brouard 10268: printf("ERROR %d: Model should be at minimum 'model=1+age.' %s\n",num_filled, line);
10269: fprintf(ficlog,"ERROR %d: Model should be at minimum 'model=1+age.' %s\n",num_filled, line);
10270: model[0]='\0';
10271: goto end;
10272: }
10273: else{
10274: if (model[0]=='+'){
10275: for(i=1; i<=strlen(model);i++)
10276: modeltemp[i-1]=model[i];
1.201 brouard 10277: strcpy(model,modeltemp);
1.197 brouard 10278: }
10279: }
1.199 brouard 10280: /* printf(" model=1+age%s modeltemp= %s, model=%s\n",model, modeltemp, model);fflush(stdout); */
1.203 brouard 10281: printf("model=1+age+%s\n",model);fflush(stdout);
1.197 brouard 10282: }
10283: /* fscanf(ficpar,"title=%s datafile=%s lastobs=%d firstpass=%d lastpass=%d\nftol=%lf stepm=%d ncovcol=%d nlstate=%d ndeath=%d maxwav=%d mle=%d weight=%d model=1+age+%s\n",title, datafile, &lastobs, &firstpass,&lastpass,&ftol, &stepm, &ncovcol, &nlstate,&ndeath, &maxwav, &mle, &weightopt,model); */
10284: /* numlinepar=numlinepar+3; /\* In general *\/ */
10285: /* printf("title=%s datafile=%s lastobs=%d firstpass=%d lastpass=%d\nftol=%e stepm=%d ncovcol=%d nlstate=%d ndeath=%d maxwav=%d mle=%d weight=%d\nmodel=1+age+%s\n", title, datafile, lastobs, firstpass,lastpass,ftol, stepm, ncovcol, nlstate,ndeath, maxwav, mle, weightopt,model); */
1.223 brouard 10286: fprintf(ficparo,"title=%s datafile=%s lastobs=%d firstpass=%d lastpass=%d\nftol=%e stepm=%d ncovcol=%d nqv=%d ntv=%d nqtv=%d nlstate=%d ndeath=%d maxwav=%d mle=%d weight=%d\nmodel=1+age+%s.\n", title, datafile, lastobs, firstpass,lastpass,ftol,stepm,ncovcol, nqv, ntv, nqtv, nlstate,ndeath,maxwav, mle, weightopt,model);
10287: fprintf(ficlog,"title=%s datafile=%s lastobs=%d firstpass=%d lastpass=%d\nftol=%e stepm=%d ncovcol=%d nqv=%d ntv=%d nqtv=%d nlstate=%d ndeath=%d maxwav=%d mle=%d weight=%d\nmodel=1+age+%s.\n", title, datafile, lastobs, firstpass,lastpass,ftol,stepm,ncovcol, nqv, ntv, nqtv, nlstate,ndeath,maxwav, mle, weightopt,model);
1.126 brouard 10288: fflush(ficlog);
1.190 brouard 10289: /* if(model[0]=='#'|| model[0]== '\0'){ */
10290: if(model[0]=='#'){
1.187 brouard 10291: printf("Error in 'model' line: model should start with 'model=1+age+' and end with '.' \n \
10292: 'model=1+age+.' or 'model=1+age+V1.' or 'model=1+age+age*age+V1+V1*age.' or \n \
10293: 'model=1+age+V1+V2.' or 'model=1+age+V1+V2+V1*V2.' etc. \n"); \
10294: if(mle != -1){
10295: printf("Fix the model line and run imach with mle=-1 to get a correct template of the parameter file.\n");
10296: exit(1);
10297: }
10298: }
1.126 brouard 10299: while((c=getc(ficpar))=='#' && c!= EOF){
10300: ungetc(c,ficpar);
10301: fgets(line, MAXLINE, ficpar);
10302: numlinepar++;
1.195 brouard 10303: if(line[1]=='q'){ /* This #q will quit imach (the answer is q) */
10304: z[0]=line[1];
10305: }
10306: /* printf("****line [1] = %c \n",line[1]); */
1.141 brouard 10307: fputs(line, stdout);
10308: //puts(line);
1.126 brouard 10309: fputs(line,ficparo);
10310: fputs(line,ficlog);
10311: }
10312: ungetc(c,ficpar);
10313:
10314:
1.145 brouard 10315: covar=matrix(0,NCOVMAX,1,n); /**< used in readdata */
1.225 brouard 10316: coqvar=matrix(1,nqv,1,n); /**< Fixed quantitative covariate */
1.233 brouard 10317: cotvar=ma3x(1,maxwav,1,ntv+nqtv,1,n); /**< Time varying covariate (dummy and quantitative)*/
1.225 brouard 10318: cotqvar=ma3x(1,maxwav,1,nqtv,1,n); /**< Time varying quantitative covariate */
1.136 brouard 10319: cptcovn=0; /*Number of covariates, i.e. number of '+' in model statement plus one, indepently of n in Vn*/
10320: /* v1+v2+v3+v2*v4+v5*age makes cptcovn = 5
10321: v1+v2*age+v2*v3 makes cptcovn = 3
10322: */
10323: if (strlen(model)>1)
1.187 brouard 10324: ncovmodel=2+nbocc(model,'+')+1; /*Number of variables including intercept and age = cptcovn + intercept + age : v1+v2+v3+v2*v4+v5*age makes 5+2=7,age*age makes 3*/
1.145 brouard 10325: else
1.187 brouard 10326: ncovmodel=2; /* Constant and age */
1.133 brouard 10327: nforce= (nlstate+ndeath-1)*nlstate; /* Number of forces ij from state i to j */
10328: npar= nforce*ncovmodel; /* Number of parameters like aij*/
1.131 brouard 10329: if(npar >MAXPARM || nlstate >NLSTATEMAX || ndeath >NDEATHMAX || ncovmodel>NCOVMAX){
10330: printf("Too complex model for current IMaCh: npar=(nlstate+ndeath-1)*nlstate*ncovmodel=%d >= %d(MAXPARM) or nlstate=%d >= %d(NLSTATEMAX) or ndeath=%d >= %d(NDEATHMAX) or ncovmodel=(k+age+#of+signs)=%d(NCOVMAX) >= %d\n",npar, MAXPARM, nlstate, NLSTATEMAX, ndeath, NDEATHMAX, ncovmodel, NCOVMAX);
10331: fprintf(ficlog,"Too complex model for current IMaCh: %d >=%d(MAXPARM) or %d >=%d(NLSTATEMAX) or %d >=%d(NDEATHMAX) or %d(NCOVMAX) >=%d\n",npar, MAXPARM, nlstate, NLSTATEMAX, ndeath, NDEATHMAX, ncovmodel, NCOVMAX);
10332: fflush(stdout);
10333: fclose (ficlog);
10334: goto end;
10335: }
1.126 brouard 10336: delti3= ma3x(1,nlstate,1,nlstate+ndeath-1,1,ncovmodel);
10337: delti=delti3[1][1];
10338: /*delti=vector(1,npar); *//* Scale of each paramater (output from hesscov)*/
10339: if(mle==-1){ /* Print a wizard for help writing covariance matrix */
1.247 brouard 10340: /* We could also provide initial parameters values giving by simple logistic regression
10341: * only one way, that is without matrix product. We will have nlstate maximizations */
10342: /* for(i=1;i<nlstate;i++){ */
10343: /* /\*reducing xi for 1 to npar to 1 to ncovmodel; *\/ */
10344: /* mlikeli(ficres,p, ncovmodel, ncovmodel, nlstate, ftol, funcnoprod); */
10345: /* } */
1.126 brouard 10346: prwizard(ncovmodel, nlstate, ndeath, model, ficparo);
1.191 brouard 10347: printf(" You chose mle=-1, look at file %s for a template of covariance matrix \n",filereso);
10348: fprintf(ficlog," You chose mle=-1, look at file %s for a template of covariance matrix \n",filereso);
1.126 brouard 10349: free_ma3x(delti3,1,nlstate,1, nlstate+ndeath-1,1,ncovmodel);
10350: fclose (ficparo);
10351: fclose (ficlog);
10352: goto end;
10353: exit(0);
1.220 brouard 10354: } else if(mle==-5) { /* Main Wizard */
1.126 brouard 10355: prwizard(ncovmodel, nlstate, ndeath, model, ficparo);
1.192 brouard 10356: printf(" You chose mle=-3, look at file %s for a template of covariance matrix \n",filereso);
10357: fprintf(ficlog," You chose mle=-3, look at file %s for a template of covariance matrix \n",filereso);
1.126 brouard 10358: param= ma3x(1,nlstate,1,nlstate+ndeath-1,1,ncovmodel);
10359: matcov=matrix(1,npar,1,npar);
1.203 brouard 10360: hess=matrix(1,npar,1,npar);
1.220 brouard 10361: } else{ /* Begin of mle != -1 or -5 */
1.145 brouard 10362: /* Read guessed parameters */
1.126 brouard 10363: /* Reads comments: lines beginning with '#' */
10364: while((c=getc(ficpar))=='#' && c!= EOF){
10365: ungetc(c,ficpar);
10366: fgets(line, MAXLINE, ficpar);
10367: numlinepar++;
1.141 brouard 10368: fputs(line,stdout);
1.126 brouard 10369: fputs(line,ficparo);
10370: fputs(line,ficlog);
10371: }
10372: ungetc(c,ficpar);
10373:
10374: param= ma3x(1,nlstate,1,nlstate+ndeath-1,1,ncovmodel);
1.251 brouard 10375: paramstart= ma3x(1,nlstate,1,nlstate+ndeath-1,1,ncovmodel);
1.126 brouard 10376: for(i=1; i <=nlstate; i++){
1.234 brouard 10377: j=0;
1.126 brouard 10378: for(jj=1; jj <=nlstate+ndeath; jj++){
1.234 brouard 10379: if(jj==i) continue;
10380: j++;
10381: fscanf(ficpar,"%1d%1d",&i1,&j1);
10382: if ((i1 != i) || (j1 != jj)){
10383: printf("Error in line parameters number %d, %1d%1d instead of %1d%1d \n \
1.126 brouard 10384: It might be a problem of design; if ncovcol and the model are correct\n \
10385: run imach with mle=-1 to get a correct template of the parameter file.\n",numlinepar, i,j, i1, j1);
1.234 brouard 10386: exit(1);
10387: }
10388: fprintf(ficparo,"%1d%1d",i1,j1);
10389: if(mle==1)
10390: printf("%1d%1d",i,jj);
10391: fprintf(ficlog,"%1d%1d",i,jj);
10392: for(k=1; k<=ncovmodel;k++){
10393: fscanf(ficpar," %lf",¶m[i][j][k]);
10394: if(mle==1){
10395: printf(" %lf",param[i][j][k]);
10396: fprintf(ficlog," %lf",param[i][j][k]);
10397: }
10398: else
10399: fprintf(ficlog," %lf",param[i][j][k]);
10400: fprintf(ficparo," %lf",param[i][j][k]);
10401: }
10402: fscanf(ficpar,"\n");
10403: numlinepar++;
10404: if(mle==1)
10405: printf("\n");
10406: fprintf(ficlog,"\n");
10407: fprintf(ficparo,"\n");
1.126 brouard 10408: }
10409: }
10410: fflush(ficlog);
1.234 brouard 10411:
1.251 brouard 10412: /* Reads parameters values */
1.126 brouard 10413: p=param[1][1];
1.251 brouard 10414: pstart=paramstart[1][1];
1.126 brouard 10415:
10416: /* Reads comments: lines beginning with '#' */
10417: while((c=getc(ficpar))=='#' && c!= EOF){
10418: ungetc(c,ficpar);
10419: fgets(line, MAXLINE, ficpar);
10420: numlinepar++;
1.141 brouard 10421: fputs(line,stdout);
1.126 brouard 10422: fputs(line,ficparo);
10423: fputs(line,ficlog);
10424: }
10425: ungetc(c,ficpar);
10426:
10427: for(i=1; i <=nlstate; i++){
10428: for(j=1; j <=nlstate+ndeath-1; j++){
1.234 brouard 10429: fscanf(ficpar,"%1d%1d",&i1,&j1);
10430: if ( (i1-i) * (j1-j) != 0){
10431: printf("Error in line parameters number %d, %1d%1d instead of %1d%1d \n",numlinepar, i,j, i1, j1);
10432: exit(1);
10433: }
10434: printf("%1d%1d",i,j);
10435: fprintf(ficparo,"%1d%1d",i1,j1);
10436: fprintf(ficlog,"%1d%1d",i1,j1);
10437: for(k=1; k<=ncovmodel;k++){
10438: fscanf(ficpar,"%le",&delti3[i][j][k]);
10439: printf(" %le",delti3[i][j][k]);
10440: fprintf(ficparo," %le",delti3[i][j][k]);
10441: fprintf(ficlog," %le",delti3[i][j][k]);
10442: }
10443: fscanf(ficpar,"\n");
10444: numlinepar++;
10445: printf("\n");
10446: fprintf(ficparo,"\n");
10447: fprintf(ficlog,"\n");
1.126 brouard 10448: }
10449: }
10450: fflush(ficlog);
1.234 brouard 10451:
1.145 brouard 10452: /* Reads covariance matrix */
1.126 brouard 10453: delti=delti3[1][1];
1.220 brouard 10454:
10455:
1.126 brouard 10456: /* free_ma3x(delti3,1,nlstate,1,nlstate+ndeath-1,1,ncovmodel); */ /* Hasn't to to freed here otherwise delti is no more allocated */
1.220 brouard 10457:
1.126 brouard 10458: /* Reads comments: lines beginning with '#' */
10459: while((c=getc(ficpar))=='#' && c!= EOF){
10460: ungetc(c,ficpar);
10461: fgets(line, MAXLINE, ficpar);
10462: numlinepar++;
1.141 brouard 10463: fputs(line,stdout);
1.126 brouard 10464: fputs(line,ficparo);
10465: fputs(line,ficlog);
10466: }
10467: ungetc(c,ficpar);
1.220 brouard 10468:
1.126 brouard 10469: matcov=matrix(1,npar,1,npar);
1.203 brouard 10470: hess=matrix(1,npar,1,npar);
1.131 brouard 10471: for(i=1; i <=npar; i++)
10472: for(j=1; j <=npar; j++) matcov[i][j]=0.;
1.220 brouard 10473:
1.194 brouard 10474: /* Scans npar lines */
1.126 brouard 10475: for(i=1; i <=npar; i++){
1.226 brouard 10476: count=fscanf(ficpar,"%1d%1d%d",&i1,&j1,&jk);
1.194 brouard 10477: if(count != 3){
1.226 brouard 10478: printf("Error! Error in parameter file %s at line %d after line starting with %1d%1d%1d\n\
1.194 brouard 10479: This is probably because your covariance matrix doesn't \n contain exactly %d lines corresponding to your model line '1+age+%s'.\n\
10480: Please run with mle=-1 to get a correct covariance matrix.\n",optionfile,numlinepar, i1,j1,jk, npar, model);
1.226 brouard 10481: fprintf(ficlog,"Error! Error in parameter file %s at line %d after line starting with %1d%1d%1d\n\
1.194 brouard 10482: This is probably because your covariance matrix doesn't \n contain exactly %d lines corresponding to your model line '1+age+%s'.\n\
10483: Please run with mle=-1 to get a correct covariance matrix.\n",optionfile,numlinepar, i1,j1,jk, npar, model);
1.226 brouard 10484: exit(1);
1.220 brouard 10485: }else{
1.226 brouard 10486: if(mle==1)
10487: printf("%1d%1d%d",i1,j1,jk);
10488: }
10489: fprintf(ficlog,"%1d%1d%d",i1,j1,jk);
10490: fprintf(ficparo,"%1d%1d%d",i1,j1,jk);
1.126 brouard 10491: for(j=1; j <=i; j++){
1.226 brouard 10492: fscanf(ficpar," %le",&matcov[i][j]);
10493: if(mle==1){
10494: printf(" %.5le",matcov[i][j]);
10495: }
10496: fprintf(ficlog," %.5le",matcov[i][j]);
10497: fprintf(ficparo," %.5le",matcov[i][j]);
1.126 brouard 10498: }
10499: fscanf(ficpar,"\n");
10500: numlinepar++;
10501: if(mle==1)
1.220 brouard 10502: printf("\n");
1.126 brouard 10503: fprintf(ficlog,"\n");
10504: fprintf(ficparo,"\n");
10505: }
1.194 brouard 10506: /* End of read covariance matrix npar lines */
1.126 brouard 10507: for(i=1; i <=npar; i++)
10508: for(j=i+1;j<=npar;j++)
1.226 brouard 10509: matcov[i][j]=matcov[j][i];
1.126 brouard 10510:
10511: if(mle==1)
10512: printf("\n");
10513: fprintf(ficlog,"\n");
10514:
10515: fflush(ficlog);
10516:
10517: /*-------- Rewriting parameter file ----------*/
10518: strcpy(rfileres,"r"); /* "Rparameterfile */
10519: strcat(rfileres,optionfilefiname); /* Parameter file first name*/
10520: strcat(rfileres,"."); /* */
10521: strcat(rfileres,optionfilext); /* Other files have txt extension */
10522: if((ficres =fopen(rfileres,"w"))==NULL) {
1.201 brouard 10523: printf("Problem writing new parameter file: %s\n", rfileres);goto end;
10524: fprintf(ficlog,"Problem writing new parameter file: %s\n", rfileres);goto end;
1.126 brouard 10525: }
10526: fprintf(ficres,"#%s\n",version);
10527: } /* End of mle != -3 */
1.218 brouard 10528:
1.186 brouard 10529: /* Main data
10530: */
1.126 brouard 10531: n= lastobs;
10532: num=lvector(1,n);
10533: moisnais=vector(1,n);
10534: annais=vector(1,n);
10535: moisdc=vector(1,n);
10536: andc=vector(1,n);
1.220 brouard 10537: weight=vector(1,n);
1.126 brouard 10538: agedc=vector(1,n);
10539: cod=ivector(1,n);
1.220 brouard 10540: for(i=1;i<=n;i++){
1.234 brouard 10541: num[i]=0;
10542: moisnais[i]=0;
10543: annais[i]=0;
10544: moisdc[i]=0;
10545: andc[i]=0;
10546: agedc[i]=0;
10547: cod[i]=0;
10548: weight[i]=1.0; /* Equal weights, 1 by default */
10549: }
1.126 brouard 10550: mint=matrix(1,maxwav,1,n);
10551: anint=matrix(1,maxwav,1,n);
1.131 brouard 10552: s=imatrix(1,maxwav+1,1,n); /* s[i][j] health state for wave i and individual j */
1.126 brouard 10553: tab=ivector(1,NCOVMAX);
1.144 brouard 10554: ncodemax=ivector(1,NCOVMAX); /* Number of code per covariate; if O and 1 only, 2**ncov; V1+V2+V3+V4=>16 */
1.192 brouard 10555: ncodemaxwundef=ivector(1,NCOVMAX); /* Number of code per covariate; if - 1 O and 1 only, 2**ncov; V1+V2+V3+V4=>16 */
1.126 brouard 10556:
1.136 brouard 10557: /* Reads data from file datafile */
10558: if (readdata(datafile, firstobs, lastobs, &imx)==1)
10559: goto end;
10560:
10561: /* Calculation of the number of parameters from char model */
1.234 brouard 10562: /* modelsav=V2+V1+V4+age*V3 strb=age*V3 stra=V2+V1+V4
1.137 brouard 10563: k=4 (age*V3) Tvar[k=4]= 3 (from V3) Tag[cptcovage=1]=4
10564: k=3 V4 Tvar[k=3]= 4 (from V4)
10565: k=2 V1 Tvar[k=2]= 1 (from V1)
10566: k=1 Tvar[1]=2 (from V2)
1.234 brouard 10567: */
10568:
10569: Tvar=ivector(1,NCOVMAX); /* Was 15 changed to NCOVMAX. */
10570: TvarsDind=ivector(1,NCOVMAX); /* */
10571: TvarsD=ivector(1,NCOVMAX); /* */
10572: TvarsQind=ivector(1,NCOVMAX); /* */
10573: TvarsQ=ivector(1,NCOVMAX); /* */
1.232 brouard 10574: TvarF=ivector(1,NCOVMAX); /* */
10575: TvarFind=ivector(1,NCOVMAX); /* */
10576: TvarV=ivector(1,NCOVMAX); /* */
10577: TvarVind=ivector(1,NCOVMAX); /* */
10578: TvarA=ivector(1,NCOVMAX); /* */
10579: TvarAind=ivector(1,NCOVMAX); /* */
1.231 brouard 10580: TvarFD=ivector(1,NCOVMAX); /* */
10581: TvarFDind=ivector(1,NCOVMAX); /* */
10582: TvarFQ=ivector(1,NCOVMAX); /* */
10583: TvarFQind=ivector(1,NCOVMAX); /* */
10584: TvarVD=ivector(1,NCOVMAX); /* */
10585: TvarVDind=ivector(1,NCOVMAX); /* */
10586: TvarVQ=ivector(1,NCOVMAX); /* */
10587: TvarVQind=ivector(1,NCOVMAX); /* */
10588:
1.230 brouard 10589: Tvalsel=vector(1,NCOVMAX); /* */
1.233 brouard 10590: Tvarsel=ivector(1,NCOVMAX); /* */
1.226 brouard 10591: Typevar=ivector(-1,NCOVMAX); /* -1 to 2 */
10592: Fixed=ivector(-1,NCOVMAX); /* -1 to 3 */
10593: Dummy=ivector(-1,NCOVMAX); /* -1 to 3 */
1.137 brouard 10594: /* V2+V1+V4+age*V3 is a model with 4 covariates (3 plus signs).
10595: For each model-covariate stores the data-covariate id. Tvar[1]=2, Tvar[2]=1, Tvar[3]=4,
10596: Tvar[4=age*V3] is 3 and 'age' is recorded in Tage.
10597: */
10598: /* For model-covariate k tells which data-covariate to use but
10599: because this model-covariate is a construction we invent a new column
10600: ncovcol + k1
10601: If already ncovcol=4 and model=V2+V1+V1*V4+age*V3
10602: Tvar[3=V1*V4]=4+1 etc */
1.227 brouard 10603: Tprod=ivector(1,NCOVMAX); /* Gives the k position of the k1 product */
10604: Tposprod=ivector(1,NCOVMAX); /* Gives the k1 product from the k position */
1.137 brouard 10605: /* Tprod[k1=1]=3(=V1*V4) for V2+V1+V1*V4+age*V3
10606: if V2+V1+V1*V4+age*V3+V3*V2 TProd[k1=2]=5 (V3*V2)
1.227 brouard 10607: Tposprod[k]=k1 , Tposprod[3]=1, Tposprod[5]=2
1.137 brouard 10608: */
1.145 brouard 10609: Tvaraff=ivector(1,NCOVMAX); /* Unclear */
10610: Tvard=imatrix(1,NCOVMAX,1,2); /* n=Tvard[k1][1] and m=Tvard[k1][2] gives the couple n,m of the k1 th product Vn*Vm
1.141 brouard 10611: * For V3*V2 (in V2+V1+V1*V4+age*V3+V3*V2), V3*V2 position is 2nd.
10612: * Tvard[k1=2][1]=3 (V3) Tvard[k1=2][2]=2(V2) */
1.145 brouard 10613: Tage=ivector(1,NCOVMAX); /* Gives the covariate id of covariates associated with age: V2 + V1 + age*V4 + V3*age
1.137 brouard 10614: 4 covariates (3 plus signs)
10615: Tage[1=V3*age]= 4; Tage[2=age*V4] = 3
10616: */
1.230 brouard 10617: Tmodelind=ivector(1,NCOVMAX);/** gives the k model position of an
1.227 brouard 10618: * individual dummy, fixed or varying:
10619: * Tmodelind[Tvaraff[3]]=9,Tvaraff[1]@9={4,
10620: * 3, 1, 0, 0, 0, 0, 0, 0},
1.230 brouard 10621: * model=V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 ,
10622: * V1 df, V2 qf, V3 & V4 dv, V5 qv
10623: * Tmodelind[1]@9={9,0,3,2,}*/
10624: TmodelInvind=ivector(1,NCOVMAX); /* TmodelInvind=Tvar[k]- ncovcol-nqv={5-2-1=2,*/
10625: TmodelInvQind=ivector(1,NCOVMAX);/** gives the k model position of an
1.228 brouard 10626: * individual quantitative, fixed or varying:
10627: * Tmodelqind[1]=1,Tvaraff[1]@9={4,
10628: * 3, 1, 0, 0, 0, 0, 0, 0},
10629: * model=V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1*/
1.186 brouard 10630: /* Main decodemodel */
10631:
1.187 brouard 10632:
1.223 brouard 10633: if(decodemodel(model, lastobs) == 1) /* In order to get Tvar[k] V4+V3+V5 p Tvar[1]@3 = {4, 3, 5}*/
1.136 brouard 10634: goto end;
10635:
1.137 brouard 10636: if((double)(lastobs-imx)/(double)imx > 1.10){
10637: nbwarn++;
10638: printf("Warning: The value of parameter lastobs=%d is big compared to the \n effective number of cases imx=%d, please adjust, \n otherwise you are allocating more memory than necessary.\n",lastobs, imx);
10639: fprintf(ficlog,"Warning: The value of parameter lastobs=%d is big compared to the \n effective number of cases imx=%d, please adjust, \n otherwise you are allocating more memory than necessary.\n",lastobs, imx);
10640: }
1.136 brouard 10641: /* if(mle==1){*/
1.137 brouard 10642: if (weightopt != 1) { /* Maximisation without weights. We can have weights different from 1 but want no weight*/
10643: for(i=1;i<=imx;i++) weight[i]=1.0; /* changed to imx */
1.136 brouard 10644: }
10645:
10646: /*-calculation of age at interview from date of interview and age at death -*/
10647: agev=matrix(1,maxwav,1,imx);
10648:
10649: if(calandcheckages(imx, maxwav, &agemin, &agemax, &nberr, &nbwarn) == 1)
10650: goto end;
10651:
1.126 brouard 10652:
1.136 brouard 10653: agegomp=(int)agemin;
10654: free_vector(moisnais,1,n);
10655: free_vector(annais,1,n);
1.126 brouard 10656: /* free_matrix(mint,1,maxwav,1,n);
10657: free_matrix(anint,1,maxwav,1,n);*/
1.215 brouard 10658: /* free_vector(moisdc,1,n); */
10659: /* free_vector(andc,1,n); */
1.145 brouard 10660: /* */
10661:
1.126 brouard 10662: wav=ivector(1,imx);
1.214 brouard 10663: /* dh=imatrix(1,lastpass-firstpass+1,1,imx); */
10664: /* bh=imatrix(1,lastpass-firstpass+1,1,imx); */
10665: /* mw=imatrix(1,lastpass-firstpass+1,1,imx); */
10666: dh=imatrix(1,lastpass-firstpass+2,1,imx); /* We are adding a wave if status is unknown at last wave but death occurs after last wave.*/
10667: bh=imatrix(1,lastpass-firstpass+2,1,imx);
10668: mw=imatrix(1,lastpass-firstpass+2,1,imx);
1.126 brouard 10669:
10670: /* Concatenates waves */
1.214 brouard 10671: /* Concatenates waves: wav[i] is the number of effective (useful waves) of individual i.
10672: Death is a valid wave (if date is known).
10673: mw[mi][i] is the number of (mi=1 to wav[i]) effective wave out of mi of individual i
10674: dh[m][i] or dh[mw[mi][i]][i] is the delay between two effective waves m=mw[mi][i]
10675: and mw[mi+1][i]. dh depends on stepm.
10676: */
10677:
1.126 brouard 10678: concatwav(wav, dh, bh, mw, s, agedc, agev, firstpass, lastpass, imx, nlstate, stepm);
1.248 brouard 10679: /* Concatenates waves */
1.145 brouard 10680:
1.215 brouard 10681: free_vector(moisdc,1,n);
10682: free_vector(andc,1,n);
10683:
1.126 brouard 10684: /* Routine tricode is to calculate cptcoveff (real number of unique covariates) and to associate covariable number and modality */
10685: nbcode=imatrix(0,NCOVMAX,0,NCOVMAX);
10686: ncodemax[1]=1;
1.145 brouard 10687: Ndum =ivector(-1,NCOVMAX);
1.225 brouard 10688: cptcoveff=0;
1.220 brouard 10689: if (ncovmodel-nagesqr > 2 ){ /* That is if covariate other than cst, age and age*age */
10690: tricode(&cptcoveff,Tvar,nbcode,imx, Ndum); /**< Fills nbcode[Tvar[j]][l]; */
1.227 brouard 10691: }
10692:
10693: ncovcombmax=pow(2,cptcoveff);
10694: invalidvarcomb=ivector(1, ncovcombmax);
10695: for(i=1;i<ncovcombmax;i++)
10696: invalidvarcomb[i]=0;
10697:
1.211 brouard 10698: /* Nbcode gives the value of the lth modality (currently 1 to 2) of jth covariate, in
1.186 brouard 10699: V2+V1*age, there are 3 covariates Tvar[2]=1 (V1).*/
1.211 brouard 10700: /* 1 to ncodemax[j] which is the maximum value of this jth covariate */
1.227 brouard 10701:
1.200 brouard 10702: /* codtab=imatrix(1,100,1,10);*/ /* codtab[h,k]=( (h-1) - mod(k-1,2**(k-1) )/2**(k-1) */
1.198 brouard 10703: /*printf(" codtab[1,1],codtab[100,10]=%d,%d\n", codtab[1][1],codtabm(100,10));*/
1.186 brouard 10704: /* codtab gives the value 1 or 2 of the hth combination of k covariates (1 or 2).*/
1.211 brouard 10705: /* nbcode[Tvaraff[j]][codtabm(h,j)]) : if there are only 2 modalities for a covariate j,
10706: * codtabm(h,j) gives its value classified at position h and nbcode gives how it is coded
10707: * (currently 0 or 1) in the data.
10708: * In a loop on h=1 to 2**k, and a loop on j (=1 to k), we get the value of
10709: * corresponding modality (h,j).
10710: */
10711:
1.145 brouard 10712: h=0;
10713: /*if (cptcovn > 0) */
1.126 brouard 10714: m=pow(2,cptcoveff);
10715:
1.144 brouard 10716: /**< codtab(h,k) k = codtab[h,k]=( (h-1) - mod(k-1,2**(k-1) )/2**(k-1) + 1
1.211 brouard 10717: * For k=4 covariates, h goes from 1 to m=2**k
10718: * codtabm(h,k)= (1 & (h-1) >> (k-1)) + 1;
10719: * #define codtabm(h,k) (1 & (h-1) >> (k-1))+1
1.186 brouard 10720: * h\k 1 2 3 4
1.143 brouard 10721: *______________________________
10722: * 1 i=1 1 i=1 1 i=1 1 i=1 1
10723: * 2 2 1 1 1
10724: * 3 i=2 1 2 1 1
10725: * 4 2 2 1 1
10726: * 5 i=3 1 i=2 1 2 1
10727: * 6 2 1 2 1
10728: * 7 i=4 1 2 2 1
10729: * 8 2 2 2 1
1.197 brouard 10730: * 9 i=5 1 i=3 1 i=2 1 2
10731: * 10 2 1 1 2
10732: * 11 i=6 1 2 1 2
10733: * 12 2 2 1 2
10734: * 13 i=7 1 i=4 1 2 2
10735: * 14 2 1 2 2
10736: * 15 i=8 1 2 2 2
10737: * 16 2 2 2 2
1.143 brouard 10738: */
1.212 brouard 10739: /* How to do the opposite? From combination h (=1 to 2**k) how to get the value on the covariates? */
1.211 brouard 10740: /* from h=5 and m, we get then number of covariates k=log(m)/log(2)=4
10741: * and the value of each covariate?
10742: * V1=1, V2=1, V3=2, V4=1 ?
10743: * h-1=4 and 4 is 0100 or reverse 0010, and +1 is 1121 ok.
10744: * h=6, 6-1=5, 5 is 0101, 1010, 2121, V1=2nd, V2=1st, V3=2nd, V4=1st.
10745: * In order to get the real value in the data, we use nbcode
10746: * nbcode[Tvar[3][2nd]]=1 and nbcode[Tvar[4][1]]=0
10747: * We are keeping this crazy system in order to be able (in the future?)
10748: * to have more than 2 values (0 or 1) for a covariate.
10749: * #define codtabm(h,k) (1 & (h-1) >> (k-1))+1
10750: * h=6, k=2? h-1=5=0101, reverse 1010, +1=2121, k=2nd position: value is 1: codtabm(6,2)=1
10751: * bbbbbbbb
10752: * 76543210
10753: * h-1 00000101 (6-1=5)
1.219 brouard 10754: *(h-1)>>(k-1)= 00000010 >> (2-1) = 1 right shift
1.211 brouard 10755: * &
10756: * 1 00000001 (1)
1.219 brouard 10757: * 00000000 = 1 & ((h-1) >> (k-1))
10758: * +1= 00000001 =1
1.211 brouard 10759: *
10760: * h=14, k=3 => h'=h-1=13, k'=k-1=2
10761: * h' 1101 =2^3+2^2+0x2^1+2^0
10762: * >>k' 11
10763: * & 00000001
10764: * = 00000001
10765: * +1 = 00000010=2 = codtabm(14,3)
10766: * Reverse h=6 and m=16?
10767: * cptcoveff=log(16)/log(2)=4 covariate: 6-1=5=0101 reversed=1010 +1=2121 =>V1=2, V2=1, V3=2, V4=1.
10768: * for (j=1 to cptcoveff) Vj=decodtabm(j,h,cptcoveff)
10769: * decodtabm(h,j,cptcoveff)= (((h-1) >> (j-1)) & 1) +1
10770: * decodtabm(h,j,cptcoveff)= (h <= (1<<cptcoveff)?(((h-1) >> (j-1)) & 1) +1 : -1)
10771: * V3=decodtabm(14,3,2**4)=2
10772: * h'=13 1101 =2^3+2^2+0x2^1+2^0
10773: *(h-1) >> (j-1) 0011 =13 >> 2
10774: * &1 000000001
10775: * = 000000001
10776: * +1= 000000010 =2
10777: * 2211
10778: * V1=1+1, V2=0+1, V3=1+1, V4=1+1
10779: * V3=2
1.220 brouard 10780: * codtabm and decodtabm are identical
1.211 brouard 10781: */
10782:
1.145 brouard 10783:
10784: free_ivector(Ndum,-1,NCOVMAX);
10785:
10786:
1.126 brouard 10787:
1.186 brouard 10788: /* Initialisation of ----------- gnuplot -------------*/
1.126 brouard 10789: strcpy(optionfilegnuplot,optionfilefiname);
10790: if(mle==-3)
1.201 brouard 10791: strcat(optionfilegnuplot,"-MORT_");
1.126 brouard 10792: strcat(optionfilegnuplot,".gp");
10793:
10794: if((ficgp=fopen(optionfilegnuplot,"w"))==NULL) {
10795: printf("Problem with file %s",optionfilegnuplot);
10796: }
10797: else{
1.204 brouard 10798: fprintf(ficgp,"\n# IMaCh-%s\n", version);
1.126 brouard 10799: fprintf(ficgp,"# %s\n", optionfilegnuplot);
1.141 brouard 10800: //fprintf(ficgp,"set missing 'NaNq'\n");
10801: fprintf(ficgp,"set datafile missing 'NaNq'\n");
1.126 brouard 10802: }
10803: /* fclose(ficgp);*/
1.186 brouard 10804:
10805:
10806: /* Initialisation of --------- index.htm --------*/
1.126 brouard 10807:
10808: strcpy(optionfilehtm,optionfilefiname); /* Main html file */
10809: if(mle==-3)
1.201 brouard 10810: strcat(optionfilehtm,"-MORT_");
1.126 brouard 10811: strcat(optionfilehtm,".htm");
10812: if((fichtm=fopen(optionfilehtm,"w"))==NULL) {
1.131 brouard 10813: printf("Problem with %s \n",optionfilehtm);
10814: exit(0);
1.126 brouard 10815: }
10816:
10817: strcpy(optionfilehtmcov,optionfilefiname); /* Only for matrix of covariance */
10818: strcat(optionfilehtmcov,"-cov.htm");
10819: if((fichtmcov=fopen(optionfilehtmcov,"w"))==NULL) {
10820: printf("Problem with %s \n",optionfilehtmcov), exit(0);
10821: }
10822: else{
10823: fprintf(fichtmcov,"<html><head>\n<title>IMaCh Cov %s</title></head>\n <body><font size=\"2\">%s <br> %s</font> \
10824: <hr size=\"2\" color=\"#EC5E5E\"> \n\
1.204 brouard 10825: Title=%s <br>Datafile=%s Firstpass=%d Lastpass=%d Stepm=%d Weight=%d Model=1+age+%s<br>\n",\
1.126 brouard 10826: optionfilehtmcov,version,fullversion,title,datafile,firstpass,lastpass,stepm, weightopt, model);
10827: }
10828:
1.213 brouard 10829: fprintf(fichtm,"<html><head>\n<head>\n<meta charset=\"utf-8\"/><meta http-equiv=\"Content-Type\" content=\"text/html; charset=utf-8\" />\n<title>IMaCh %s</title></head>\n <body><font size=\"7\"><a href=http:/euroreves.ined.fr/imach>IMaCh for Interpolated Markov Chain</a> </font><br>\n<font size=\"3\">Sponsored by Copyright (C) 2002-2015 <a href=http://www.ined.fr>INED</a>-EUROREVES-Institut de longévité-2013-2016-Japan Society for the Promotion of Sciences 日本学術振興会 (<a href=https://www.jsps.go.jp/english/e-grants/>Grant-in-Aid for Scientific Research 25293121</a>) - <a href=https://software.intel.com/en-us>Intel Software 2015-2018</a></font><br> \
1.204 brouard 10830: <hr size=\"2\" color=\"#EC5E5E\"> \n\
10831: <font size=\"2\">IMaCh-%s <br> %s</font> \
1.126 brouard 10832: <hr size=\"2\" color=\"#EC5E5E\"> \n\
1.204 brouard 10833: Title=%s <br>Datafile=%s Firstpass=%d Lastpass=%d Stepm=%d Weight=%d Model=1+age+%s<br>\n\
1.126 brouard 10834: \n\
10835: <hr size=\"2\" color=\"#EC5E5E\">\
10836: <ul><li><h4>Parameter files</h4>\n\
10837: - Parameter file: <a href=\"%s.%s\">%s.%s</a><br>\n\
10838: - Copy of the parameter file: <a href=\"o%s\">o%s</a><br>\n\
10839: - Log file of the run: <a href=\"%s\">%s</a><br>\n\
10840: - Gnuplot file name: <a href=\"%s\">%s</a><br>\n\
10841: - Date and time at start: %s</ul>\n",\
10842: optionfilehtm,version,fullversion,title,datafile,firstpass,lastpass,stepm, weightopt, model,\
10843: optionfilefiname,optionfilext,optionfilefiname,optionfilext,\
10844: fileres,fileres,\
10845: filelog,filelog,optionfilegnuplot,optionfilegnuplot,strstart);
10846: fflush(fichtm);
10847:
10848: strcpy(pathr,path);
10849: strcat(pathr,optionfilefiname);
1.184 brouard 10850: #ifdef WIN32
10851: _chdir(optionfilefiname); /* Move to directory named optionfile */
10852: #else
1.126 brouard 10853: chdir(optionfilefiname); /* Move to directory named optionfile */
1.184 brouard 10854: #endif
10855:
1.126 brouard 10856:
1.220 brouard 10857: /* Calculates basic frequencies. Computes observed prevalence at single age
10858: and for any valid combination of covariates
1.126 brouard 10859: and prints on file fileres'p'. */
1.251 brouard 10860: freqsummary(fileres, p, pstart, agemin, agemax, s, agev, nlstate, imx, Tvaraff, invalidvarcomb, nbcode, ncodemax,mint,anint,strstart, \
1.227 brouard 10861: firstpass, lastpass, stepm, weightopt, model);
1.126 brouard 10862:
10863: fprintf(fichtm,"\n");
10864: fprintf(fichtm,"<br>Total number of observations=%d <br>\n\
10865: Youngest age at first (selected) pass %.2f, oldest age %.2f<br>\n\
10866: Interval (in months) between two waves: Min=%d Max=%d Mean=%.2lf<br>\n",\
10867: imx,agemin,agemax,jmin,jmax,jmean);
10868: pmmij= matrix(1,nlstate+ndeath,1,nlstate+ndeath); /* creation */
1.220 brouard 10869: oldms= matrix(1,nlstate+ndeath,1,nlstate+ndeath); /* creation */
10870: newms= matrix(1,nlstate+ndeath,1,nlstate+ndeath); /* creation */
10871: savms= matrix(1,nlstate+ndeath,1,nlstate+ndeath); /* creation */
10872: oldm=oldms; newm=newms; savm=savms; /* Keeps fixed addresses to free */
1.218 brouard 10873:
1.126 brouard 10874: /* For Powell, parameters are in a vector p[] starting at p[1]
10875: so we point p on param[1][1] so that p[1] maps on param[1][1][1] */
10876: p=param[1][1]; /* *(*(*(param +1)+1)+0) */
10877:
10878: globpr=0; /* To get the number ipmx of contributions and the sum of weights*/
1.186 brouard 10879: /* For mortality only */
1.126 brouard 10880: if (mle==-3){
1.136 brouard 10881: ximort=matrix(1,NDIM,1,NDIM);
1.248 brouard 10882: for(i=1;i<=NDIM;i++)
10883: for(j=1;j<=NDIM;j++)
10884: ximort[i][j]=0.;
1.186 brouard 10885: /* ximort=gsl_matrix_alloc(1,NDIM,1,NDIM); */
1.126 brouard 10886: cens=ivector(1,n);
10887: ageexmed=vector(1,n);
10888: agecens=vector(1,n);
10889: dcwave=ivector(1,n);
1.223 brouard 10890:
1.126 brouard 10891: for (i=1; i<=imx; i++){
10892: dcwave[i]=-1;
10893: for (m=firstpass; m<=lastpass; m++)
1.226 brouard 10894: if (s[m][i]>nlstate) {
10895: dcwave[i]=m;
10896: /* printf("i=%d j=%d s=%d dcwave=%d\n",i,j, s[j][i],dcwave[i]);*/
10897: break;
10898: }
1.126 brouard 10899: }
1.226 brouard 10900:
1.126 brouard 10901: for (i=1; i<=imx; i++) {
10902: if (wav[i]>0){
1.226 brouard 10903: ageexmed[i]=agev[mw[1][i]][i];
10904: j=wav[i];
10905: agecens[i]=1.;
10906:
10907: if (ageexmed[i]> 1 && wav[i] > 0){
10908: agecens[i]=agev[mw[j][i]][i];
10909: cens[i]= 1;
10910: }else if (ageexmed[i]< 1)
10911: cens[i]= -1;
10912: if (agedc[i]< AGESUP && agedc[i]>1 && dcwave[i]>firstpass && dcwave[i]<=lastpass)
10913: cens[i]=0 ;
1.126 brouard 10914: }
10915: else cens[i]=-1;
10916: }
10917:
10918: for (i=1;i<=NDIM;i++) {
10919: for (j=1;j<=NDIM;j++)
1.226 brouard 10920: ximort[i][j]=(i == j ? 1.0 : 0.0);
1.126 brouard 10921: }
10922:
1.145 brouard 10923: /*p[1]=0.0268; p[NDIM]=0.083;*/
1.126 brouard 10924: /*printf("%lf %lf", p[1], p[2]);*/
10925:
10926:
1.136 brouard 10927: #ifdef GSL
10928: printf("GSL optimization\n"); fprintf(ficlog,"Powell\n");
1.162 brouard 10929: #else
1.126 brouard 10930: printf("Powell\n"); fprintf(ficlog,"Powell\n");
1.136 brouard 10931: #endif
1.201 brouard 10932: strcpy(filerespow,"POW-MORT_");
10933: strcat(filerespow,fileresu);
1.126 brouard 10934: if((ficrespow=fopen(filerespow,"w"))==NULL) {
10935: printf("Problem with resultfile: %s\n", filerespow);
10936: fprintf(ficlog,"Problem with resultfile: %s\n", filerespow);
10937: }
1.136 brouard 10938: #ifdef GSL
10939: fprintf(ficrespow,"# GSL optimization\n# iter -2*LL");
1.162 brouard 10940: #else
1.126 brouard 10941: fprintf(ficrespow,"# Powell\n# iter -2*LL");
1.136 brouard 10942: #endif
1.126 brouard 10943: /* for (i=1;i<=nlstate;i++)
10944: for(j=1;j<=nlstate+ndeath;j++)
10945: if(j!=i)fprintf(ficrespow," p%1d%1d",i,j);
10946: */
10947: fprintf(ficrespow,"\n");
1.136 brouard 10948: #ifdef GSL
10949: /* gsl starts here */
10950: T = gsl_multimin_fminimizer_nmsimplex;
10951: gsl_multimin_fminimizer *sfm = NULL;
10952: gsl_vector *ss, *x;
10953: gsl_multimin_function minex_func;
10954:
10955: /* Initial vertex size vector */
10956: ss = gsl_vector_alloc (NDIM);
10957:
10958: if (ss == NULL){
10959: GSL_ERROR_VAL ("failed to allocate space for ss", GSL_ENOMEM, 0);
10960: }
10961: /* Set all step sizes to 1 */
10962: gsl_vector_set_all (ss, 0.001);
10963:
10964: /* Starting point */
1.126 brouard 10965:
1.136 brouard 10966: x = gsl_vector_alloc (NDIM);
10967:
10968: if (x == NULL){
10969: gsl_vector_free(ss);
10970: GSL_ERROR_VAL ("failed to allocate space for x", GSL_ENOMEM, 0);
10971: }
10972:
10973: /* Initialize method and iterate */
10974: /* p[1]=0.0268; p[NDIM]=0.083; */
1.186 brouard 10975: /* gsl_vector_set(x, 0, 0.0268); */
10976: /* gsl_vector_set(x, 1, 0.083); */
1.136 brouard 10977: gsl_vector_set(x, 0, p[1]);
10978: gsl_vector_set(x, 1, p[2]);
10979:
10980: minex_func.f = &gompertz_f;
10981: minex_func.n = NDIM;
10982: minex_func.params = (void *)&p; /* ??? */
10983:
10984: sfm = gsl_multimin_fminimizer_alloc (T, NDIM);
10985: gsl_multimin_fminimizer_set (sfm, &minex_func, x, ss);
10986:
10987: printf("Iterations beginning .....\n\n");
10988: printf("Iter. # Intercept Slope -Log Likelihood Simplex size\n");
10989:
10990: iteri=0;
10991: while (rval == GSL_CONTINUE){
10992: iteri++;
10993: status = gsl_multimin_fminimizer_iterate(sfm);
10994:
10995: if (status) printf("error: %s\n", gsl_strerror (status));
10996: fflush(0);
10997:
10998: if (status)
10999: break;
11000:
11001: rval = gsl_multimin_test_size (gsl_multimin_fminimizer_size (sfm), 1e-6);
11002: ssval = gsl_multimin_fminimizer_size (sfm);
11003:
11004: if (rval == GSL_SUCCESS)
11005: printf ("converged to a local maximum at\n");
11006:
11007: printf("%5d ", iteri);
11008: for (it = 0; it < NDIM; it++){
11009: printf ("%10.5f ", gsl_vector_get (sfm->x, it));
11010: }
11011: printf("f() = %-10.5f ssize = %.7f\n", sfm->fval, ssval);
11012: }
11013:
11014: printf("\n\n Please note: Program should be run many times with varying starting points to detemine global maximum\n\n");
11015:
11016: gsl_vector_free(x); /* initial values */
11017: gsl_vector_free(ss); /* inital step size */
11018: for (it=0; it<NDIM; it++){
11019: p[it+1]=gsl_vector_get(sfm->x,it);
11020: fprintf(ficrespow," %.12lf", p[it]);
11021: }
11022: gsl_multimin_fminimizer_free (sfm); /* p *(sfm.x.data) et p *(sfm.x.data+1) */
11023: #endif
11024: #ifdef POWELL
11025: powell(p,ximort,NDIM,ftol,&iter,&fret,gompertz);
11026: #endif
1.126 brouard 11027: fclose(ficrespow);
11028:
1.203 brouard 11029: hesscov(matcov, hess, p, NDIM, delti, 1e-4, gompertz);
1.126 brouard 11030:
11031: for(i=1; i <=NDIM; i++)
11032: for(j=i+1;j<=NDIM;j++)
1.220 brouard 11033: matcov[i][j]=matcov[j][i];
1.126 brouard 11034:
11035: printf("\nCovariance matrix\n ");
1.203 brouard 11036: fprintf(ficlog,"\nCovariance matrix\n ");
1.126 brouard 11037: for(i=1; i <=NDIM; i++) {
11038: for(j=1;j<=NDIM;j++){
1.220 brouard 11039: printf("%f ",matcov[i][j]);
11040: fprintf(ficlog,"%f ",matcov[i][j]);
1.126 brouard 11041: }
1.203 brouard 11042: printf("\n "); fprintf(ficlog,"\n ");
1.126 brouard 11043: }
11044:
11045: printf("iter=%d MLE=%f Eq=%lf*exp(%lf*(age-%d))\n",iter,-gompertz(p),p[1],p[2],agegomp);
1.193 brouard 11046: for (i=1;i<=NDIM;i++) {
1.126 brouard 11047: printf("%f [%f ; %f]\n",p[i],p[i]-2*sqrt(matcov[i][i]),p[i]+2*sqrt(matcov[i][i]));
1.193 brouard 11048: fprintf(ficlog,"%f [%f ; %f]\n",p[i],p[i]-2*sqrt(matcov[i][i]),p[i]+2*sqrt(matcov[i][i]));
11049: }
1.126 brouard 11050: lsurv=vector(1,AGESUP);
11051: lpop=vector(1,AGESUP);
11052: tpop=vector(1,AGESUP);
11053: lsurv[agegomp]=100000;
11054:
11055: for (k=agegomp;k<=AGESUP;k++) {
11056: agemortsup=k;
11057: if (p[1]*exp(p[2]*(k-agegomp))>1) break;
11058: }
11059:
11060: for (k=agegomp;k<agemortsup;k++)
11061: lsurv[k+1]=lsurv[k]-lsurv[k]*(p[1]*exp(p[2]*(k-agegomp)));
11062:
11063: for (k=agegomp;k<agemortsup;k++){
11064: lpop[k]=(lsurv[k]+lsurv[k+1])/2.;
11065: sumlpop=sumlpop+lpop[k];
11066: }
11067:
11068: tpop[agegomp]=sumlpop;
11069: for (k=agegomp;k<(agemortsup-3);k++){
11070: /* tpop[k+1]=2;*/
11071: tpop[k+1]=tpop[k]-lpop[k];
11072: }
11073:
11074:
11075: printf("\nAge lx qx dx Lx Tx e(x)\n");
11076: for (k=agegomp;k<(agemortsup-2);k++)
11077: printf("%d %.0lf %lf %.0lf %.0lf %.0lf %lf\n",k,lsurv[k],p[1]*exp(p[2]*(k-agegomp)),(p[1]*exp(p[2]*(k-agegomp)))*lsurv[k],lpop[k],tpop[k],tpop[k]/lsurv[k]);
11078:
11079:
11080: replace_back_to_slash(pathc,pathcd); /* Even gnuplot wants a / */
1.220 brouard 11081: ageminpar=50;
11082: agemaxpar=100;
1.194 brouard 11083: if(ageminpar == AGEOVERFLOW ||agemaxpar == AGEOVERFLOW){
11084: printf("Warning! Error in gnuplot file with ageminpar %f or agemaxpar %f overflow\n\
11085: This is probably because your parameter file doesn't \n contain the exact number of lines (or columns) corresponding to your model line.\n\
11086: Please run with mle=-1 to get a correct covariance matrix.\n",ageminpar,agemaxpar);
11087: fprintf(ficlog,"Warning! Error in gnuplot file with ageminpar %f or agemaxpar %f overflow\n\
11088: This is probably because your parameter file doesn't \n contain the exact number of lines (or columns) corresponding to your model line.\n\
11089: Please run with mle=-1 to get a correct covariance matrix.\n",ageminpar,agemaxpar);
1.220 brouard 11090: }else{
11091: printf("Warning! ageminpar %f and agemaxpar %f have been fixed because for simplification until it is fixed...\n\n",ageminpar,agemaxpar);
11092: fprintf(ficlog,"Warning! ageminpar %f and agemaxpar %f have been fixed because for simplification until it is fixed...\n\n",ageminpar,agemaxpar);
1.201 brouard 11093: printinggnuplotmort(fileresu, optionfilefiname,ageminpar,agemaxpar,fage, pathc,p);
1.220 brouard 11094: }
1.201 brouard 11095: printinghtmlmort(fileresu,title,datafile, firstpass, lastpass, \
1.126 brouard 11096: stepm, weightopt,\
11097: model,imx,p,matcov,agemortsup);
11098:
11099: free_vector(lsurv,1,AGESUP);
11100: free_vector(lpop,1,AGESUP);
11101: free_vector(tpop,1,AGESUP);
1.220 brouard 11102: free_matrix(ximort,1,NDIM,1,NDIM);
1.136 brouard 11103: free_ivector(cens,1,n);
11104: free_vector(agecens,1,n);
11105: free_ivector(dcwave,1,n);
1.220 brouard 11106: #ifdef GSL
1.136 brouard 11107: #endif
1.186 brouard 11108: } /* Endof if mle==-3 mortality only */
1.205 brouard 11109: /* Standard */
11110: else{ /* For mle !=- 3, could be 0 or 1 or 4 etc. */
11111: globpr=0;/* Computes sum of likelihood for globpr=1 and funcone */
11112: /* Computes likelihood for initial parameters, uses funcone to compute gpimx and gsw */
1.132 brouard 11113: likelione(ficres, p, npar, nlstate, &globpr, &ipmx, &sw, &fretone, funcone); /* Prints the contributions to the likelihood */
1.126 brouard 11114: printf("First Likeli=%12.6f ipmx=%ld sw=%12.6f",fretone,ipmx,sw);
11115: for (k=1; k<=npar;k++)
11116: printf(" %d %8.5f",k,p[k]);
11117: printf("\n");
1.205 brouard 11118: if(mle>=1){ /* Could be 1 or 2, Real Maximization */
11119: /* mlikeli uses func not funcone */
1.247 brouard 11120: /* for(i=1;i<nlstate;i++){ */
11121: /* /\*reducing xi for 1 to npar to 1 to ncovmodel; *\/ */
11122: /* mlikeli(ficres,p, ncovmodel, ncovmodel, nlstate, ftol, funcnoprod); */
11123: /* } */
1.205 brouard 11124: mlikeli(ficres,p, npar, ncovmodel, nlstate, ftol, func);
11125: }
11126: if(mle==0) {/* No optimization, will print the likelihoods for the datafile */
11127: globpr=0;/* Computes sum of likelihood for globpr=1 and funcone */
11128: /* Computes likelihood for initial parameters, uses funcone to compute gpimx and gsw */
11129: likelione(ficres, p, npar, nlstate, &globpr, &ipmx, &sw, &fretone, funcone); /* Prints the contributions to the likelihood */
11130: }
11131: globpr=1; /* again, to print the individual contributions using computed gpimx and gsw */
1.126 brouard 11132: likelione(ficres, p, npar, nlstate, &globpr, &ipmx, &sw, &fretone, funcone); /* Prints the contributions to the likelihood */
11133: printf("Second Likeli=%12.6f ipmx=%ld sw=%12.6f",fretone,ipmx,sw);
11134: for (k=1; k<=npar;k++)
11135: printf(" %d %8.5f",k,p[k]);
11136: printf("\n");
11137:
11138: /*--------- results files --------------*/
1.224 brouard 11139: fprintf(ficres,"title=%s datafile=%s lastobs=%d firstpass=%d lastpass=%d\nftol=%e stepm=%d ncovcol=%d nqv=%d ntv=%d nqtv=%d nlstate=%d ndeath=%d maxwav=%d mle= 0 weight=%d\nmodel=1+age+%s.\n", title, datafile, lastobs, firstpass,lastpass,ftol, stepm, ncovcol, nqv, ntv, nqtv, nlstate, ndeath, maxwav, weightopt,model);
1.126 brouard 11140:
11141:
11142: fprintf(ficres,"# Parameters nlstate*nlstate*ncov a12*1 + b12 * age + ...\n");
11143: printf("# Parameters nlstate*nlstate*ncov a12*1 + b12 * age + ...\n");
11144: fprintf(ficlog,"# Parameters nlstate*nlstate*ncov a12*1 + b12 * age + ...\n");
11145: for(i=1,jk=1; i <=nlstate; i++){
11146: for(k=1; k <=(nlstate+ndeath); k++){
1.225 brouard 11147: if (k != i) {
11148: printf("%d%d ",i,k);
11149: fprintf(ficlog,"%d%d ",i,k);
11150: fprintf(ficres,"%1d%1d ",i,k);
11151: for(j=1; j <=ncovmodel; j++){
11152: printf("%12.7f ",p[jk]);
11153: fprintf(ficlog,"%12.7f ",p[jk]);
11154: fprintf(ficres,"%12.7f ",p[jk]);
11155: jk++;
11156: }
11157: printf("\n");
11158: fprintf(ficlog,"\n");
11159: fprintf(ficres,"\n");
11160: }
1.126 brouard 11161: }
11162: }
1.203 brouard 11163: if(mle != 0){
11164: /* Computing hessian and covariance matrix only at a peak of the Likelihood, that is after optimization */
1.126 brouard 11165: ftolhess=ftol; /* Usually correct */
1.203 brouard 11166: hesscov(matcov, hess, p, npar, delti, ftolhess, func);
11167: printf("Parameters and 95%% confidence intervals\n W is simply the result of the division of the parameter by the square root of covariance of the parameter.\n And Wald-based confidence intervals plus and minus 1.96 * W .\n But be careful that parameters are highly correlated because incidence of disability is highly correlated to incidence of recovery.\n It might be better to visualize the covariance matrix. See the page 'Matrix of variance-covariance of one-step probabilities' and its graphs.\n");
11168: fprintf(ficlog, "Parameters, Wald tests and Wald-based confidence intervals\n W is simply the result of the division of the parameter by the square root of covariance of the parameter.\n And Wald-based confidence intervals plus and minus 1.96 * W \n It might be better to visualize the covariance matrix. See the page 'Matrix of variance-covariance of one-step probabilities' and its graphs.\n");
11169: for(i=1,jk=1; i <=nlstate; i++){
1.225 brouard 11170: for(k=1; k <=(nlstate+ndeath); k++){
11171: if (k != i) {
11172: printf("%d%d ",i,k);
11173: fprintf(ficlog,"%d%d ",i,k);
11174: for(j=1; j <=ncovmodel; j++){
11175: printf("%12.7f W=%8.3f CI=[%12.7f ; %12.7f] ",p[jk], p[jk]/sqrt(matcov[jk][jk]), p[jk]-1.96*sqrt(matcov[jk][jk]),p[jk]+1.96*sqrt(matcov[jk][jk]));
11176: fprintf(ficlog,"%12.7f W=%8.3f CI=[%12.7f ; %12.7f] ",p[jk], p[jk]/sqrt(matcov[jk][jk]), p[jk]-1.96*sqrt(matcov[jk][jk]),p[jk]+1.96*sqrt(matcov[jk][jk]));
11177: jk++;
11178: }
11179: printf("\n");
11180: fprintf(ficlog,"\n");
11181: }
11182: }
1.193 brouard 11183: }
1.203 brouard 11184: } /* end of hesscov and Wald tests */
1.225 brouard 11185:
1.203 brouard 11186: /* */
1.126 brouard 11187: fprintf(ficres,"# Scales (for hessian or gradient estimation)\n");
11188: printf("# Scales (for hessian or gradient estimation)\n");
11189: fprintf(ficlog,"# Scales (for hessian or gradient estimation)\n");
11190: for(i=1,jk=1; i <=nlstate; i++){
11191: for(j=1; j <=nlstate+ndeath; j++){
1.225 brouard 11192: if (j!=i) {
11193: fprintf(ficres,"%1d%1d",i,j);
11194: printf("%1d%1d",i,j);
11195: fprintf(ficlog,"%1d%1d",i,j);
11196: for(k=1; k<=ncovmodel;k++){
11197: printf(" %.5e",delti[jk]);
11198: fprintf(ficlog," %.5e",delti[jk]);
11199: fprintf(ficres," %.5e",delti[jk]);
11200: jk++;
11201: }
11202: printf("\n");
11203: fprintf(ficlog,"\n");
11204: fprintf(ficres,"\n");
11205: }
1.126 brouard 11206: }
11207: }
11208:
11209: fprintf(ficres,"# Covariance matrix \n# 121 Var(a12)\n# 122 Cov(b12,a12) Var(b12)\n# ...\n# 232 Cov(b23,a12) Cov(b23,b12) ... Var (b23)\n");
1.203 brouard 11210: if(mle >= 1) /* To big for the screen */
1.126 brouard 11211: printf("# Covariance matrix \n# 121 Var(a12)\n# 122 Cov(b12,a12) Var(b12)\n# ...\n# 232 Cov(b23,a12) Cov(b23,b12) ... Var (b23)\n");
11212: fprintf(ficlog,"# Covariance matrix \n# 121 Var(a12)\n# 122 Cov(b12,a12) Var(b12)\n# ...\n# 232 Cov(b23,a12) Cov(b23,b12) ... Var (b23)\n");
11213: /* # 121 Var(a12)\n\ */
11214: /* # 122 Cov(b12,a12) Var(b12)\n\ */
11215: /* # 131 Cov(a13,a12) Cov(a13,b12, Var(a13)\n\ */
11216: /* # 132 Cov(b13,a12) Cov(b13,b12, Cov(b13,a13) Var(b13)\n\ */
11217: /* # 212 Cov(a21,a12) Cov(a21,b12, Cov(a21,a13) Cov(a21,b13) Var(a21)\n\ */
11218: /* # 212 Cov(b21,a12) Cov(b21,b12, Cov(b21,a13) Cov(b21,b13) Cov(b21,a21) Var(b21)\n\ */
11219: /* # 232 Cov(a23,a12) Cov(a23,b12, Cov(a23,a13) Cov(a23,b13) Cov(a23,a21) Cov(a23,b21) Var(a23)\n\ */
11220: /* # 232 Cov(b23,a12) Cov(b23,b12) ... Var (b23)\n" */
11221:
11222:
11223: /* Just to have a covariance matrix which will be more understandable
11224: even is we still don't want to manage dictionary of variables
11225: */
11226: for(itimes=1;itimes<=2;itimes++){
11227: jj=0;
11228: for(i=1; i <=nlstate; i++){
1.225 brouard 11229: for(j=1; j <=nlstate+ndeath; j++){
11230: if(j==i) continue;
11231: for(k=1; k<=ncovmodel;k++){
11232: jj++;
11233: ca[0]= k+'a'-1;ca[1]='\0';
11234: if(itimes==1){
11235: if(mle>=1)
11236: printf("#%1d%1d%d",i,j,k);
11237: fprintf(ficlog,"#%1d%1d%d",i,j,k);
11238: fprintf(ficres,"#%1d%1d%d",i,j,k);
11239: }else{
11240: if(mle>=1)
11241: printf("%1d%1d%d",i,j,k);
11242: fprintf(ficlog,"%1d%1d%d",i,j,k);
11243: fprintf(ficres,"%1d%1d%d",i,j,k);
11244: }
11245: ll=0;
11246: for(li=1;li <=nlstate; li++){
11247: for(lj=1;lj <=nlstate+ndeath; lj++){
11248: if(lj==li) continue;
11249: for(lk=1;lk<=ncovmodel;lk++){
11250: ll++;
11251: if(ll<=jj){
11252: cb[0]= lk +'a'-1;cb[1]='\0';
11253: if(ll<jj){
11254: if(itimes==1){
11255: if(mle>=1)
11256: printf(" Cov(%s%1d%1d,%s%1d%1d)",ca,i,j,cb, li,lj);
11257: fprintf(ficlog," Cov(%s%1d%1d,%s%1d%1d)",ca,i,j,cb, li,lj);
11258: fprintf(ficres," Cov(%s%1d%1d,%s%1d%1d)",ca,i,j,cb, li,lj);
11259: }else{
11260: if(mle>=1)
11261: printf(" %.5e",matcov[jj][ll]);
11262: fprintf(ficlog," %.5e",matcov[jj][ll]);
11263: fprintf(ficres," %.5e",matcov[jj][ll]);
11264: }
11265: }else{
11266: if(itimes==1){
11267: if(mle>=1)
11268: printf(" Var(%s%1d%1d)",ca,i,j);
11269: fprintf(ficlog," Var(%s%1d%1d)",ca,i,j);
11270: fprintf(ficres," Var(%s%1d%1d)",ca,i,j);
11271: }else{
11272: if(mle>=1)
11273: printf(" %.7e",matcov[jj][ll]);
11274: fprintf(ficlog," %.7e",matcov[jj][ll]);
11275: fprintf(ficres," %.7e",matcov[jj][ll]);
11276: }
11277: }
11278: }
11279: } /* end lk */
11280: } /* end lj */
11281: } /* end li */
11282: if(mle>=1)
11283: printf("\n");
11284: fprintf(ficlog,"\n");
11285: fprintf(ficres,"\n");
11286: numlinepar++;
11287: } /* end k*/
11288: } /*end j */
1.126 brouard 11289: } /* end i */
11290: } /* end itimes */
11291:
11292: fflush(ficlog);
11293: fflush(ficres);
1.225 brouard 11294: while(fgets(line, MAXLINE, ficpar)) {
11295: /* If line starts with a # it is a comment */
11296: if (line[0] == '#') {
11297: numlinepar++;
11298: fputs(line,stdout);
11299: fputs(line,ficparo);
11300: fputs(line,ficlog);
11301: continue;
11302: }else
11303: break;
11304: }
11305:
1.209 brouard 11306: /* while((c=getc(ficpar))=='#' && c!= EOF){ */
11307: /* ungetc(c,ficpar); */
11308: /* fgets(line, MAXLINE, ficpar); */
11309: /* fputs(line,stdout); */
11310: /* fputs(line,ficparo); */
11311: /* } */
11312: /* ungetc(c,ficpar); */
1.126 brouard 11313:
11314: estepm=0;
1.209 brouard 11315: if((num_filled=sscanf(line,"agemin=%lf agemax=%lf bage=%lf fage=%lf estepm=%d ftolpl=%lf\n",&ageminpar,&agemaxpar, &bage, &fage, &estepm, &ftolpl)) !=EOF){
1.225 brouard 11316:
11317: if (num_filled != 6) {
11318: printf("Error: Not 6 parameters in line, for example:agemin=60 agemax=95 bage=55 fage=95 estepm=24 ftolpl=6e-4\n, your line=%s . Probably you are running an older format.\n",line);
11319: fprintf(ficlog,"Error: Not 6 parameters in line, for example:agemin=60 agemax=95 bage=55 fage=95 estepm=24 ftolpl=6e-4\n, your line=%s . Probably you are running an older format.\n",line);
11320: goto end;
11321: }
11322: printf("agemin=%lf agemax=%lf bage=%lf fage=%lf estepm=%d ftolpl=%lf\n",ageminpar,agemaxpar, bage, fage, estepm, ftolpl);
11323: }
11324: /* ftolpl=6*ftol*1.e5; /\* 6.e-3 make convergences in less than 80 loops for the prevalence limit *\/ */
11325: /*ftolpl=6.e-4;*/ /* 6.e-3 make convergences in less than 80 loops for the prevalence limit */
11326:
1.209 brouard 11327: /* fscanf(ficpar,"agemin=%lf agemax=%lf bage=%lf fage=%lf estepm=%d ftolpl=%\n",&ageminpar,&agemaxpar, &bage, &fage, &estepm); */
1.126 brouard 11328: if (estepm==0 || estepm < stepm) estepm=stepm;
11329: if (fage <= 2) {
11330: bage = ageminpar;
11331: fage = agemaxpar;
11332: }
11333:
11334: fprintf(ficres,"# agemin agemax for life expectancy, bage fage (if mle==0 ie no data nor Max likelihood).\n");
1.211 brouard 11335: fprintf(ficres,"agemin=%.0f agemax=%.0f bage=%.0f fage=%.0f estepm=%d ftolpl=%e\n",ageminpar,agemaxpar,bage,fage, estepm, ftolpl);
11336: fprintf(ficparo,"agemin=%.0f agemax=%.0f bage=%.0f fage=%.0f estepm=%d, ftolpl=%e\n",ageminpar,agemaxpar,bage,fage, estepm, ftolpl);
1.220 brouard 11337:
1.186 brouard 11338: /* Other stuffs, more or less useful */
1.254 brouard 11339: while(fgets(line, MAXLINE, ficpar)) {
11340: /* If line starts with a # it is a comment */
11341: if (line[0] == '#') {
11342: numlinepar++;
11343: fputs(line,stdout);
11344: fputs(line,ficparo);
11345: fputs(line,ficlog);
11346: continue;
11347: }else
11348: break;
11349: }
11350:
11351: if((num_filled=sscanf(line,"begin-prev-date=%lf/%lf/%lf end-prev-date=%lf/%lf/%lf mov_average=%d\n",&jprev1, &mprev1,&anprev1,&jprev2, &mprev2,&anprev2,&mobilav)) !=EOF){
11352:
11353: if (num_filled != 7) {
11354: printf("Error: Not 7 (data)parameters in line but %d, for example:begin-prev-date=1/1/1990 end-prev-date=1/6/2004 mov_average=0\n, your line=%s . Probably you are running an older format.\n",num_filled,line);
11355: fprintf(ficlog,"Error: Not 7 (data)parameters in line but %d, for example:begin-prev-date=1/1/1990 end-prev-date=1/6/2004 mov_average=0\n, your line=%s . Probably you are running an older format.\n",num_filled,line);
11356: goto end;
11357: }
11358: printf("begin-prev-date=%.lf/%.lf/%.lf end-prev-date=%.lf/%.lf/%.lf mov_average=%d\n",jprev1, mprev1,anprev1,jprev2, mprev2,anprev2,mobilav);
11359: fprintf(ficparo,"begin-prev-date=%.lf/%.lf/%.lf end-prev-date=%.lf/%.lf/%.lf mov_average=%d\n",jprev1, mprev1,anprev1,jprev2, mprev2,anprev2,mobilav);
11360: fprintf(ficres,"begin-prev-date=%.lf/%.lf/%.lf end-prev-date=%.lf/%.lf/%.lf mov_average=%d\n",jprev1, mprev1,anprev1,jprev2, mprev2,anprev2,mobilav);
11361: fprintf(ficlog,"begin-prev-date=%.lf/%.lf/%.lf end-prev-date=%.lf/%.lf/%.lf mov_average=%d\n",jprev1, mprev1,anprev1,jprev2, mprev2,anprev2,mobilav);
1.126 brouard 11362: }
1.254 brouard 11363:
11364: while(fgets(line, MAXLINE, ficpar)) {
11365: /* If line starts with a # it is a comment */
11366: if (line[0] == '#') {
11367: numlinepar++;
11368: fputs(line,stdout);
11369: fputs(line,ficparo);
11370: fputs(line,ficlog);
11371: continue;
11372: }else
11373: break;
1.126 brouard 11374: }
11375:
11376:
11377: dateprev1=anprev1+(mprev1-1)/12.+(jprev1-1)/365.;
11378: dateprev2=anprev2+(mprev2-1)/12.+(jprev2-1)/365.;
11379:
1.254 brouard 11380: if((num_filled=sscanf(line,"pop_based=%d\n",&popbased)) !=EOF){
11381: if (num_filled != 1) {
11382: printf("Error: Not 1 (data)parameters in line but %d, for example:pop_based=0\n, your line=%s . Probably you are running an older format.\n",num_filled,line);
11383: fprintf(ficlog,"Error: Not 1 (data)parameters in line but %d, for example: pop_based=1\n, your line=%s . Probably you are running an older format.\n",num_filled,line);
11384: goto end;
11385: }
11386: printf("pop_based=%d\n",popbased);
11387: fprintf(ficlog,"pop_based=%d\n",popbased);
11388: fprintf(ficparo,"pop_based=%d\n",popbased);
11389: fprintf(ficres,"pop_based=%d\n",popbased);
11390: }
11391:
1.258 brouard 11392: /* Results */
11393: nresult=0;
11394: do{
11395: if(!fgets(line, MAXLINE, ficpar)){
11396: endishere=1;
11397: parameterline=14;
11398: }else if (line[0] == '#') {
11399: /* If line starts with a # it is a comment */
1.254 brouard 11400: numlinepar++;
11401: fputs(line,stdout);
11402: fputs(line,ficparo);
11403: fputs(line,ficlog);
11404: continue;
1.258 brouard 11405: }else if(sscanf(line,"prevforecast=%[^\n]\n",modeltemp))
11406: parameterline=11;
11407: else if(sscanf(line,"backcast=%[^\n]\n",modeltemp))
11408: parameterline=12;
11409: else if(sscanf(line,"result:%[^\n]\n",modeltemp))
11410: parameterline=13;
11411: else{
11412: parameterline=14;
1.254 brouard 11413: }
1.258 brouard 11414: switch (parameterline){
11415: case 11:
11416: if((num_filled=sscanf(line,"prevforecast=%d starting-proj-date=%lf/%lf/%lf final-proj-date=%lf/%lf/%lf mobil_average=%d\n",&prevfcast,&jproj1,&mproj1,&anproj1,&jproj2,&mproj2,&anproj2,&mobilavproj)) !=EOF){
11417: if (num_filled != 8) {
11418: printf("Error: Not 8 (data)parameters in line but %d, for example:prevforecast=1 starting-proj-date=1/1/1990 final-proj-date=1/1/2000 mobil_average=0\n, your line=%s . Probably you are running an older format.\n",num_filled,line);
11419: fprintf(ficlog,"Error: Not 8 (data)parameters in line but %d, for example:prevforecast=1 starting-proj-date=1/1/1990 final-proj-date=1/1/2000 mov_average=0\n, your line=%s . Probably you are running an older format.\n",num_filled,line);
11420: goto end;
11421: }
11422: fprintf(ficparo,"prevforecast=%d starting-proj-date=%.lf/%.lf/%.lf final-proj-date=%.lf/%.lf/%.lf mobil_average=%d\n",prevfcast,jproj1,mproj1,anproj1,jproj2,mproj2,anproj2,mobilavproj);
11423: printf("prevforecast=%d starting-proj-date=%.lf/%.lf/%.lf final-proj-date=%.lf/%.lf/%.lf mobil_average=%d\n",prevfcast,jproj1,mproj1,anproj1,jproj2,mproj2,anproj2,mobilavproj);
11424: fprintf(ficlog,"prevforecast=%d starting-proj-date=%.lf/%.lf/%.lf final-proj-date=%.lf/%.lf/%.lf mobil_average=%d\n",prevfcast,jproj1,mproj1,anproj1,jproj2,mproj2,anproj2,mobilavproj);
11425: fprintf(ficres,"prevforecast=%d starting-proj-date=%.lf/%.lf/%.lf final-proj-date=%.lf/%.lf/%.lf mobil_average=%d\n",prevfcast,jproj1,mproj1,anproj1,jproj2,mproj2,anproj2,mobilavproj);
11426: /* day and month of proj2 are not used but only year anproj2.*/
11427: }
1.254 brouard 11428: break;
1.258 brouard 11429: case 12:
11430: /*fscanf(ficpar,"backcast=%d starting-back-date=%lf/%lf/%lf final-back-date=%lf/%lf/%lf mobil_average=%d\n",&backcast,&jback1,&mback1,&anback1,&jback2,&mback2,&anback2,&mobilavproj);*/
11431: if((num_filled=sscanf(line,"backcast=%d starting-back-date=%lf/%lf/%lf final-back-date=%lf/%lf/%lf mobil_average=%d\n",&backcast,&jback1,&mback1,&anback1,&jback2,&mback2,&anback2,&mobilavproj)) !=EOF){
11432: if (num_filled != 8) {
1.262 brouard 11433: printf("Error: Not 8 (data)parameters in line but %d, for example:backcast=1 starting-back-date=1/1/1990 final-back-date=1/1/1970 mobil_average=1\n, your line=%s . Probably you are running an older format.\n",num_filled,line);
11434: fprintf(ficlog,"Error: Not 8 (data)parameters in line but %d, for example:backcast=1 starting-back-date=1/1/1990 final-back-date=1/1/1970 mobil_average=1\n, your line=%s . Probably you are running an older format.\n",num_filled,line);
1.258 brouard 11435: goto end;
11436: }
11437: printf("backcast=%d starting-back-date=%.lf/%.lf/%.lf final-back-date=%.lf/%.lf/%.lf mobil_average=%d\n",backcast,jback1,mback1,anback1,jback2,mback2,anback2,mobilavproj);
11438: fprintf(ficparo,"backcast=%d starting-back-date=%.lf/%.lf/%.lf final-back-date=%.lf/%.lf/%.lf mobil_average=%d\n",backcast,jback1,mback1,anback1,jback2,mback2,anback2,mobilavproj);
11439: fprintf(ficlog,"backcast=%d starting-back-date=%.lf/%.lf/%.lf final-back-date=%.lf/%.lf/%.lf mobil_average=%d\n",backcast,jback1,mback1,anback1,jback2,mback2,anback2,mobilavproj);
11440: fprintf(ficres,"backcast=%d starting-back-date=%.lf/%.lf/%.lf final-back-date=%.lf/%.lf/%.lf mobil_average=%d\n",backcast,jback1,mback1,anback1,jback2,mback2,anback2,mobilavproj);
11441: /* day and month of proj2 are not used but only year anproj2.*/
11442: }
1.230 brouard 11443: break;
1.258 brouard 11444: case 13:
11445: if((num_filled=sscanf(line,"result:%[^\n]\n",resultline)) !=EOF){
11446: if (num_filled == 0){
11447: resultline[0]='\0';
11448: printf("Warning %d: no result line! It should be at minimum 'result: V2=0 V1=1 or result:.\n%s\n", num_filled, line);
11449: fprintf(ficlog,"Warning %d: no result line! It should be at minimum 'result: V2=0 V1=1 or result:.\n%s\n", num_filled, line);
11450: break;
11451: } else if (num_filled != 1){
11452: printf("ERROR %d: result line! It should be at minimum 'result: V2=0 V1=1 or result:.' %s\n",num_filled, line);
11453: fprintf(ficlog,"ERROR %d: result line! It should be at minimum 'result: V2=0 V1=1 or result:.' %s\n",num_filled, line);
11454: }
11455: nresult++; /* Sum of resultlines */
11456: printf("Result %d: result=%s\n",nresult, resultline);
11457: if(nresult > MAXRESULTLINES){
11458: printf("ERROR: Current version of IMaCh limits the number of resultlines to %d, you used %d\n",MAXRESULTLINES,nresult);
11459: fprintf(ficlog,"ERROR: Current version of IMaCh limits the number of resultlines to %d, you used %d\n",MAXRESULTLINES,nresult);
11460: goto end;
11461: }
11462: decoderesult(resultline, nresult); /* Fills TKresult[nresult] combination and Tresult[nresult][k4+1] combination values */
11463: fprintf(ficparo,"result: %s\n",resultline);
11464: fprintf(ficres,"result: %s\n",resultline);
11465: fprintf(ficlog,"result: %s\n",resultline);
1.230 brouard 11466: break;
1.258 brouard 11467: case 14:
1.259 brouard 11468: if(ncovmodel >2 && nresult==0 ){
11469: printf("ERROR: no result lines! It should be at minimum 'result: V2=0 V1=1 or result:.' %s\n",line);
1.258 brouard 11470: goto end;
11471: }
1.259 brouard 11472: break;
1.258 brouard 11473: default:
11474: nresult=1;
11475: decoderesult(".",nresult ); /* No covariate */
11476: }
11477: } /* End switch parameterline */
11478: }while(endishere==0); /* End do */
1.126 brouard 11479:
1.230 brouard 11480: /* freqsummary(fileres, agemin, agemax, s, agev, nlstate, imx,Tvaraff,nbcode, ncodemax,mint,anint); */
1.145 brouard 11481: /* ,dateprev1,dateprev2,jprev1, mprev1,anprev1,jprev2, mprev2,anprev2); */
1.126 brouard 11482:
11483: replace_back_to_slash(pathc,pathcd); /* Even gnuplot wants a / */
1.194 brouard 11484: if(ageminpar == AGEOVERFLOW ||agemaxpar == -AGEOVERFLOW){
1.230 brouard 11485: printf("Warning! Error in gnuplot file with ageminpar %f or agemaxpar %f overflow\n\
1.194 brouard 11486: This is probably because your parameter file doesn't \n contain the exact number of lines (or columns) corresponding to your model line.\n\
11487: Please run with mle=-1 to get a correct covariance matrix.\n",ageminpar,agemaxpar);
1.230 brouard 11488: fprintf(ficlog,"Warning! Error in gnuplot file with ageminpar %f or agemaxpar %f overflow\n\
1.194 brouard 11489: This is probably because your parameter file doesn't \n contain the exact number of lines (or columns) corresponding to your model line.\n\
11490: Please run with mle=-1 to get a correct covariance matrix.\n",ageminpar,agemaxpar);
1.220 brouard 11491: }else{
1.218 brouard 11492: printinggnuplot(fileresu, optionfilefiname,ageminpar,agemaxpar,fage, prevfcast, backcast, pathc,p);
1.220 brouard 11493: }
11494: printinghtml(fileresu,title,datafile, firstpass, lastpass, stepm, weightopt, \
1.258 brouard 11495: model,imx,jmin,jmax,jmean,rfileres,popforecast,mobilav,prevfcast,mobilavproj,backcast, estepm, \
1.225 brouard 11496: jprev1,mprev1,anprev1,dateprev1,jprev2,mprev2,anprev2,dateprev2);
1.220 brouard 11497:
1.225 brouard 11498: /*------------ free_vector -------------*/
11499: /* chdir(path); */
1.220 brouard 11500:
1.215 brouard 11501: /* free_ivector(wav,1,imx); */ /* Moved after last prevalence call */
11502: /* free_imatrix(dh,1,lastpass-firstpass+2,1,imx); */
11503: /* free_imatrix(bh,1,lastpass-firstpass+2,1,imx); */
11504: /* free_imatrix(mw,1,lastpass-firstpass+2,1,imx); */
1.126 brouard 11505: free_lvector(num,1,n);
11506: free_vector(agedc,1,n);
11507: /*free_matrix(covar,0,NCOVMAX,1,n);*/
11508: /*free_matrix(covar,1,NCOVMAX,1,n);*/
11509: fclose(ficparo);
11510: fclose(ficres);
1.220 brouard 11511:
11512:
1.186 brouard 11513: /* Other results (useful)*/
1.220 brouard 11514:
11515:
1.126 brouard 11516: /*--------------- Prevalence limit (period or stable prevalence) --------------*/
1.180 brouard 11517: /*#include "prevlim.h"*/ /* Use ficrespl, ficlog */
11518: prlim=matrix(1,nlstate,1,nlstate);
1.209 brouard 11519: prevalence_limit(p, prlim, ageminpar, agemaxpar, ftolpl, &ncvyear);
1.126 brouard 11520: fclose(ficrespl);
11521:
11522: /*------------- h Pij x at various ages ------------*/
1.180 brouard 11523: /*#include "hpijx.h"*/
11524: hPijx(p, bage, fage);
1.145 brouard 11525: fclose(ficrespij);
1.227 brouard 11526:
1.220 brouard 11527: /* ncovcombmax= pow(2,cptcoveff); */
1.219 brouard 11528: /*-------------- Variance of one-step probabilities---*/
1.145 brouard 11529: k=1;
1.126 brouard 11530: varprob(optionfilefiname, matcov, p, delti, nlstate, bage, fage,k,Tvar,nbcode, ncodemax,strstart);
1.227 brouard 11531:
1.219 brouard 11532: /* Prevalence for each covariates in probs[age][status][cov] */
1.218 brouard 11533: probs= ma3x(1,AGESUP,1,nlstate+ndeath, 1,ncovcombmax);
1.126 brouard 11534: for(i=1;i<=AGESUP;i++)
1.219 brouard 11535: for(j=1;j<=nlstate+ndeath;j++) /* ndeath is useless but a necessity to be compared with mobaverages */
1.225 brouard 11536: for(k=1;k<=ncovcombmax;k++)
11537: probs[i][j][k]=0.;
1.219 brouard 11538: prevalence(probs, ageminpar, agemaxpar, s, agev, nlstate, imx, Tvar, nbcode, ncodemax, mint, anint, dateprev1, dateprev2, firstpass, lastpass);
11539: if (mobilav!=0 ||mobilavproj !=0 ) {
11540: mobaverages= ma3x(1, AGESUP,1,nlstate+ndeath, 1,ncovcombmax);
1.227 brouard 11541: for(i=1;i<=AGESUP;i++)
11542: for(j=1;j<=nlstate;j++)
11543: for(k=1;k<=ncovcombmax;k++)
11544: mobaverages[i][j][k]=0.;
1.219 brouard 11545: mobaverage=mobaverages;
11546: if (mobilav!=0) {
1.235 brouard 11547: printf("Movingaveraging observed prevalence\n");
1.258 brouard 11548: fprintf(ficlog,"Movingaveraging observed prevalence\n");
1.227 brouard 11549: if (movingaverage(probs, ageminpar, agemaxpar, mobaverage, mobilav)!=0){
11550: fprintf(ficlog," Error in movingaverage mobilav=%d\n",mobilav);
11551: printf(" Error in movingaverage mobilav=%d\n",mobilav);
11552: }
1.219 brouard 11553: }
11554: /* /\* Prevalence for each covariates in probs[age][status][cov] *\/ */
11555: /* prevalence(probs, ageminpar, agemaxpar, s, agev, nlstate, imx, Tvar, nbcode, ncodemax, mint, anint, dateprev1, dateprev2, firstpass, lastpass); */
11556: else if (mobilavproj !=0) {
1.235 brouard 11557: printf("Movingaveraging projected observed prevalence\n");
1.258 brouard 11558: fprintf(ficlog,"Movingaveraging projected observed prevalence\n");
1.227 brouard 11559: if (movingaverage(probs, ageminpar, agemaxpar, mobaverage, mobilavproj)!=0){
11560: fprintf(ficlog," Error in movingaverage mobilavproj=%d\n",mobilavproj);
11561: printf(" Error in movingaverage mobilavproj=%d\n",mobilavproj);
11562: }
1.219 brouard 11563: }
11564: }/* end if moving average */
1.227 brouard 11565:
1.126 brouard 11566: /*---------- Forecasting ------------------*/
11567: /*if((stepm == 1) && (strcmp(model,".")==0)){*/
11568: if(prevfcast==1){
11569: /* if(stepm ==1){*/
1.225 brouard 11570: prevforecast(fileresu, anproj1, mproj1, jproj1, agemin, agemax, dateprev1, dateprev2, mobilavproj, bage, fage, firstpass, lastpass, anproj2, p, cptcoveff);
1.126 brouard 11571: }
1.217 brouard 11572: if(backcast==1){
1.219 brouard 11573: ddnewms=matrix(1,nlstate+ndeath,1,nlstate+ndeath);
11574: ddoldms=matrix(1,nlstate+ndeath,1,nlstate+ndeath);
11575: ddsavms=matrix(1,nlstate+ndeath,1,nlstate+ndeath);
11576:
11577: /*--------------- Back Prevalence limit (period or stable prevalence) --------------*/
11578:
11579: bprlim=matrix(1,nlstate,1,nlstate);
11580: back_prevalence_limit(p, bprlim, ageminpar, agemaxpar, ftolpl, &ncvyear, dateprev1, dateprev2, firstpass, lastpass, mobilavproj);
11581: fclose(ficresplb);
11582:
1.222 brouard 11583: hBijx(p, bage, fage, mobaverage);
11584: fclose(ficrespijb);
1.219 brouard 11585: free_matrix(bprlim,1,nlstate,1,nlstate); /*here or after loop ? */
11586:
11587: /* prevbackforecast(fileresu, anback1, mback1, jback1, agemin, agemax, dateprev1, dateprev2, mobilavproj,
1.225 brouard 11588: bage, fage, firstpass, lastpass, anback2, p, cptcoveff); */
1.219 brouard 11589: free_matrix(ddnewms, 1, nlstate+ndeath, 1, nlstate+ndeath);
11590: free_matrix(ddsavms, 1, nlstate+ndeath, 1, nlstate+ndeath);
11591: free_matrix(ddoldms, 1, nlstate+ndeath, 1, nlstate+ndeath);
11592: }
1.217 brouard 11593:
1.186 brouard 11594:
11595: /* ------ Other prevalence ratios------------ */
1.126 brouard 11596:
1.215 brouard 11597: free_ivector(wav,1,imx);
11598: free_imatrix(dh,1,lastpass-firstpass+2,1,imx);
11599: free_imatrix(bh,1,lastpass-firstpass+2,1,imx);
11600: free_imatrix(mw,1,lastpass-firstpass+2,1,imx);
1.218 brouard 11601:
11602:
1.127 brouard 11603: /*---------- Health expectancies, no variances ------------*/
1.218 brouard 11604:
1.201 brouard 11605: strcpy(filerese,"E_");
11606: strcat(filerese,fileresu);
1.126 brouard 11607: if((ficreseij=fopen(filerese,"w"))==NULL) {
11608: printf("Problem with Health Exp. resultfile: %s\n", filerese); exit(0);
11609: fprintf(ficlog,"Problem with Health Exp. resultfile: %s\n", filerese); exit(0);
11610: }
1.208 brouard 11611: printf("Computing Health Expectancies: result on file '%s' ...", filerese);fflush(stdout);
11612: fprintf(ficlog,"Computing Health Expectancies: result on file '%s' ...", filerese);fflush(ficlog);
1.238 brouard 11613:
11614: pstamp(ficreseij);
1.219 brouard 11615:
1.235 brouard 11616: i1=pow(2,cptcoveff); /* Number of combination of dummy covariates */
11617: if (cptcovn < 1){i1=1;}
11618:
11619: for(nres=1; nres <= nresult; nres++) /* For each resultline */
11620: for(k=1; k<=i1;k++){ /* For any combination of dummy covariates, fixed and varying */
1.253 brouard 11621: if(i1 != 1 && TKresult[nres]!= k)
1.235 brouard 11622: continue;
1.219 brouard 11623: fprintf(ficreseij,"\n#****** ");
1.235 brouard 11624: printf("\n#****** ");
1.225 brouard 11625: for(j=1;j<=cptcoveff;j++) {
1.227 brouard 11626: fprintf(ficreseij,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
1.235 brouard 11627: printf("V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
11628: }
11629: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
11630: printf(" V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11631: fprintf(ficreseij," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
1.219 brouard 11632: }
11633: fprintf(ficreseij,"******\n");
1.235 brouard 11634: printf("******\n");
1.219 brouard 11635:
11636: eij=ma3x(1,nlstate,1,nlstate,(int) bage, (int) fage);
11637: oldm=oldms;savm=savms;
1.235 brouard 11638: evsij(eij, p, nlstate, stepm, (int) bage, (int)fage, oldm, savm, k, estepm, strstart, nres);
1.127 brouard 11639:
1.219 brouard 11640: free_ma3x(eij,1,nlstate,1,nlstate,(int) bage, (int)fage);
1.127 brouard 11641: }
11642: fclose(ficreseij);
1.208 brouard 11643: printf("done evsij\n");fflush(stdout);
11644: fprintf(ficlog,"done evsij\n");fflush(ficlog);
1.218 brouard 11645:
1.227 brouard 11646: /*---------- State-specific expectancies and variances ------------*/
1.218 brouard 11647:
11648:
1.201 brouard 11649: strcpy(filerest,"T_");
11650: strcat(filerest,fileresu);
1.127 brouard 11651: if((ficrest=fopen(filerest,"w"))==NULL) {
11652: printf("Problem with total LE resultfile: %s\n", filerest);goto end;
11653: fprintf(ficlog,"Problem with total LE resultfile: %s\n", filerest);goto end;
11654: }
1.208 brouard 11655: printf("Computing Total Life expectancies with their standard errors: file '%s' ...\n", filerest); fflush(stdout);
11656: fprintf(ficlog,"Computing Total Life expectancies with their standard errors: file '%s' ...\n", filerest); fflush(ficlog);
1.218 brouard 11657:
1.126 brouard 11658:
1.201 brouard 11659: strcpy(fileresstde,"STDE_");
11660: strcat(fileresstde,fileresu);
1.126 brouard 11661: if((ficresstdeij=fopen(fileresstde,"w"))==NULL) {
1.227 brouard 11662: printf("Problem with State specific Exp. and std errors resultfile: %s\n", fileresstde); exit(0);
11663: fprintf(ficlog,"Problem with State specific Exp. and std errors resultfile: %s\n", fileresstde); exit(0);
1.126 brouard 11664: }
1.227 brouard 11665: printf(" Computing State-specific Expectancies and standard errors: result on file '%s' \n", fileresstde);
11666: fprintf(ficlog," Computing State-specific Expectancies and standard errors: result on file '%s' \n", fileresstde);
1.126 brouard 11667:
1.201 brouard 11668: strcpy(filerescve,"CVE_");
11669: strcat(filerescve,fileresu);
1.126 brouard 11670: if((ficrescveij=fopen(filerescve,"w"))==NULL) {
1.227 brouard 11671: printf("Problem with Covar. State-specific Exp. resultfile: %s\n", filerescve); exit(0);
11672: fprintf(ficlog,"Problem with Covar. State-specific Exp. resultfile: %s\n", filerescve); exit(0);
1.126 brouard 11673: }
1.227 brouard 11674: printf(" Computing Covar. of State-specific Expectancies: result on file '%s' \n", filerescve);
11675: fprintf(ficlog," Computing Covar. of State-specific Expectancies: result on file '%s' \n", filerescve);
1.126 brouard 11676:
1.201 brouard 11677: strcpy(fileresv,"V_");
11678: strcat(fileresv,fileresu);
1.126 brouard 11679: if((ficresvij=fopen(fileresv,"w"))==NULL) {
11680: printf("Problem with variance resultfile: %s\n", fileresv);exit(0);
11681: fprintf(ficlog,"Problem with variance resultfile: %s\n", fileresv);exit(0);
11682: }
1.227 brouard 11683: printf(" Computing Variance-covariance of State-specific Expectancies: file '%s' ... ", fileresv);fflush(stdout);
11684: fprintf(ficlog," Computing Variance-covariance of State-specific Expectancies: file '%s' ... ", fileresv);fflush(ficlog);
1.126 brouard 11685:
1.145 brouard 11686: /*for(cptcov=1,k=0;cptcov<=i1;cptcov++){
11687: for(cptcod=1;cptcod<=ncodemax[cptcov];cptcod++){*/
11688:
1.235 brouard 11689: i1=pow(2,cptcoveff); /* Number of combination of dummy covariates */
11690: if (cptcovn < 1){i1=1;}
11691:
11692: for(nres=1; nres <= nresult; nres++) /* For each resultline */
11693: for(k=1; k<=i1;k++){ /* For any combination of dummy covariates, fixed and varying */
1.253 brouard 11694: if(i1 != 1 && TKresult[nres]!= k)
1.235 brouard 11695: continue;
1.242 brouard 11696: printf("\n#****** Result for:");
11697: fprintf(ficrest,"\n#****** Result for:");
11698: fprintf(ficlog,"\n#****** Result for:");
1.227 brouard 11699: for(j=1;j<=cptcoveff;j++){
11700: printf("V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
11701: fprintf(ficrest,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
11702: fprintf(ficlog,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
11703: }
1.235 brouard 11704: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
11705: printf(" V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11706: fprintf(ficrest," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11707: fprintf(ficlog," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11708: }
1.208 brouard 11709: fprintf(ficrest,"******\n");
1.227 brouard 11710: fprintf(ficlog,"******\n");
11711: printf("******\n");
1.208 brouard 11712:
11713: fprintf(ficresstdeij,"\n#****** ");
11714: fprintf(ficrescveij,"\n#****** ");
1.225 brouard 11715: for(j=1;j<=cptcoveff;j++) {
1.227 brouard 11716: fprintf(ficresstdeij,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
11717: fprintf(ficrescveij,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
1.208 brouard 11718: }
1.235 brouard 11719: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
11720: fprintf(ficresstdeij," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11721: fprintf(ficrescveij," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11722: }
1.208 brouard 11723: fprintf(ficresstdeij,"******\n");
11724: fprintf(ficrescveij,"******\n");
11725:
11726: fprintf(ficresvij,"\n#****** ");
1.238 brouard 11727: /* pstamp(ficresvij); */
1.225 brouard 11728: for(j=1;j<=cptcoveff;j++)
1.227 brouard 11729: fprintf(ficresvij,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
1.235 brouard 11730: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
11731: fprintf(ficresvij," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11732: }
1.208 brouard 11733: fprintf(ficresvij,"******\n");
11734:
11735: eij=ma3x(1,nlstate,1,nlstate,(int) bage, (int) fage);
11736: oldm=oldms;savm=savms;
1.235 brouard 11737: printf(" cvevsij ");
11738: fprintf(ficlog, " cvevsij ");
11739: cvevsij(eij, p, nlstate, stepm, (int) bage, (int)fage, oldm, savm, k, estepm, delti, matcov, strstart, nres);
1.208 brouard 11740: printf(" end cvevsij \n ");
11741: fprintf(ficlog, " end cvevsij \n ");
11742:
11743: /*
11744: */
11745: /* goto endfree; */
11746:
11747: vareij=ma3x(1,nlstate,1,nlstate,(int) bage, (int) fage);
11748: pstamp(ficrest);
11749:
11750:
11751: for(vpopbased=0; vpopbased <= popbased; vpopbased++){ /* Done for vpopbased=0 and vpopbased=1 if popbased==1*/
1.227 brouard 11752: oldm=oldms;savm=savms; /* ZZ Segmentation fault */
11753: cptcod= 0; /* To be deleted */
11754: printf("varevsij vpopbased=%d \n",vpopbased);
11755: fprintf(ficlog, "varevsij vpopbased=%d \n",vpopbased);
1.235 brouard 11756: varevsij(optionfilefiname, vareij, matcov, p, delti, nlstate, stepm, (int) bage, (int) fage, oldm, savm, prlim, ftolpl, &ncvyear, k, estepm, cptcov,cptcod,vpopbased,mobilav, strstart, nres); /* cptcod not initialized Intel */
1.227 brouard 11757: fprintf(ficrest,"# Total life expectancy with std error and decomposition into time to be expected in each health state\n# (weighted average of eij where weights are ");
11758: if(vpopbased==1)
11759: fprintf(ficrest,"the age specific prevalence observed (cross-sectionally) in the population i.e cross-sectionally\n in each health state (popbased=1) (mobilav=%d)\n",mobilav);
11760: else
11761: fprintf(ficrest,"the age specific period (stable) prevalences in each health state \n");
11762: fprintf(ficrest,"# Age popbased mobilav e.. (std) ");
11763: for (i=1;i<=nlstate;i++) fprintf(ficrest,"e.%d (std) ",i);
11764: fprintf(ficrest,"\n");
11765: /* printf("Which p?\n"); for(i=1;i<=npar;i++)printf("p[i=%d]=%lf,",i,p[i]);printf("\n"); */
11766: epj=vector(1,nlstate+1);
11767: printf("Computing age specific period (stable) prevalences in each health state \n");
11768: fprintf(ficlog,"Computing age specific period (stable) prevalences in each health state \n");
11769: for(age=bage; age <=fage ;age++){
1.235 brouard 11770: prevalim(prlim, nlstate, p, age, oldm, savm, ftolpl, &ncvyear, k, nres); /*ZZ Is it the correct prevalim */
1.227 brouard 11771: if (vpopbased==1) {
11772: if(mobilav ==0){
11773: for(i=1; i<=nlstate;i++)
11774: prlim[i][i]=probs[(int)age][i][k];
11775: }else{ /* mobilav */
11776: for(i=1; i<=nlstate;i++)
11777: prlim[i][i]=mobaverage[(int)age][i][k];
11778: }
11779: }
1.219 brouard 11780:
1.227 brouard 11781: fprintf(ficrest," %4.0f %d %d",age, vpopbased, mobilav);
11782: /* fprintf(ficrest," %4.0f %d %d %d %d",age, vpopbased, mobilav,Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]); */ /* to be done */
11783: /* printf(" age %4.0f ",age); */
11784: for(j=1, epj[nlstate+1]=0.;j <=nlstate;j++){
11785: for(i=1, epj[j]=0.;i <=nlstate;i++) {
11786: epj[j] += prlim[i][i]*eij[i][j][(int)age];
11787: /*ZZZ printf("%lf %lf ", prlim[i][i] ,eij[i][j][(int)age]);*/
11788: /* printf("%lf %lf ", prlim[i][i] ,eij[i][j][(int)age]); */
11789: }
11790: epj[nlstate+1] +=epj[j];
11791: }
11792: /* printf(" age %4.0f \n",age); */
1.219 brouard 11793:
1.227 brouard 11794: for(i=1, vepp=0.;i <=nlstate;i++)
11795: for(j=1;j <=nlstate;j++)
11796: vepp += vareij[i][j][(int)age];
11797: fprintf(ficrest," %7.3f (%7.3f)", epj[nlstate+1],sqrt(vepp));
11798: for(j=1;j <=nlstate;j++){
11799: fprintf(ficrest," %7.3f (%7.3f)", epj[j],sqrt(vareij[j][j][(int)age]));
11800: }
11801: fprintf(ficrest,"\n");
11802: }
1.208 brouard 11803: } /* End vpopbased */
11804: free_ma3x(eij,1,nlstate,1,nlstate,(int) bage, (int)fage);
11805: free_ma3x(vareij,1,nlstate,1,nlstate,(int) bage, (int)fage);
11806: free_vector(epj,1,nlstate+1);
1.235 brouard 11807: printf("done selection\n");fflush(stdout);
11808: fprintf(ficlog,"done selection\n");fflush(ficlog);
1.208 brouard 11809:
1.145 brouard 11810: /*}*/
1.235 brouard 11811: } /* End k selection */
1.227 brouard 11812:
11813: printf("done State-specific expectancies\n");fflush(stdout);
11814: fprintf(ficlog,"done State-specific expectancies\n");fflush(ficlog);
11815:
1.126 brouard 11816: /*------- Variance of period (stable) prevalence------*/
1.227 brouard 11817:
1.201 brouard 11818: strcpy(fileresvpl,"VPL_");
11819: strcat(fileresvpl,fileresu);
1.126 brouard 11820: if((ficresvpl=fopen(fileresvpl,"w"))==NULL) {
11821: printf("Problem with variance of period (stable) prevalence resultfile: %s\n", fileresvpl);
11822: exit(0);
11823: }
1.208 brouard 11824: printf("Computing Variance-covariance of period (stable) prevalence: file '%s' ...", fileresvpl);fflush(stdout);
11825: fprintf(ficlog, "Computing Variance-covariance of period (stable) prevalence: file '%s' ...", fileresvpl);fflush(ficlog);
1.227 brouard 11826:
1.145 brouard 11827: /*for(cptcov=1,k=0;cptcov<=i1;cptcov++){
11828: for(cptcod=1;cptcod<=ncodemax[cptcov];cptcod++){*/
1.227 brouard 11829:
1.235 brouard 11830: i1=pow(2,cptcoveff);
11831: if (cptcovn < 1){i1=1;}
11832:
11833: for(nres=1; nres <= nresult; nres++) /* For each resultline */
11834: for(k=1; k<=i1;k++){
1.253 brouard 11835: if(i1 != 1 && TKresult[nres]!= k)
1.235 brouard 11836: continue;
1.227 brouard 11837: fprintf(ficresvpl,"\n#****** ");
11838: printf("\n#****** ");
11839: fprintf(ficlog,"\n#****** ");
11840: for(j=1;j<=cptcoveff;j++) {
11841: fprintf(ficresvpl,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
11842: fprintf(ficlog,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
11843: printf("V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
11844: }
1.235 brouard 11845: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
11846: printf(" V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11847: fprintf(ficresvpl," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11848: fprintf(ficlog," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11849: }
1.227 brouard 11850: fprintf(ficresvpl,"******\n");
11851: printf("******\n");
11852: fprintf(ficlog,"******\n");
11853:
11854: varpl=matrix(1,nlstate,(int) bage, (int) fage);
11855: oldm=oldms;savm=savms;
1.235 brouard 11856: varprevlim(fileres, varpl, matcov, p, delti, nlstate, stepm, (int) bage, (int) fage, oldm, savm, prlim, ftolpl, &ncvyear, k, strstart, nres);
1.227 brouard 11857: free_matrix(varpl,1,nlstate,(int) bage, (int)fage);
1.145 brouard 11858: /*}*/
1.126 brouard 11859: }
1.227 brouard 11860:
1.126 brouard 11861: fclose(ficresvpl);
1.208 brouard 11862: printf("done variance-covariance of period prevalence\n");fflush(stdout);
11863: fprintf(ficlog,"done variance-covariance of period prevalence\n");fflush(ficlog);
1.227 brouard 11864:
11865: free_vector(weight,1,n);
11866: free_imatrix(Tvard,1,NCOVMAX,1,2);
11867: free_imatrix(s,1,maxwav+1,1,n);
11868: free_matrix(anint,1,maxwav,1,n);
11869: free_matrix(mint,1,maxwav,1,n);
11870: free_ivector(cod,1,n);
11871: free_ivector(tab,1,NCOVMAX);
11872: fclose(ficresstdeij);
11873: fclose(ficrescveij);
11874: fclose(ficresvij);
11875: fclose(ficrest);
11876: fclose(ficpar);
11877:
11878:
1.126 brouard 11879: /*---------- End : free ----------------*/
1.219 brouard 11880: if (mobilav!=0 ||mobilavproj !=0)
11881: free_ma3x(mobaverages,1, AGESUP,1,nlstate+ndeath, 1,ncovcombmax); /* We need to have a squared matrix with prevalence of the dead! */
1.218 brouard 11882: free_ma3x(probs,1,AGESUP,1,nlstate+ndeath, 1,ncovcombmax);
1.220 brouard 11883: free_matrix(prlim,1,nlstate,1,nlstate); /*here or after loop ? */
11884: free_matrix(pmmij,1,nlstate+ndeath,1,nlstate+ndeath);
1.126 brouard 11885: } /* mle==-3 arrives here for freeing */
1.227 brouard 11886: /* endfree:*/
11887: free_matrix(oldms, 1,nlstate+ndeath,1,nlstate+ndeath);
11888: free_matrix(newms, 1,nlstate+ndeath,1,nlstate+ndeath);
11889: free_matrix(savms, 1,nlstate+ndeath,1,nlstate+ndeath);
11890: free_ma3x(cotqvar,1,maxwav,1,nqtv,1,n);
1.233 brouard 11891: free_ma3x(cotvar,1,maxwav,1,ntv+nqtv,1,n);
1.227 brouard 11892: free_matrix(coqvar,1,maxwav,1,n);
11893: free_matrix(covar,0,NCOVMAX,1,n);
11894: free_matrix(matcov,1,npar,1,npar);
11895: free_matrix(hess,1,npar,1,npar);
11896: /*free_vector(delti,1,npar);*/
11897: free_ma3x(delti3,1,nlstate,1, nlstate+ndeath-1,1,ncovmodel);
11898: free_matrix(agev,1,maxwav,1,imx);
11899: free_ma3x(param,1,nlstate,1, nlstate+ndeath-1,1,ncovmodel);
11900:
11901: free_ivector(ncodemax,1,NCOVMAX);
11902: free_ivector(ncodemaxwundef,1,NCOVMAX);
11903: free_ivector(Dummy,-1,NCOVMAX);
11904: free_ivector(Fixed,-1,NCOVMAX);
1.238 brouard 11905: free_ivector(DummyV,1,NCOVMAX);
11906: free_ivector(FixedV,1,NCOVMAX);
1.227 brouard 11907: free_ivector(Typevar,-1,NCOVMAX);
11908: free_ivector(Tvar,1,NCOVMAX);
1.234 brouard 11909: free_ivector(TvarsQ,1,NCOVMAX);
11910: free_ivector(TvarsQind,1,NCOVMAX);
11911: free_ivector(TvarsD,1,NCOVMAX);
11912: free_ivector(TvarsDind,1,NCOVMAX);
1.231 brouard 11913: free_ivector(TvarFD,1,NCOVMAX);
11914: free_ivector(TvarFDind,1,NCOVMAX);
1.232 brouard 11915: free_ivector(TvarF,1,NCOVMAX);
11916: free_ivector(TvarFind,1,NCOVMAX);
11917: free_ivector(TvarV,1,NCOVMAX);
11918: free_ivector(TvarVind,1,NCOVMAX);
11919: free_ivector(TvarA,1,NCOVMAX);
11920: free_ivector(TvarAind,1,NCOVMAX);
1.231 brouard 11921: free_ivector(TvarFQ,1,NCOVMAX);
11922: free_ivector(TvarFQind,1,NCOVMAX);
11923: free_ivector(TvarVD,1,NCOVMAX);
11924: free_ivector(TvarVDind,1,NCOVMAX);
11925: free_ivector(TvarVQ,1,NCOVMAX);
11926: free_ivector(TvarVQind,1,NCOVMAX);
1.230 brouard 11927: free_ivector(Tvarsel,1,NCOVMAX);
11928: free_vector(Tvalsel,1,NCOVMAX);
1.227 brouard 11929: free_ivector(Tposprod,1,NCOVMAX);
11930: free_ivector(Tprod,1,NCOVMAX);
11931: free_ivector(Tvaraff,1,NCOVMAX);
11932: free_ivector(invalidvarcomb,1,ncovcombmax);
11933: free_ivector(Tage,1,NCOVMAX);
11934: free_ivector(Tmodelind,1,NCOVMAX);
1.228 brouard 11935: free_ivector(TmodelInvind,1,NCOVMAX);
11936: free_ivector(TmodelInvQind,1,NCOVMAX);
1.227 brouard 11937:
11938: free_imatrix(nbcode,0,NCOVMAX,0,NCOVMAX);
11939: /* free_imatrix(codtab,1,100,1,10); */
1.126 brouard 11940: fflush(fichtm);
11941: fflush(ficgp);
11942:
1.227 brouard 11943:
1.126 brouard 11944: if((nberr >0) || (nbwarn>0)){
1.216 brouard 11945: printf("End of Imach with %d errors and/or %d warnings. Please look at the log file for details.\n",nberr,nbwarn);
11946: fprintf(ficlog,"End of Imach with %d errors and/or warnings %d. Please look at the log file for details.\n",nberr,nbwarn);
1.126 brouard 11947: }else{
11948: printf("End of Imach\n");
11949: fprintf(ficlog,"End of Imach\n");
11950: }
11951: printf("See log file on %s\n",filelog);
11952: /* gettimeofday(&end_time, (struct timezone*)0);*/ /* after time */
1.157 brouard 11953: /*(void) gettimeofday(&end_time,&tzp);*/
11954: rend_time = time(NULL);
11955: end_time = *localtime(&rend_time);
11956: /* tml = *localtime(&end_time.tm_sec); */
11957: strcpy(strtend,asctime(&end_time));
1.126 brouard 11958: printf("Local time at start %s\nLocal time at end %s",strstart, strtend);
11959: fprintf(ficlog,"Local time at start %s\nLocal time at end %s\n",strstart, strtend);
1.157 brouard 11960: printf("Total time used %s\n", asc_diff_time(rend_time -rstart_time,tmpout));
1.227 brouard 11961:
1.157 brouard 11962: printf("Total time was %.0lf Sec.\n", difftime(rend_time,rstart_time));
11963: fprintf(ficlog,"Total time used %s\n", asc_diff_time(rend_time -rstart_time,tmpout));
11964: fprintf(ficlog,"Total time was %.0lf Sec.\n", difftime(rend_time,rstart_time));
1.126 brouard 11965: /* printf("Total time was %d uSec.\n", total_usecs);*/
11966: /* if(fileappend(fichtm,optionfilehtm)){ */
11967: fprintf(fichtm,"<br>Local time at start %s<br>Local time at end %s<br>\n</body></html>",strstart, strtend);
11968: fclose(fichtm);
11969: fprintf(fichtmcov,"<br>Local time at start %s<br>Local time at end %s<br>\n</body></html>",strstart, strtend);
11970: fclose(fichtmcov);
11971: fclose(ficgp);
11972: fclose(ficlog);
11973: /*------ End -----------*/
1.227 brouard 11974:
11975:
11976: printf("Before Current directory %s!\n",pathcd);
1.184 brouard 11977: #ifdef WIN32
1.227 brouard 11978: if (_chdir(pathcd) != 0)
11979: printf("Can't move to directory %s!\n",path);
11980: if(_getcwd(pathcd,MAXLINE) > 0)
1.184 brouard 11981: #else
1.227 brouard 11982: if(chdir(pathcd) != 0)
11983: printf("Can't move to directory %s!\n", path);
11984: if (getcwd(pathcd, MAXLINE) > 0)
1.184 brouard 11985: #endif
1.126 brouard 11986: printf("Current directory %s!\n",pathcd);
11987: /*strcat(plotcmd,CHARSEPARATOR);*/
11988: sprintf(plotcmd,"gnuplot");
1.157 brouard 11989: #ifdef _WIN32
1.126 brouard 11990: sprintf(plotcmd,"\"%sgnuplot.exe\"",pathimach);
11991: #endif
11992: if(!stat(plotcmd,&info)){
1.158 brouard 11993: printf("Error or gnuplot program not found: '%s'\n",plotcmd);fflush(stdout);
1.126 brouard 11994: if(!stat(getenv("GNUPLOTBIN"),&info)){
1.158 brouard 11995: printf("Error or gnuplot program not found: '%s' Environment GNUPLOTBIN not set.\n",plotcmd);fflush(stdout);
1.126 brouard 11996: }else
11997: strcpy(pplotcmd,plotcmd);
1.157 brouard 11998: #ifdef __unix
1.126 brouard 11999: strcpy(plotcmd,GNUPLOTPROGRAM);
12000: if(!stat(plotcmd,&info)){
1.158 brouard 12001: printf("Error gnuplot program not found: '%s'\n",plotcmd);fflush(stdout);
1.126 brouard 12002: }else
12003: strcpy(pplotcmd,plotcmd);
12004: #endif
12005: }else
12006: strcpy(pplotcmd,plotcmd);
12007:
12008: sprintf(plotcmd,"%s %s",pplotcmd, optionfilegnuplot);
1.158 brouard 12009: printf("Starting graphs with: '%s'\n",plotcmd);fflush(stdout);
1.227 brouard 12010:
1.126 brouard 12011: if((outcmd=system(plotcmd)) != 0){
1.158 brouard 12012: printf("gnuplot command might not be in your path: '%s', err=%d\n", plotcmd, outcmd);
1.154 brouard 12013: printf("\n Trying if gnuplot resides on the same directory that IMaCh\n");
1.152 brouard 12014: sprintf(plotcmd,"%sgnuplot %s", pathimach, optionfilegnuplot);
1.150 brouard 12015: if((outcmd=system(plotcmd)) != 0)
1.153 brouard 12016: printf("\n Still a problem with gnuplot command %s, err=%d\n", plotcmd, outcmd);
1.126 brouard 12017: }
1.158 brouard 12018: printf(" Successful, please wait...");
1.126 brouard 12019: while (z[0] != 'q') {
12020: /* chdir(path); */
1.154 brouard 12021: printf("\nType e to edit results with your browser, g to graph again and q for exit: ");
1.126 brouard 12022: scanf("%s",z);
12023: /* if (z[0] == 'c') system("./imach"); */
12024: if (z[0] == 'e') {
1.158 brouard 12025: #ifdef __APPLE__
1.152 brouard 12026: sprintf(pplotcmd, "open %s", optionfilehtm);
1.157 brouard 12027: #elif __linux
12028: sprintf(pplotcmd, "xdg-open %s", optionfilehtm);
1.153 brouard 12029: #else
1.152 brouard 12030: sprintf(pplotcmd, "%s", optionfilehtm);
1.153 brouard 12031: #endif
12032: printf("Starting browser with: %s",pplotcmd);fflush(stdout);
12033: system(pplotcmd);
1.126 brouard 12034: }
12035: else if (z[0] == 'g') system(plotcmd);
12036: else if (z[0] == 'q') exit(0);
12037: }
1.227 brouard 12038: end:
1.126 brouard 12039: while (z[0] != 'q') {
1.195 brouard 12040: printf("\nType q for exiting: "); fflush(stdout);
1.126 brouard 12041: scanf("%s",z);
12042: }
12043: }
FreeBSD-CVSweb <freebsd-cvsweb@FreeBSD.org>
![[BACK]](/icons/small/back.gif){kind=icon}
Return to imach.c CVS log ![[TXT]](/icons/small/text.gif){kind=icon}
![[DIR]](/icons/small/dir.gif){kind=icon}
Up to [Local Repository] / imach / src