Annotation of imach/src/imach.c, revision 1.248
1.248 ! brouard 1: /* $Id: imach.c,v 1.247 2016/09/02 11:11:21 brouard Exp $
1.126 brouard 2: $State: Exp $
1.163 brouard 3: $Log: imach.c,v $
1.248 ! brouard 4: Revision 1.247 2016/09/02 11:11:21 brouard
! 5: *** empty log message ***
! 6:
1.247 brouard 7: Revision 1.246 2016/09/02 08:49:22 brouard
8: *** empty log message ***
9:
1.246 brouard 10: Revision 1.245 2016/09/02 07:25:01 brouard
11: *** empty log message ***
12:
1.245 brouard 13: Revision 1.244 2016/09/02 07:17:34 brouard
14: *** empty log message ***
15:
1.244 brouard 16: Revision 1.243 2016/09/02 06:45:35 brouard
17: *** empty log message ***
18:
1.243 brouard 19: Revision 1.242 2016/08/30 15:01:20 brouard
20: Summary: Fixing a lots
21:
1.242 brouard 22: Revision 1.241 2016/08/29 17:17:25 brouard
23: Summary: gnuplot problem in Back projection to fix
24:
1.241 brouard 25: Revision 1.240 2016/08/29 07:53:18 brouard
26: Summary: Better
27:
1.240 brouard 28: Revision 1.239 2016/08/26 15:51:03 brouard
29: Summary: Improvement in Powell output in order to copy and paste
30:
31: Author:
32:
1.239 brouard 33: Revision 1.238 2016/08/26 14:23:35 brouard
34: Summary: Starting tests of 0.99
35:
1.238 brouard 36: Revision 1.237 2016/08/26 09:20:19 brouard
37: Summary: to valgrind
38:
1.237 brouard 39: Revision 1.236 2016/08/25 10:50:18 brouard
40: *** empty log message ***
41:
1.236 brouard 42: Revision 1.235 2016/08/25 06:59:23 brouard
43: *** empty log message ***
44:
1.235 brouard 45: Revision 1.234 2016/08/23 16:51:20 brouard
46: *** empty log message ***
47:
1.234 brouard 48: Revision 1.233 2016/08/23 07:40:50 brouard
49: Summary: not working
50:
1.233 brouard 51: Revision 1.232 2016/08/22 14:20:21 brouard
52: Summary: not working
53:
1.232 brouard 54: Revision 1.231 2016/08/22 07:17:15 brouard
55: Summary: not working
56:
1.231 brouard 57: Revision 1.230 2016/08/22 06:55:53 brouard
58: Summary: Not working
59:
1.230 brouard 60: Revision 1.229 2016/07/23 09:45:53 brouard
61: Summary: Completing for func too
62:
1.229 brouard 63: Revision 1.228 2016/07/22 17:45:30 brouard
64: Summary: Fixing some arrays, still debugging
65:
1.227 brouard 66: Revision 1.226 2016/07/12 18:42:34 brouard
67: Summary: temp
68:
1.226 brouard 69: Revision 1.225 2016/07/12 08:40:03 brouard
70: Summary: saving but not running
71:
1.225 brouard 72: Revision 1.224 2016/07/01 13:16:01 brouard
73: Summary: Fixes
74:
1.224 brouard 75: Revision 1.223 2016/02/19 09:23:35 brouard
76: Summary: temporary
77:
1.223 brouard 78: Revision 1.222 2016/02/17 08:14:50 brouard
79: Summary: Probably last 0.98 stable version 0.98r6
80:
1.222 brouard 81: Revision 1.221 2016/02/15 23:35:36 brouard
82: Summary: minor bug
83:
1.220 brouard 84: Revision 1.219 2016/02/15 00:48:12 brouard
85: *** empty log message ***
86:
1.219 brouard 87: Revision 1.218 2016/02/12 11:29:23 brouard
88: Summary: 0.99 Back projections
89:
1.218 brouard 90: Revision 1.217 2015/12/23 17:18:31 brouard
91: Summary: Experimental backcast
92:
1.217 brouard 93: Revision 1.216 2015/12/18 17:32:11 brouard
94: Summary: 0.98r4 Warning and status=-2
95:
96: Version 0.98r4 is now:
97: - displaying an error when status is -1, date of interview unknown and date of death known;
98: - permitting a status -2 when the vital status is unknown at a known date of right truncation.
99: Older changes concerning s=-2, dating from 2005 have been supersed.
100:
1.216 brouard 101: Revision 1.215 2015/12/16 08:52:24 brouard
102: Summary: 0.98r4 working
103:
1.215 brouard 104: Revision 1.214 2015/12/16 06:57:54 brouard
105: Summary: temporary not working
106:
1.214 brouard 107: Revision 1.213 2015/12/11 18:22:17 brouard
108: Summary: 0.98r4
109:
1.213 brouard 110: Revision 1.212 2015/11/21 12:47:24 brouard
111: Summary: minor typo
112:
1.212 brouard 113: Revision 1.211 2015/11/21 12:41:11 brouard
114: Summary: 0.98r3 with some graph of projected cross-sectional
115:
116: Author: Nicolas Brouard
117:
1.211 brouard 118: Revision 1.210 2015/11/18 17:41:20 brouard
119: Summary: Start working on projected prevalences
120:
1.210 brouard 121: Revision 1.209 2015/11/17 22:12:03 brouard
122: Summary: Adding ftolpl parameter
123: Author: N Brouard
124:
125: We had difficulties to get smoothed confidence intervals. It was due
126: to the period prevalence which wasn't computed accurately. The inner
127: parameter ftolpl is now an outer parameter of the .imach parameter
128: file after estepm. If ftolpl is small 1.e-4 and estepm too,
129: computation are long.
130:
1.209 brouard 131: Revision 1.208 2015/11/17 14:31:57 brouard
132: Summary: temporary
133:
1.208 brouard 134: Revision 1.207 2015/10/27 17:36:57 brouard
135: *** empty log message ***
136:
1.207 brouard 137: Revision 1.206 2015/10/24 07:14:11 brouard
138: *** empty log message ***
139:
1.206 brouard 140: Revision 1.205 2015/10/23 15:50:53 brouard
141: Summary: 0.98r3 some clarification for graphs on likelihood contributions
142:
1.205 brouard 143: Revision 1.204 2015/10/01 16:20:26 brouard
144: Summary: Some new graphs of contribution to likelihood
145:
1.204 brouard 146: Revision 1.203 2015/09/30 17:45:14 brouard
147: Summary: looking at better estimation of the hessian
148:
149: Also a better criteria for convergence to the period prevalence And
150: therefore adding the number of years needed to converge. (The
151: prevalence in any alive state shold sum to one
152:
1.203 brouard 153: Revision 1.202 2015/09/22 19:45:16 brouard
154: Summary: Adding some overall graph on contribution to likelihood. Might change
155:
1.202 brouard 156: Revision 1.201 2015/09/15 17:34:58 brouard
157: Summary: 0.98r0
158:
159: - Some new graphs like suvival functions
160: - Some bugs fixed like model=1+age+V2.
161:
1.201 brouard 162: Revision 1.200 2015/09/09 16:53:55 brouard
163: Summary: Big bug thanks to Flavia
164:
165: Even model=1+age+V2. did not work anymore
166:
1.200 brouard 167: Revision 1.199 2015/09/07 14:09:23 brouard
168: Summary: 0.98q6 changing default small png format for graph to vectorized svg.
169:
1.199 brouard 170: Revision 1.198 2015/09/03 07:14:39 brouard
171: Summary: 0.98q5 Flavia
172:
1.198 brouard 173: Revision 1.197 2015/09/01 18:24:39 brouard
174: *** empty log message ***
175:
1.197 brouard 176: Revision 1.196 2015/08/18 23:17:52 brouard
177: Summary: 0.98q5
178:
1.196 brouard 179: Revision 1.195 2015/08/18 16:28:39 brouard
180: Summary: Adding a hack for testing purpose
181:
182: After reading the title, ftol and model lines, if the comment line has
183: a q, starting with #q, the answer at the end of the run is quit. It
184: permits to run test files in batch with ctest. The former workaround was
185: $ echo q | imach foo.imach
186:
1.195 brouard 187: Revision 1.194 2015/08/18 13:32:00 brouard
188: Summary: Adding error when the covariance matrix doesn't contain the exact number of lines required by the model line.
189:
1.194 brouard 190: Revision 1.193 2015/08/04 07:17:42 brouard
191: Summary: 0.98q4
192:
1.193 brouard 193: Revision 1.192 2015/07/16 16:49:02 brouard
194: Summary: Fixing some outputs
195:
1.192 brouard 196: Revision 1.191 2015/07/14 10:00:33 brouard
197: Summary: Some fixes
198:
1.191 brouard 199: Revision 1.190 2015/05/05 08:51:13 brouard
200: Summary: Adding digits in output parameters (7 digits instead of 6)
201:
202: Fix 1+age+.
203:
1.190 brouard 204: Revision 1.189 2015/04/30 14:45:16 brouard
205: Summary: 0.98q2
206:
1.189 brouard 207: Revision 1.188 2015/04/30 08:27:53 brouard
208: *** empty log message ***
209:
1.188 brouard 210: Revision 1.187 2015/04/29 09:11:15 brouard
211: *** empty log message ***
212:
1.187 brouard 213: Revision 1.186 2015/04/23 12:01:52 brouard
214: Summary: V1*age is working now, version 0.98q1
215:
216: Some codes had been disabled in order to simplify and Vn*age was
217: working in the optimization phase, ie, giving correct MLE parameters,
218: but, as usual, outputs were not correct and program core dumped.
219:
1.186 brouard 220: Revision 1.185 2015/03/11 13:26:42 brouard
221: Summary: Inclusion of compile and links command line for Intel Compiler
222:
1.185 brouard 223: Revision 1.184 2015/03/11 11:52:39 brouard
224: Summary: Back from Windows 8. Intel Compiler
225:
1.184 brouard 226: Revision 1.183 2015/03/10 20:34:32 brouard
227: Summary: 0.98q0, trying with directest, mnbrak fixed
228:
229: We use directest instead of original Powell test; probably no
230: incidence on the results, but better justifications;
231: We fixed Numerical Recipes mnbrak routine which was wrong and gave
232: wrong results.
233:
1.183 brouard 234: Revision 1.182 2015/02/12 08:19:57 brouard
235: Summary: Trying to keep directest which seems simpler and more general
236: Author: Nicolas Brouard
237:
1.182 brouard 238: Revision 1.181 2015/02/11 23:22:24 brouard
239: Summary: Comments on Powell added
240:
241: Author:
242:
1.181 brouard 243: Revision 1.180 2015/02/11 17:33:45 brouard
244: Summary: Finishing move from main to function (hpijx and prevalence_limit)
245:
1.180 brouard 246: Revision 1.179 2015/01/04 09:57:06 brouard
247: Summary: back to OS/X
248:
1.179 brouard 249: Revision 1.178 2015/01/04 09:35:48 brouard
250: *** empty log message ***
251:
1.178 brouard 252: Revision 1.177 2015/01/03 18:40:56 brouard
253: Summary: Still testing ilc32 on OSX
254:
1.177 brouard 255: Revision 1.176 2015/01/03 16:45:04 brouard
256: *** empty log message ***
257:
1.176 brouard 258: Revision 1.175 2015/01/03 16:33:42 brouard
259: *** empty log message ***
260:
1.175 brouard 261: Revision 1.174 2015/01/03 16:15:49 brouard
262: Summary: Still in cross-compilation
263:
1.174 brouard 264: Revision 1.173 2015/01/03 12:06:26 brouard
265: Summary: trying to detect cross-compilation
266:
1.173 brouard 267: Revision 1.172 2014/12/27 12:07:47 brouard
268: Summary: Back from Visual Studio and Intel, options for compiling for Windows XP
269:
1.172 brouard 270: Revision 1.171 2014/12/23 13:26:59 brouard
271: Summary: Back from Visual C
272:
273: Still problem with utsname.h on Windows
274:
1.171 brouard 275: Revision 1.170 2014/12/23 11:17:12 brouard
276: Summary: Cleaning some \%% back to %%
277:
278: The escape was mandatory for a specific compiler (which one?), but too many warnings.
279:
1.170 brouard 280: Revision 1.169 2014/12/22 23:08:31 brouard
281: Summary: 0.98p
282:
283: Outputs some informations on compiler used, OS etc. Testing on different platforms.
284:
1.169 brouard 285: Revision 1.168 2014/12/22 15:17:42 brouard
1.170 brouard 286: Summary: update
1.169 brouard 287:
1.168 brouard 288: Revision 1.167 2014/12/22 13:50:56 brouard
289: Summary: Testing uname and compiler version and if compiled 32 or 64
290:
291: Testing on Linux 64
292:
1.167 brouard 293: Revision 1.166 2014/12/22 11:40:47 brouard
294: *** empty log message ***
295:
1.166 brouard 296: Revision 1.165 2014/12/16 11:20:36 brouard
297: Summary: After compiling on Visual C
298:
299: * imach.c (Module): Merging 1.61 to 1.162
300:
1.165 brouard 301: Revision 1.164 2014/12/16 10:52:11 brouard
302: Summary: Merging with Visual C after suppressing some warnings for unused variables. Also fixing Saito's bug 0.98Xn
303:
304: * imach.c (Module): Merging 1.61 to 1.162
305:
1.164 brouard 306: Revision 1.163 2014/12/16 10:30:11 brouard
307: * imach.c (Module): Merging 1.61 to 1.162
308:
1.163 brouard 309: Revision 1.162 2014/09/25 11:43:39 brouard
310: Summary: temporary backup 0.99!
311:
1.162 brouard 312: Revision 1.1 2014/09/16 11:06:58 brouard
313: Summary: With some code (wrong) for nlopt
314:
315: Author:
316:
317: Revision 1.161 2014/09/15 20:41:41 brouard
318: Summary: Problem with macro SQR on Intel compiler
319:
1.161 brouard 320: Revision 1.160 2014/09/02 09:24:05 brouard
321: *** empty log message ***
322:
1.160 brouard 323: Revision 1.159 2014/09/01 10:34:10 brouard
324: Summary: WIN32
325: Author: Brouard
326:
1.159 brouard 327: Revision 1.158 2014/08/27 17:11:51 brouard
328: *** empty log message ***
329:
1.158 brouard 330: Revision 1.157 2014/08/27 16:26:55 brouard
331: Summary: Preparing windows Visual studio version
332: Author: Brouard
333:
334: In order to compile on Visual studio, time.h is now correct and time_t
335: and tm struct should be used. difftime should be used but sometimes I
336: just make the differences in raw time format (time(&now).
337: Trying to suppress #ifdef LINUX
338: Add xdg-open for __linux in order to open default browser.
339:
1.157 brouard 340: Revision 1.156 2014/08/25 20:10:10 brouard
341: *** empty log message ***
342:
1.156 brouard 343: Revision 1.155 2014/08/25 18:32:34 brouard
344: Summary: New compile, minor changes
345: Author: Brouard
346:
1.155 brouard 347: Revision 1.154 2014/06/20 17:32:08 brouard
348: Summary: Outputs now all graphs of convergence to period prevalence
349:
1.154 brouard 350: Revision 1.153 2014/06/20 16:45:46 brouard
351: Summary: If 3 live state, convergence to period prevalence on same graph
352: Author: Brouard
353:
1.153 brouard 354: Revision 1.152 2014/06/18 17:54:09 brouard
355: Summary: open browser, use gnuplot on same dir than imach if not found in the path
356:
1.152 brouard 357: Revision 1.151 2014/06/18 16:43:30 brouard
358: *** empty log message ***
359:
1.151 brouard 360: Revision 1.150 2014/06/18 16:42:35 brouard
361: Summary: If gnuplot is not in the path try on same directory than imach binary (OSX)
362: Author: brouard
363:
1.150 brouard 364: Revision 1.149 2014/06/18 15:51:14 brouard
365: Summary: Some fixes in parameter files errors
366: Author: Nicolas Brouard
367:
1.149 brouard 368: Revision 1.148 2014/06/17 17:38:48 brouard
369: Summary: Nothing new
370: Author: Brouard
371:
372: Just a new packaging for OS/X version 0.98nS
373:
1.148 brouard 374: Revision 1.147 2014/06/16 10:33:11 brouard
375: *** empty log message ***
376:
1.147 brouard 377: Revision 1.146 2014/06/16 10:20:28 brouard
378: Summary: Merge
379: Author: Brouard
380:
381: Merge, before building revised version.
382:
1.146 brouard 383: Revision 1.145 2014/06/10 21:23:15 brouard
384: Summary: Debugging with valgrind
385: Author: Nicolas Brouard
386:
387: Lot of changes in order to output the results with some covariates
388: After the Edimburgh REVES conference 2014, it seems mandatory to
389: improve the code.
390: No more memory valgrind error but a lot has to be done in order to
391: continue the work of splitting the code into subroutines.
392: Also, decodemodel has been improved. Tricode is still not
393: optimal. nbcode should be improved. Documentation has been added in
394: the source code.
395:
1.144 brouard 396: Revision 1.143 2014/01/26 09:45:38 brouard
397: Summary: Version 0.98nR (to be improved, but gives same optimization results as 0.98k. Nice, promising
398:
399: * imach.c (Module): Trying to merge old staffs together while being at Tokyo. Not tested...
400: (Module): Version 0.98nR Running ok, but output format still only works for three covariates.
401:
1.143 brouard 402: Revision 1.142 2014/01/26 03:57:36 brouard
403: Summary: gnuplot changed plot w l 1 has to be changed to plot w l lt 2
404:
405: * imach.c (Module): Trying to merge old staffs together while being at Tokyo. Not tested...
406:
1.142 brouard 407: Revision 1.141 2014/01/26 02:42:01 brouard
408: * imach.c (Module): Trying to merge old staffs together while being at Tokyo. Not tested...
409:
1.141 brouard 410: Revision 1.140 2011/09/02 10:37:54 brouard
411: Summary: times.h is ok with mingw32 now.
412:
1.140 brouard 413: Revision 1.139 2010/06/14 07:50:17 brouard
414: After the theft of my laptop, I probably lost some lines of codes which were not uploaded to the CVS tree.
415: I remember having already fixed agemin agemax which are pointers now but not cvs saved.
416:
1.139 brouard 417: Revision 1.138 2010/04/30 18:19:40 brouard
418: *** empty log message ***
419:
1.138 brouard 420: Revision 1.137 2010/04/29 18:11:38 brouard
421: (Module): Checking covariates for more complex models
422: than V1+V2. A lot of change to be done. Unstable.
423:
1.137 brouard 424: Revision 1.136 2010/04/26 20:30:53 brouard
425: (Module): merging some libgsl code. Fixing computation
426: of likelione (using inter/intrapolation if mle = 0) in order to
427: get same likelihood as if mle=1.
428: Some cleaning of code and comments added.
429:
1.136 brouard 430: Revision 1.135 2009/10/29 15:33:14 brouard
431: (Module): Now imach stops if date of birth, at least year of birth, is not given. Some cleaning of the code.
432:
1.135 brouard 433: Revision 1.134 2009/10/29 13:18:53 brouard
434: (Module): Now imach stops if date of birth, at least year of birth, is not given. Some cleaning of the code.
435:
1.134 brouard 436: Revision 1.133 2009/07/06 10:21:25 brouard
437: just nforces
438:
1.133 brouard 439: Revision 1.132 2009/07/06 08:22:05 brouard
440: Many tings
441:
1.132 brouard 442: Revision 1.131 2009/06/20 16:22:47 brouard
443: Some dimensions resccaled
444:
1.131 brouard 445: Revision 1.130 2009/05/26 06:44:34 brouard
446: (Module): Max Covariate is now set to 20 instead of 8. A
447: lot of cleaning with variables initialized to 0. Trying to make
448: V2+V3*age+V1+V4 strb=V3*age+V1+V4 working better.
449:
1.130 brouard 450: Revision 1.129 2007/08/31 13:49:27 lievre
451: Modification of the way of exiting when the covariate is not binary in order to see on the window the error message before exiting
452:
1.129 lievre 453: Revision 1.128 2006/06/30 13:02:05 brouard
454: (Module): Clarifications on computing e.j
455:
1.128 brouard 456: Revision 1.127 2006/04/28 18:11:50 brouard
457: (Module): Yes the sum of survivors was wrong since
458: imach-114 because nhstepm was no more computed in the age
459: loop. Now we define nhstepma in the age loop.
460: (Module): In order to speed up (in case of numerous covariates) we
461: compute health expectancies (without variances) in a first step
462: and then all the health expectancies with variances or standard
463: deviation (needs data from the Hessian matrices) which slows the
464: computation.
465: In the future we should be able to stop the program is only health
466: expectancies and graph are needed without standard deviations.
467:
1.127 brouard 468: Revision 1.126 2006/04/28 17:23:28 brouard
469: (Module): Yes the sum of survivors was wrong since
470: imach-114 because nhstepm was no more computed in the age
471: loop. Now we define nhstepma in the age loop.
472: Version 0.98h
473:
1.126 brouard 474: Revision 1.125 2006/04/04 15:20:31 lievre
475: Errors in calculation of health expectancies. Age was not initialized.
476: Forecasting file added.
477:
478: Revision 1.124 2006/03/22 17:13:53 lievre
479: Parameters are printed with %lf instead of %f (more numbers after the comma).
480: The log-likelihood is printed in the log file
481:
482: Revision 1.123 2006/03/20 10:52:43 brouard
483: * imach.c (Module): <title> changed, corresponds to .htm file
484: name. <head> headers where missing.
485:
486: * imach.c (Module): Weights can have a decimal point as for
487: English (a comma might work with a correct LC_NUMERIC environment,
488: otherwise the weight is truncated).
489: Modification of warning when the covariates values are not 0 or
490: 1.
491: Version 0.98g
492:
493: Revision 1.122 2006/03/20 09:45:41 brouard
494: (Module): Weights can have a decimal point as for
495: English (a comma might work with a correct LC_NUMERIC environment,
496: otherwise the weight is truncated).
497: Modification of warning when the covariates values are not 0 or
498: 1.
499: Version 0.98g
500:
501: Revision 1.121 2006/03/16 17:45:01 lievre
502: * imach.c (Module): Comments concerning covariates added
503:
504: * imach.c (Module): refinements in the computation of lli if
505: status=-2 in order to have more reliable computation if stepm is
506: not 1 month. Version 0.98f
507:
508: Revision 1.120 2006/03/16 15:10:38 lievre
509: (Module): refinements in the computation of lli if
510: status=-2 in order to have more reliable computation if stepm is
511: not 1 month. Version 0.98f
512:
513: Revision 1.119 2006/03/15 17:42:26 brouard
514: (Module): Bug if status = -2, the loglikelihood was
515: computed as likelihood omitting the logarithm. Version O.98e
516:
517: Revision 1.118 2006/03/14 18:20:07 brouard
518: (Module): varevsij Comments added explaining the second
519: table of variances if popbased=1 .
520: (Module): Covariances of eij, ekl added, graphs fixed, new html link.
521: (Module): Function pstamp added
522: (Module): Version 0.98d
523:
524: Revision 1.117 2006/03/14 17:16:22 brouard
525: (Module): varevsij Comments added explaining the second
526: table of variances if popbased=1 .
527: (Module): Covariances of eij, ekl added, graphs fixed, new html link.
528: (Module): Function pstamp added
529: (Module): Version 0.98d
530:
531: Revision 1.116 2006/03/06 10:29:27 brouard
532: (Module): Variance-covariance wrong links and
533: varian-covariance of ej. is needed (Saito).
534:
535: Revision 1.115 2006/02/27 12:17:45 brouard
536: (Module): One freematrix added in mlikeli! 0.98c
537:
538: Revision 1.114 2006/02/26 12:57:58 brouard
539: (Module): Some improvements in processing parameter
540: filename with strsep.
541:
542: Revision 1.113 2006/02/24 14:20:24 brouard
543: (Module): Memory leaks checks with valgrind and:
544: datafile was not closed, some imatrix were not freed and on matrix
545: allocation too.
546:
547: Revision 1.112 2006/01/30 09:55:26 brouard
548: (Module): Back to gnuplot.exe instead of wgnuplot.exe
549:
550: Revision 1.111 2006/01/25 20:38:18 brouard
551: (Module): Lots of cleaning and bugs added (Gompertz)
552: (Module): Comments can be added in data file. Missing date values
553: can be a simple dot '.'.
554:
555: Revision 1.110 2006/01/25 00:51:50 brouard
556: (Module): Lots of cleaning and bugs added (Gompertz)
557:
558: Revision 1.109 2006/01/24 19:37:15 brouard
559: (Module): Comments (lines starting with a #) are allowed in data.
560:
561: Revision 1.108 2006/01/19 18:05:42 lievre
562: Gnuplot problem appeared...
563: To be fixed
564:
565: Revision 1.107 2006/01/19 16:20:37 brouard
566: Test existence of gnuplot in imach path
567:
568: Revision 1.106 2006/01/19 13:24:36 brouard
569: Some cleaning and links added in html output
570:
571: Revision 1.105 2006/01/05 20:23:19 lievre
572: *** empty log message ***
573:
574: Revision 1.104 2005/09/30 16:11:43 lievre
575: (Module): sump fixed, loop imx fixed, and simplifications.
576: (Module): If the status is missing at the last wave but we know
577: that the person is alive, then we can code his/her status as -2
578: (instead of missing=-1 in earlier versions) and his/her
579: contributions to the likelihood is 1 - Prob of dying from last
580: health status (= 1-p13= p11+p12 in the easiest case of somebody in
581: the healthy state at last known wave). Version is 0.98
582:
583: Revision 1.103 2005/09/30 15:54:49 lievre
584: (Module): sump fixed, loop imx fixed, and simplifications.
585:
586: Revision 1.102 2004/09/15 17:31:30 brouard
587: Add the possibility to read data file including tab characters.
588:
589: Revision 1.101 2004/09/15 10:38:38 brouard
590: Fix on curr_time
591:
592: Revision 1.100 2004/07/12 18:29:06 brouard
593: Add version for Mac OS X. Just define UNIX in Makefile
594:
595: Revision 1.99 2004/06/05 08:57:40 brouard
596: *** empty log message ***
597:
598: Revision 1.98 2004/05/16 15:05:56 brouard
599: New version 0.97 . First attempt to estimate force of mortality
600: directly from the data i.e. without the need of knowing the health
601: state at each age, but using a Gompertz model: log u =a + b*age .
602: This is the basic analysis of mortality and should be done before any
603: other analysis, in order to test if the mortality estimated from the
604: cross-longitudinal survey is different from the mortality estimated
605: from other sources like vital statistic data.
606:
607: The same imach parameter file can be used but the option for mle should be -3.
608:
1.133 brouard 609: Agnès, who wrote this part of the code, tried to keep most of the
1.126 brouard 610: former routines in order to include the new code within the former code.
611:
612: The output is very simple: only an estimate of the intercept and of
613: the slope with 95% confident intervals.
614:
615: Current limitations:
616: A) Even if you enter covariates, i.e. with the
617: model= V1+V2 equation for example, the programm does only estimate a unique global model without covariates.
618: B) There is no computation of Life Expectancy nor Life Table.
619:
620: Revision 1.97 2004/02/20 13:25:42 lievre
621: Version 0.96d. Population forecasting command line is (temporarily)
622: suppressed.
623:
624: Revision 1.96 2003/07/15 15:38:55 brouard
625: * imach.c (Repository): Errors in subdirf, 2, 3 while printing tmpout is
626: rewritten within the same printf. Workaround: many printfs.
627:
628: Revision 1.95 2003/07/08 07:54:34 brouard
629: * imach.c (Repository):
630: (Repository): Using imachwizard code to output a more meaningful covariance
631: matrix (cov(a12,c31) instead of numbers.
632:
633: Revision 1.94 2003/06/27 13:00:02 brouard
634: Just cleaning
635:
636: Revision 1.93 2003/06/25 16:33:55 brouard
637: (Module): On windows (cygwin) function asctime_r doesn't
638: exist so I changed back to asctime which exists.
639: (Module): Version 0.96b
640:
641: Revision 1.92 2003/06/25 16:30:45 brouard
642: (Module): On windows (cygwin) function asctime_r doesn't
643: exist so I changed back to asctime which exists.
644:
645: Revision 1.91 2003/06/25 15:30:29 brouard
646: * imach.c (Repository): Duplicated warning errors corrected.
647: (Repository): Elapsed time after each iteration is now output. It
648: helps to forecast when convergence will be reached. Elapsed time
649: is stamped in powell. We created a new html file for the graphs
650: concerning matrix of covariance. It has extension -cov.htm.
651:
652: Revision 1.90 2003/06/24 12:34:15 brouard
653: (Module): Some bugs corrected for windows. Also, when
654: mle=-1 a template is output in file "or"mypar.txt with the design
655: of the covariance matrix to be input.
656:
657: Revision 1.89 2003/06/24 12:30:52 brouard
658: (Module): Some bugs corrected for windows. Also, when
659: mle=-1 a template is output in file "or"mypar.txt with the design
660: of the covariance matrix to be input.
661:
662: Revision 1.88 2003/06/23 17:54:56 brouard
663: * 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.
664:
665: Revision 1.87 2003/06/18 12:26:01 brouard
666: Version 0.96
667:
668: Revision 1.86 2003/06/17 20:04:08 brouard
669: (Module): Change position of html and gnuplot routines and added
670: routine fileappend.
671:
672: Revision 1.85 2003/06/17 13:12:43 brouard
673: * imach.c (Repository): Check when date of death was earlier that
674: current date of interview. It may happen when the death was just
675: prior to the death. In this case, dh was negative and likelihood
676: was wrong (infinity). We still send an "Error" but patch by
677: assuming that the date of death was just one stepm after the
678: interview.
679: (Repository): Because some people have very long ID (first column)
680: we changed int to long in num[] and we added a new lvector for
681: memory allocation. But we also truncated to 8 characters (left
682: truncation)
683: (Repository): No more line truncation errors.
684:
685: Revision 1.84 2003/06/13 21:44:43 brouard
686: * imach.c (Repository): Replace "freqsummary" at a correct
687: place. It differs from routine "prevalence" which may be called
688: many times. Probs is memory consuming and must be used with
689: parcimony.
690: Version 0.95a3 (should output exactly the same maximization than 0.8a2)
691:
692: Revision 1.83 2003/06/10 13:39:11 lievre
693: *** empty log message ***
694:
695: Revision 1.82 2003/06/05 15:57:20 brouard
696: Add log in imach.c and fullversion number is now printed.
697:
698: */
699: /*
700: Interpolated Markov Chain
701:
702: Short summary of the programme:
703:
1.227 brouard 704: This program computes Healthy Life Expectancies or State-specific
705: (if states aren't health statuses) Expectancies from
706: cross-longitudinal data. Cross-longitudinal data consist in:
707:
708: -1- a first survey ("cross") where individuals from different ages
709: are interviewed on their health status or degree of disability (in
710: the case of a health survey which is our main interest)
711:
712: -2- at least a second wave of interviews ("longitudinal") which
713: measure each change (if any) in individual health status. Health
714: expectancies are computed from the time spent in each health state
715: according to a model. More health states you consider, more time is
716: necessary to reach the Maximum Likelihood of the parameters involved
717: in the model. The simplest model is the multinomial logistic model
718: where pij is the probability to be observed in state j at the second
719: wave conditional to be observed in state i at the first
720: wave. Therefore the model is: log(pij/pii)= aij + bij*age+ cij*sex +
721: etc , where 'age' is age and 'sex' is a covariate. If you want to
722: have a more complex model than "constant and age", you should modify
723: the program where the markup *Covariates have to be included here
724: again* invites you to do it. More covariates you add, slower the
1.126 brouard 725: convergence.
726:
727: The advantage of this computer programme, compared to a simple
728: multinomial logistic model, is clear when the delay between waves is not
729: identical for each individual. Also, if a individual missed an
730: intermediate interview, the information is lost, but taken into
731: account using an interpolation or extrapolation.
732:
733: hPijx is the probability to be observed in state i at age x+h
734: conditional to the observed state i at age x. The delay 'h' can be
735: split into an exact number (nh*stepm) of unobserved intermediate
736: states. This elementary transition (by month, quarter,
737: semester or year) is modelled as a multinomial logistic. The hPx
738: matrix is simply the matrix product of nh*stepm elementary matrices
739: and the contribution of each individual to the likelihood is simply
740: hPijx.
741:
742: Also this programme outputs the covariance matrix of the parameters but also
1.218 brouard 743: of the life expectancies. It also computes the period (stable) prevalence.
744:
745: Back prevalence and projections:
1.227 brouard 746:
747: - back_prevalence_limit(double *p, double **bprlim, double ageminpar,
748: double agemaxpar, double ftolpl, int *ncvyearp, double
749: dateprev1,double dateprev2, int firstpass, int lastpass, int
750: mobilavproj)
751:
752: Computes the back prevalence limit for any combination of
753: covariate values k at any age between ageminpar and agemaxpar and
754: returns it in **bprlim. In the loops,
755:
756: - **bprevalim(**bprlim, ***mobaverage, nlstate, *p, age, **oldm,
757: **savm, **dnewm, **doldm, **dsavm, ftolpl, ncvyearp, k);
758:
759: - hBijx Back Probability to be in state i at age x-h being in j at x
1.218 brouard 760: Computes for any combination of covariates k and any age between bage and fage
761: p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
762: oldm=oldms;savm=savms;
1.227 brouard 763:
764: - hbxij(p3mat,nhstepm,agedeb,hstepm,p,nlstate,stepm,oldm,savm, k);
1.218 brouard 765: Computes the transition matrix starting at age 'age' over
766: 'nhstepm*hstepm*stepm' months (i.e. until
767: age (in years) age+nhstepm*hstepm*stepm/12) by multiplying
1.227 brouard 768: nhstepm*hstepm matrices.
769:
770: Returns p3mat[i][j][h] after calling
771: p3mat[i][j][h]=matprod2(newm,
772: bmij(pmmij,cov,ncovmodel,x,nlstate,prevacurrent, dnewm, doldm,
773: dsavm,ij),\ 1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath,
774: oldm);
1.226 brouard 775:
776: Important routines
777:
778: - func (or funcone), computes logit (pij) distinguishing
779: o fixed variables (single or product dummies or quantitative);
780: o varying variables by:
781: (1) wave (single, product dummies, quantitative),
782: (2) by age (can be month) age (done), age*age (done), age*Vn where Vn can be:
783: % fixed dummy (treated) or quantitative (not done because time-consuming);
784: % varying dummy (not done) or quantitative (not done);
785: - Tricode which tests the modality of dummy variables (in order to warn with wrong or empty modalities)
786: and returns the number of efficient covariates cptcoveff and modalities nbcode[Tvar[k]][1]= 0 and nbcode[Tvar[k]][2]= 1 usually.
787: - printinghtml which outputs results like life expectancy in and from a state for a combination of modalities of dummy variables
788: o There are 2*cptcoveff combinations of (0,1) for cptcoveff variables. Outputting only combinations with people, éliminating 1 1 if
789: race White (0 0), Black vs White (1 0), Hispanic (0 1) and 1 1 being meaningless.
1.218 brouard 790:
1.226 brouard 791:
792:
1.133 brouard 793: Authors: Nicolas Brouard (brouard@ined.fr) and Agnès Lièvre (lievre@ined.fr).
794: Institut national d'études démographiques, Paris.
1.126 brouard 795: This software have been partly granted by Euro-REVES, a concerted action
796: from the European Union.
797: It is copyrighted identically to a GNU software product, ie programme and
798: software can be distributed freely for non commercial use. Latest version
799: can be accessed at http://euroreves.ined.fr/imach .
800:
801: Help to debug: LD_PRELOAD=/usr/local/lib/libnjamd.so ./imach foo.imach
802: or better on gdb : set env LD_PRELOAD=/usr/local/lib/libnjamd.so
803:
804: **********************************************************************/
805: /*
806: main
807: read parameterfile
808: read datafile
809: concatwav
810: freqsummary
811: if (mle >= 1)
812: mlikeli
813: print results files
814: if mle==1
815: computes hessian
816: read end of parameter file: agemin, agemax, bage, fage, estepm
817: begin-prev-date,...
818: open gnuplot file
819: open html file
1.145 brouard 820: period (stable) prevalence | pl_nom 1-1 2-2 etc by covariate
821: for age prevalim() | #****** V1=0 V2=1 V3=1 V4=0 ******
822: | 65 1 0 2 1 3 1 4 0 0.96326 0.03674
823: freexexit2 possible for memory heap.
824:
825: h Pij x | pij_nom ficrestpij
826: # Cov Agex agex+h hpijx with i,j= 1-1 1-2 1-3 2-1 2-2 2-3
827: 1 85 85 1.00000 0.00000 0.00000 0.00000 1.00000 0.00000
828: 1 85 86 0.68299 0.22291 0.09410 0.71093 0.00000 0.28907
829:
830: 1 65 99 0.00364 0.00322 0.99314 0.00350 0.00310 0.99340
831: 1 65 100 0.00214 0.00204 0.99581 0.00206 0.00196 0.99597
832: variance of p one-step probabilities varprob | prob_nom ficresprob #One-step probabilities and stand. devi in ()
833: Standard deviation of one-step probabilities | probcor_nom ficresprobcor #One-step probabilities and correlation matrix
834: Matrix of variance covariance of one-step probabilities | probcov_nom ficresprobcov #One-step probabilities and covariance matrix
835:
1.126 brouard 836: forecasting if prevfcast==1 prevforecast call prevalence()
837: health expectancies
838: Variance-covariance of DFLE
839: prevalence()
840: movingaverage()
841: varevsij()
842: if popbased==1 varevsij(,popbased)
843: total life expectancies
844: Variance of period (stable) prevalence
845: end
846: */
847:
1.187 brouard 848: /* #define DEBUG */
849: /* #define DEBUGBRENT */
1.203 brouard 850: /* #define DEBUGLINMIN */
851: /* #define DEBUGHESS */
852: #define DEBUGHESSIJ
1.224 brouard 853: /* #define LINMINORIGINAL /\* Don't use loop on scale in linmin (accepting nan) *\/ */
1.165 brouard 854: #define POWELL /* Instead of NLOPT */
1.224 brouard 855: #define POWELLNOF3INFF1TEST /* Skip test */
1.186 brouard 856: /* #define POWELLORIGINAL /\* Don't use Directest to decide new direction but original Powell test *\/ */
857: /* #define MNBRAKORIGINAL /\* Don't use mnbrak fix *\/ */
1.126 brouard 858:
859: #include <math.h>
860: #include <stdio.h>
861: #include <stdlib.h>
862: #include <string.h>
1.226 brouard 863: #include <ctype.h>
1.159 brouard 864:
865: #ifdef _WIN32
866: #include <io.h>
1.172 brouard 867: #include <windows.h>
868: #include <tchar.h>
1.159 brouard 869: #else
1.126 brouard 870: #include <unistd.h>
1.159 brouard 871: #endif
1.126 brouard 872:
873: #include <limits.h>
874: #include <sys/types.h>
1.171 brouard 875:
876: #if defined(__GNUC__)
877: #include <sys/utsname.h> /* Doesn't work on Windows */
878: #endif
879:
1.126 brouard 880: #include <sys/stat.h>
881: #include <errno.h>
1.159 brouard 882: /* extern int errno; */
1.126 brouard 883:
1.157 brouard 884: /* #ifdef LINUX */
885: /* #include <time.h> */
886: /* #include "timeval.h" */
887: /* #else */
888: /* #include <sys/time.h> */
889: /* #endif */
890:
1.126 brouard 891: #include <time.h>
892:
1.136 brouard 893: #ifdef GSL
894: #include <gsl/gsl_errno.h>
895: #include <gsl/gsl_multimin.h>
896: #endif
897:
1.167 brouard 898:
1.162 brouard 899: #ifdef NLOPT
900: #include <nlopt.h>
901: typedef struct {
902: double (* function)(double [] );
903: } myfunc_data ;
904: #endif
905:
1.126 brouard 906: /* #include <libintl.h> */
907: /* #define _(String) gettext (String) */
908:
1.141 brouard 909: #define MAXLINE 1024 /* Was 256. Overflow with 312 with 2 states and 4 covariates. Should be ok */
1.126 brouard 910:
911: #define GNUPLOTPROGRAM "gnuplot"
912: /*#define GNUPLOTPROGRAM "..\\gp37mgw\\wgnuplot"*/
913: #define FILENAMELENGTH 132
914:
915: #define GLOCK_ERROR_NOPATH -1 /* empty path */
916: #define GLOCK_ERROR_GETCWD -2 /* cannot get cwd */
917:
1.144 brouard 918: #define MAXPARM 128 /**< Maximum number of parameters for the optimization */
919: #define NPARMAX 64 /**< (nlstate+ndeath-1)*nlstate*ncovmodel */
1.126 brouard 920:
921: #define NINTERVMAX 8
1.144 brouard 922: #define NLSTATEMAX 8 /**< Maximum number of live states (for func) */
923: #define NDEATHMAX 8 /**< Maximum number of dead states (for func) */
924: #define NCOVMAX 20 /**< Maximum number of covariates, including generated covariates V1*V2 */
1.197 brouard 925: #define codtabm(h,k) (1 & (h-1) >> (k-1))+1
1.211 brouard 926: /*#define decodtabm(h,k,cptcoveff)= (h <= (1<<cptcoveff)?(((h-1) >> (k-1)) & 1) +1 : -1)*/
927: #define decodtabm(h,k,cptcoveff) (((h-1) >> (k-1)) & 1) +1
1.126 brouard 928: #define MAXN 20000
1.144 brouard 929: #define YEARM 12. /**< Number of months per year */
1.218 brouard 930: /* #define AGESUP 130 */
931: #define AGESUP 150
932: #define AGEMARGE 25 /* Marge for agemin and agemax for(iage=agemin-AGEMARGE; iage <= agemax+3+AGEMARGE; iage++) */
1.126 brouard 933: #define AGEBASE 40
1.194 brouard 934: #define AGEOVERFLOW 1.e20
1.164 brouard 935: #define AGEGOMP 10 /**< Minimal age for Gompertz adjustment */
1.157 brouard 936: #ifdef _WIN32
937: #define DIRSEPARATOR '\\'
938: #define CHARSEPARATOR "\\"
939: #define ODIRSEPARATOR '/'
940: #else
1.126 brouard 941: #define DIRSEPARATOR '/'
942: #define CHARSEPARATOR "/"
943: #define ODIRSEPARATOR '\\'
944: #endif
945:
1.248 ! brouard 946: /* $Id: imach.c,v 1.247 2016/09/02 11:11:21 brouard Exp $ */
1.126 brouard 947: /* $State: Exp $ */
1.196 brouard 948: #include "version.h"
949: char version[]=__IMACH_VERSION__;
1.224 brouard 950: 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";
1.248 ! brouard 951: char fullversion[]="$Revision: 1.247 $ $Date: 2016/09/02 11:11:21 $";
1.126 brouard 952: char strstart[80];
953: char optionfilext[10], optionfilefiname[FILENAMELENGTH];
1.130 brouard 954: int erreur=0, nberr=0, nbwarn=0; /* Error number, number of errors number of warnings */
1.187 brouard 955: int nagesqr=0, nforce=0; /* nagesqr=1 if model is including age*age, number of forces */
1.145 brouard 956: /* Number of covariates model=V2+V1+ V3*age+V2*V4 */
957: int cptcovn=0; /**< cptcovn number of covariates added in the model (excepting constant and age and age*product) */
958: int cptcovt=0; /**< cptcovt number of covariates added in the model (excepting constant and age) */
1.225 brouard 959: int cptcovs=0; /**< cptcovs number of simple covariates in the model V2+V1 =2 */
960: int cptcovsnq=0; /**< cptcovsnq number of simple covariates in the model but non quantitative V2+V1 =2 */
1.145 brouard 961: int cptcovage=0; /**< Number of covariates with age: V3*age only =1 */
962: int cptcovprodnoage=0; /**< Number of covariate products without age */
963: int cptcoveff=0; /* Total number of covariates to vary for printing results */
1.233 brouard 964: int ncovf=0; /* Total number of effective fixed covariates (dummy or quantitative) in the model */
965: int ncovv=0; /* Total number of effective (wave) varying covariates (dummy or quantitative) in the model */
1.232 brouard 966: int ncova=0; /* Total number of effective (wave and stepm) varying with age covariates (dummy of quantitative) in the model */
1.234 brouard 967: int nsd=0; /**< Total number of single dummy variables (output) */
968: int nsq=0; /**< Total number of single quantitative variables (output) */
1.232 brouard 969: int ncoveff=0; /* Total number of effective fixed dummy covariates in the model */
1.225 brouard 970: int nqfveff=0; /**< nqfveff Number of Quantitative Fixed Variables Effective */
1.224 brouard 971: int ntveff=0; /**< ntveff number of effective time varying variables */
972: int nqtveff=0; /**< ntqveff number of effective time varying quantitative variables */
1.145 brouard 973: int cptcov=0; /* Working variable */
1.218 brouard 974: int ncovcombmax=NCOVMAX; /* Maximum calculated number of covariate combination = pow(2, cptcoveff) */
1.126 brouard 975: int npar=NPARMAX;
976: int nlstate=2; /* Number of live states */
977: int ndeath=1; /* Number of dead states */
1.130 brouard 978: int ncovmodel=0, ncovcol=0; /* Total number of covariables including constant a12*1 +b12*x ncovmodel=2 */
1.223 brouard 979: int nqv=0, ntv=0, nqtv=0; /* Total number of quantitative variables, time variable (dummy), quantitative and time variable */
1.126 brouard 980: int popbased=0;
981:
982: int *wav; /* Number of waves for this individuual 0 is possible */
1.130 brouard 983: int maxwav=0; /* Maxim number of waves */
984: int jmin=0, jmax=0; /* min, max spacing between 2 waves */
985: int ijmin=0, ijmax=0; /* Individuals having jmin and jmax */
986: int gipmx=0, gsw=0; /* Global variables on the number of contributions
1.126 brouard 987: to the likelihood and the sum of weights (done by funcone)*/
1.130 brouard 988: int mle=1, weightopt=0;
1.126 brouard 989: int **mw; /* mw[mi][i] is number of the mi wave for this individual */
990: int **dh; /* dh[mi][i] is number of steps between mi,mi+1 for this individual */
991: int **bh; /* bh[mi][i] is the bias (+ or -) for this individual if the delay between
992: * wave mi and wave mi+1 is not an exact multiple of stepm. */
1.162 brouard 993: int countcallfunc=0; /* Count the number of calls to func */
1.230 brouard 994: int selected(int kvar); /* Is covariate kvar selected for printing results */
995:
1.130 brouard 996: double jmean=1; /* Mean space between 2 waves */
1.145 brouard 997: double **matprod2(); /* test */
1.126 brouard 998: double **oldm, **newm, **savm; /* Working pointers to matrices */
999: double **oldms, **newms, **savms; /* Fixed working pointers to matrices */
1.218 brouard 1000: double **ddnewms, **ddoldms, **ddsavms; /* for freeing later */
1001:
1.136 brouard 1002: /*FILE *fic ; */ /* Used in readdata only */
1.217 brouard 1003: FILE *ficpar, *ficparo,*ficres, *ficresp, *ficresphtm, *ficresphtmfr, *ficrespl, *ficresplb,*ficrespij, *ficrespijb, *ficrest,*ficresf, *ficresfb,*ficrespop;
1.126 brouard 1004: FILE *ficlog, *ficrespow;
1.130 brouard 1005: int globpr=0; /* Global variable for printing or not */
1.126 brouard 1006: double fretone; /* Only one call to likelihood */
1.130 brouard 1007: long ipmx=0; /* Number of contributions */
1.126 brouard 1008: double sw; /* Sum of weights */
1009: char filerespow[FILENAMELENGTH];
1010: char fileresilk[FILENAMELENGTH]; /* File of individual contributions to the likelihood */
1011: FILE *ficresilk;
1012: FILE *ficgp,*ficresprob,*ficpop, *ficresprobcov, *ficresprobcor;
1013: FILE *ficresprobmorprev;
1014: FILE *fichtm, *fichtmcov; /* Html File */
1015: FILE *ficreseij;
1016: char filerese[FILENAMELENGTH];
1017: FILE *ficresstdeij;
1018: char fileresstde[FILENAMELENGTH];
1019: FILE *ficrescveij;
1020: char filerescve[FILENAMELENGTH];
1021: FILE *ficresvij;
1022: char fileresv[FILENAMELENGTH];
1023: FILE *ficresvpl;
1024: char fileresvpl[FILENAMELENGTH];
1025: char title[MAXLINE];
1.234 brouard 1026: char model[MAXLINE]; /**< The model line */
1.217 brouard 1027: char optionfile[FILENAMELENGTH], datafile[FILENAMELENGTH], filerespl[FILENAMELENGTH], fileresplb[FILENAMELENGTH];
1.126 brouard 1028: char plotcmd[FILENAMELENGTH], pplotcmd[FILENAMELENGTH];
1029: char tmpout[FILENAMELENGTH], tmpout2[FILENAMELENGTH];
1030: char command[FILENAMELENGTH];
1031: int outcmd=0;
1032:
1.217 brouard 1033: char fileres[FILENAMELENGTH], filerespij[FILENAMELENGTH], filerespijb[FILENAMELENGTH], filereso[FILENAMELENGTH], rfileres[FILENAMELENGTH];
1.202 brouard 1034: char fileresu[FILENAMELENGTH]; /* fileres without r in front */
1.126 brouard 1035: char filelog[FILENAMELENGTH]; /* Log file */
1036: char filerest[FILENAMELENGTH];
1037: char fileregp[FILENAMELENGTH];
1038: char popfile[FILENAMELENGTH];
1039:
1040: char optionfilegnuplot[FILENAMELENGTH], optionfilehtm[FILENAMELENGTH], optionfilehtmcov[FILENAMELENGTH] ;
1041:
1.157 brouard 1042: /* struct timeval start_time, end_time, curr_time, last_time, forecast_time; */
1043: /* struct timezone tzp; */
1044: /* extern int gettimeofday(); */
1045: struct tm tml, *gmtime(), *localtime();
1046:
1047: extern time_t time();
1048:
1049: struct tm start_time, end_time, curr_time, last_time, forecast_time;
1050: time_t rstart_time, rend_time, rcurr_time, rlast_time, rforecast_time; /* raw time */
1051: struct tm tm;
1052:
1.126 brouard 1053: char strcurr[80], strfor[80];
1054:
1055: char *endptr;
1056: long lval;
1057: double dval;
1058:
1059: #define NR_END 1
1060: #define FREE_ARG char*
1061: #define FTOL 1.0e-10
1062:
1063: #define NRANSI
1.240 brouard 1064: #define ITMAX 200
1065: #define ITPOWMAX 20 /* This is now multiplied by the number of parameters */
1.126 brouard 1066:
1067: #define TOL 2.0e-4
1068:
1069: #define CGOLD 0.3819660
1070: #define ZEPS 1.0e-10
1071: #define SHFT(a,b,c,d) (a)=(b);(b)=(c);(c)=(d);
1072:
1073: #define GOLD 1.618034
1074: #define GLIMIT 100.0
1075: #define TINY 1.0e-20
1076:
1077: static double maxarg1,maxarg2;
1078: #define FMAX(a,b) (maxarg1=(a),maxarg2=(b),(maxarg1)>(maxarg2)? (maxarg1):(maxarg2))
1079: #define FMIN(a,b) (maxarg1=(a),maxarg2=(b),(maxarg1)<(maxarg2)? (maxarg1):(maxarg2))
1080:
1081: #define SIGN(a,b) ((b)>0.0 ? fabs(a) : -fabs(a))
1082: #define rint(a) floor(a+0.5)
1.166 brouard 1083: /* http://www.thphys.uni-heidelberg.de/~robbers/cmbeasy/doc/html/myutils_8h-source.html */
1.183 brouard 1084: #define mytinydouble 1.0e-16
1.166 brouard 1085: /* #define DEQUAL(a,b) (fabs((a)-(b))<mytinydouble) */
1086: /* http://www.thphys.uni-heidelberg.de/~robbers/cmbeasy/doc/html/mynrutils_8h-source.html */
1087: /* static double dsqrarg; */
1088: /* #define DSQR(a) (DEQUAL((dsqrarg=(a)),0.0) ? 0.0 : dsqrarg*dsqrarg) */
1.126 brouard 1089: static double sqrarg;
1090: #define SQR(a) ((sqrarg=(a)) == 0.0 ? 0.0 :sqrarg*sqrarg)
1091: #define SWAP(a,b) {temp=(a);(a)=(b);(b)=temp;}
1092: int agegomp= AGEGOMP;
1093:
1094: int imx;
1095: int stepm=1;
1096: /* Stepm, step in month: minimum step interpolation*/
1097:
1098: int estepm;
1099: /* Estepm, step in month to interpolate survival function in order to approximate Life Expectancy*/
1100:
1101: int m,nb;
1102: long *num;
1.197 brouard 1103: int firstpass=0, lastpass=4,*cod, *cens;
1.192 brouard 1104: int *ncodemax; /* ncodemax[j]= Number of modalities of the j th
1105: covariate for which somebody answered excluding
1106: undefined. Usually 2: 0 and 1. */
1107: int *ncodemaxwundef; /* ncodemax[j]= Number of modalities of the j th
1108: covariate for which somebody answered including
1109: undefined. Usually 3: -1, 0 and 1. */
1.126 brouard 1110: double **agev,*moisnais, *annais, *moisdc, *andc,**mint, **anint;
1.218 brouard 1111: double **pmmij, ***probs; /* Global pointer */
1.219 brouard 1112: double ***mobaverage, ***mobaverages; /* New global variable */
1.126 brouard 1113: double *ageexmed,*agecens;
1114: double dateintmean=0;
1115:
1116: double *weight;
1117: int **s; /* Status */
1.141 brouard 1118: double *agedc;
1.145 brouard 1119: double **covar; /**< covar[j,i], value of jth covariate for individual i,
1.141 brouard 1120: * covar=matrix(0,NCOVMAX,1,n);
1.187 brouard 1121: * cov[Tage[kk]+2]=covar[Tvar[Tage[kk]]][i]*age; */
1.225 brouard 1122: double **coqvar; /* Fixed quantitative covariate iqv */
1123: double ***cotvar; /* Time varying covariate itv */
1124: double ***cotqvar; /* Time varying quantitative covariate itqv */
1.141 brouard 1125: double idx;
1126: int **nbcode, *Tvar; /**< model=V2 => Tvar[1]= 2 */
1.234 brouard 1127: /* V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1128: /*k 1 2 3 4 5 6 7 8 9 */
1129: /*Tvar[k]= 5 4 3 6 5 2 7 1 1 */
1130: /* Tndvar[k] 1 2 3 4 5 */
1131: /*TDvar 4 3 6 7 1 */ /* For outputs only; combination of dummies fixed or varying */
1132: /* Tns[k] 1 2 2 4 5 */ /* Number of single cova */
1133: /* TvarsD[k] 1 2 3 */ /* Number of single dummy cova */
1134: /* TvarsDind 2 3 9 */ /* position K of single dummy cova */
1135: /* TvarsQ[k] 1 2 */ /* Number of single quantitative cova */
1136: /* TvarsQind 1 6 */ /* position K of single quantitative cova */
1137: /* Tprod[i]=k 4 7 */
1138: /* Tage[i]=k 5 8 */
1139: /* */
1140: /* Type */
1141: /* V 1 2 3 4 5 */
1142: /* F F V V V */
1143: /* D Q D D Q */
1144: /* */
1145: int *TvarsD;
1146: int *TvarsDind;
1147: int *TvarsQ;
1148: int *TvarsQind;
1149:
1.235 brouard 1150: #define MAXRESULTLINES 10
1151: int nresult=0;
1152: int TKresult[MAXRESULTLINES];
1.237 brouard 1153: int Tresult[MAXRESULTLINES][NCOVMAX];/* For dummy variable , value (output) */
1154: int Tinvresult[MAXRESULTLINES][NCOVMAX];/* For dummy variable , value (output) */
1.235 brouard 1155: int Tvresult[MAXRESULTLINES][NCOVMAX]; /* For dummy variable , variable # (output) */
1156: double Tqresult[MAXRESULTLINES][NCOVMAX]; /* For quantitative variable , value (output) */
1.237 brouard 1157: double Tqinvresult[MAXRESULTLINES][NCOVMAX]; /* For quantitative variable , value (output) */
1.235 brouard 1158: int Tvqresult[MAXRESULTLINES][NCOVMAX]; /* For quantitative variable , variable # (output) */
1159:
1.234 brouard 1160: /* 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 1161: 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 */
1162: 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 */
1163: 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 */
1164: int *TvarVind; /**< TvarVind[1]=1, TvarVind[2]=2 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1165: 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 */
1166: 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 1167: int *TvarFD; /**< TvarFD[1]=V1 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1168: int *TvarFDind; /* TvarFDind[1]=9 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1169: int *TvarFQ; /* TvarFQ[1]=V2 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple fixed quantitative variable */
1170: int *TvarFQind; /* TvarFQind[1]=6 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple fixed quantitative variable */
1171: int *TvarVD; /* TvarVD[1]=V5 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple fixed quantitative variable */
1172: int *TvarVDind; /* TvarVDind[1]=1 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */ /* Only simple fixed quantitative variable */
1173: 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 */
1174: 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 */
1175:
1.230 brouard 1176: int *Tvarsel; /**< Selected covariates for output */
1177: double *Tvalsel; /**< Selected modality value of covariate for output */
1.226 brouard 1178: int *Typevar; /**< 0 for simple covariate (dummy, quantitative, fixed or varying), 1 for age product, 2 for product */
1.227 brouard 1179: int *Fixed; /** Fixed[k] 0=fixed, 1 varying, 2 fixed with age product, 3 varying with age product */
1180: 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 1181: int *DummyV; /** Dummy[v] 0=dummy (0 1), 1 quantitative */
1182: int *FixedV; /** FixedV[v] 0 fixed, 1 varying */
1.197 brouard 1183: int *Tage;
1.227 brouard 1184: int anyvaryingduminmodel=0; /**< Any varying dummy in Model=1 yes, 0 no, to avoid a loop on waves in freq */
1.228 brouard 1185: 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 1186: 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*/
1187: 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 1188: int *Ndum; /** Freq of modality (tricode */
1.200 brouard 1189: /* int **codtab;*/ /**< codtab=imatrix(1,100,1,10); */
1.227 brouard 1190: int **Tvard;
1191: int *Tprod;/**< Gives the k position of the k1 product */
1.238 brouard 1192: /* Tprod[k1=1]=3(=V1*V4) for V2+V1+V1*V4+age*V3 */
1.227 brouard 1193: int *Tposprod; /**< Gives the k1 product from the k position */
1.238 brouard 1194: /* if V2+V1+V1*V4+age*V3+V3*V2 TProd[k1=2]=5 (V3*V2) */
1195: /* Tposprod[k]=k1 , Tposprod[3]=1, Tposprod[5(V3*V2)]=2 (2nd product without age) */
1.227 brouard 1196: int cptcovprod, *Tvaraff, *invalidvarcomb;
1.126 brouard 1197: double *lsurv, *lpop, *tpop;
1198:
1.231 brouard 1199: #define FD 1; /* Fixed dummy covariate */
1200: #define FQ 2; /* Fixed quantitative covariate */
1201: #define FP 3; /* Fixed product covariate */
1202: #define FPDD 7; /* Fixed product dummy*dummy covariate */
1203: #define FPDQ 8; /* Fixed product dummy*quantitative covariate */
1204: #define FPQQ 9; /* Fixed product quantitative*quantitative covariate */
1205: #define VD 10; /* Varying dummy covariate */
1206: #define VQ 11; /* Varying quantitative covariate */
1207: #define VP 12; /* Varying product covariate */
1208: #define VPDD 13; /* Varying product dummy*dummy covariate */
1209: #define VPDQ 14; /* Varying product dummy*quantitative covariate */
1210: #define VPQQ 15; /* Varying product quantitative*quantitative covariate */
1211: #define APFD 16; /* Age product * fixed dummy covariate */
1212: #define APFQ 17; /* Age product * fixed quantitative covariate */
1213: #define APVD 18; /* Age product * varying dummy covariate */
1214: #define APVQ 19; /* Age product * varying quantitative covariate */
1215:
1216: #define FTYPE 1; /* Fixed covariate */
1217: #define VTYPE 2; /* Varying covariate (loop in wave) */
1218: #define ATYPE 2; /* Age product covariate (loop in dh within wave)*/
1219:
1220: struct kmodel{
1221: int maintype; /* main type */
1222: int subtype; /* subtype */
1223: };
1224: struct kmodel modell[NCOVMAX];
1225:
1.143 brouard 1226: double ftol=FTOL; /**< Tolerance for computing Max Likelihood */
1227: double ftolhess; /**< Tolerance for computing hessian */
1.126 brouard 1228:
1229: /**************** split *************************/
1230: static int split( char *path, char *dirc, char *name, char *ext, char *finame )
1231: {
1232: /* From a file name with (full) path (either Unix or Windows) we extract the directory (dirc)
1233: the name of the file (name), its extension only (ext) and its first part of the name (finame)
1234: */
1235: char *ss; /* pointer */
1.186 brouard 1236: int l1=0, l2=0; /* length counters */
1.126 brouard 1237:
1238: l1 = strlen(path ); /* length of path */
1239: if ( l1 == 0 ) return( GLOCK_ERROR_NOPATH );
1240: ss= strrchr( path, DIRSEPARATOR ); /* find last / */
1241: if ( ss == NULL ) { /* no directory, so determine current directory */
1242: strcpy( name, path ); /* we got the fullname name because no directory */
1243: /*if(strrchr(path, ODIRSEPARATOR )==NULL)
1244: printf("Warning you should use %s as a separator\n",DIRSEPARATOR);*/
1245: /* get current working directory */
1246: /* extern char* getcwd ( char *buf , int len);*/
1.184 brouard 1247: #ifdef WIN32
1248: if (_getcwd( dirc, FILENAME_MAX ) == NULL ) {
1249: #else
1250: if (getcwd(dirc, FILENAME_MAX) == NULL) {
1251: #endif
1.126 brouard 1252: return( GLOCK_ERROR_GETCWD );
1253: }
1254: /* got dirc from getcwd*/
1255: printf(" DIRC = %s \n",dirc);
1.205 brouard 1256: } else { /* strip directory from path */
1.126 brouard 1257: ss++; /* after this, the filename */
1258: l2 = strlen( ss ); /* length of filename */
1259: if ( l2 == 0 ) return( GLOCK_ERROR_NOPATH );
1260: strcpy( name, ss ); /* save file name */
1261: strncpy( dirc, path, l1 - l2 ); /* now the directory */
1.186 brouard 1262: dirc[l1-l2] = '\0'; /* add zero */
1.126 brouard 1263: printf(" DIRC2 = %s \n",dirc);
1264: }
1265: /* We add a separator at the end of dirc if not exists */
1266: l1 = strlen( dirc ); /* length of directory */
1267: if( dirc[l1-1] != DIRSEPARATOR ){
1268: dirc[l1] = DIRSEPARATOR;
1269: dirc[l1+1] = 0;
1270: printf(" DIRC3 = %s \n",dirc);
1271: }
1272: ss = strrchr( name, '.' ); /* find last / */
1273: if (ss >0){
1274: ss++;
1275: strcpy(ext,ss); /* save extension */
1276: l1= strlen( name);
1277: l2= strlen(ss)+1;
1278: strncpy( finame, name, l1-l2);
1279: finame[l1-l2]= 0;
1280: }
1281:
1282: return( 0 ); /* we're done */
1283: }
1284:
1285:
1286: /******************************************/
1287:
1288: void replace_back_to_slash(char *s, char*t)
1289: {
1290: int i;
1291: int lg=0;
1292: i=0;
1293: lg=strlen(t);
1294: for(i=0; i<= lg; i++) {
1295: (s[i] = t[i]);
1296: if (t[i]== '\\') s[i]='/';
1297: }
1298: }
1299:
1.132 brouard 1300: char *trimbb(char *out, char *in)
1.137 brouard 1301: { /* Trim multiple blanks in line but keeps first blanks if line starts with blanks */
1.132 brouard 1302: char *s;
1303: s=out;
1304: while (*in != '\0'){
1.137 brouard 1305: while( *in == ' ' && *(in+1) == ' '){ /* && *(in+1) != '\0'){*/
1.132 brouard 1306: in++;
1307: }
1308: *out++ = *in++;
1309: }
1310: *out='\0';
1311: return s;
1312: }
1313:
1.187 brouard 1314: /* char *substrchaine(char *out, char *in, char *chain) */
1315: /* { */
1316: /* /\* Substract chain 'chain' from 'in', return and output 'out' *\/ */
1317: /* char *s, *t; */
1318: /* t=in;s=out; */
1319: /* while ((*in != *chain) && (*in != '\0')){ */
1320: /* *out++ = *in++; */
1321: /* } */
1322:
1323: /* /\* *in matches *chain *\/ */
1324: /* while ((*in++ == *chain++) && (*in != '\0')){ */
1325: /* printf("*in = %c, *out= %c *chain= %c \n", *in, *out, *chain); */
1326: /* } */
1327: /* in--; chain--; */
1328: /* while ( (*in != '\0')){ */
1329: /* printf("Bef *in = %c, *out= %c *chain= %c \n", *in, *out, *chain); */
1330: /* *out++ = *in++; */
1331: /* printf("Aft *in = %c, *out= %c *chain= %c \n", *in, *out, *chain); */
1332: /* } */
1333: /* *out='\0'; */
1334: /* out=s; */
1335: /* return out; */
1336: /* } */
1337: char *substrchaine(char *out, char *in, char *chain)
1338: {
1339: /* Substract chain 'chain' from 'in', return and output 'out' */
1340: /* in="V1+V1*age+age*age+V2", chain="age*age" */
1341:
1342: char *strloc;
1343:
1344: strcpy (out, in);
1345: strloc = strstr(out, chain); /* strloc points to out at age*age+V2 */
1346: printf("Bef strloc=%s chain=%s out=%s \n", strloc, chain, out);
1347: if(strloc != NULL){
1348: /* will affect out */ /* strloc+strlenc(chain)=+V2 */ /* Will also work in Unicode */
1349: memmove(strloc,strloc+strlen(chain), strlen(strloc+strlen(chain))+1);
1350: /* strcpy (strloc, strloc +strlen(chain));*/
1351: }
1352: printf("Aft strloc=%s chain=%s in=%s out=%s \n", strloc, chain, in, out);
1353: return out;
1354: }
1355:
1356:
1.145 brouard 1357: char *cutl(char *blocc, char *alocc, char *in, char occ)
1358: {
1.187 brouard 1359: /* cuts string in into blocc and alocc where blocc ends before FIRST occurence of char 'occ'
1.145 brouard 1360: and alocc starts after first occurence of char 'occ' : ex cutv(blocc,alocc,"abcdef2ghi2j",'2')
1.187 brouard 1361: gives blocc="abcdef" and alocc="ghi2j".
1.145 brouard 1362: If occ is not found blocc is null and alocc is equal to in. Returns blocc
1363: */
1.160 brouard 1364: char *s, *t;
1.145 brouard 1365: t=in;s=in;
1366: while ((*in != occ) && (*in != '\0')){
1367: *alocc++ = *in++;
1368: }
1369: if( *in == occ){
1370: *(alocc)='\0';
1371: s=++in;
1372: }
1373:
1374: if (s == t) {/* occ not found */
1375: *(alocc-(in-s))='\0';
1376: in=s;
1377: }
1378: while ( *in != '\0'){
1379: *blocc++ = *in++;
1380: }
1381:
1382: *blocc='\0';
1383: return t;
1384: }
1.137 brouard 1385: char *cutv(char *blocc, char *alocc, char *in, char occ)
1386: {
1.187 brouard 1387: /* cuts string in into blocc and alocc where blocc ends before LAST occurence of char 'occ'
1.137 brouard 1388: and alocc starts after last occurence of char 'occ' : ex cutv(blocc,alocc,"abcdef2ghi2j",'2')
1389: gives blocc="abcdef2ghi" and alocc="j".
1390: If occ is not found blocc is null and alocc is equal to in. Returns alocc
1391: */
1392: char *s, *t;
1393: t=in;s=in;
1394: while (*in != '\0'){
1395: while( *in == occ){
1396: *blocc++ = *in++;
1397: s=in;
1398: }
1399: *blocc++ = *in++;
1400: }
1401: if (s == t) /* occ not found */
1402: *(blocc-(in-s))='\0';
1403: else
1404: *(blocc-(in-s)-1)='\0';
1405: in=s;
1406: while ( *in != '\0'){
1407: *alocc++ = *in++;
1408: }
1409:
1410: *alocc='\0';
1411: return s;
1412: }
1413:
1.126 brouard 1414: int nbocc(char *s, char occ)
1415: {
1416: int i,j=0;
1417: int lg=20;
1418: i=0;
1419: lg=strlen(s);
1420: for(i=0; i<= lg; i++) {
1.234 brouard 1421: if (s[i] == occ ) j++;
1.126 brouard 1422: }
1423: return j;
1424: }
1425:
1.137 brouard 1426: /* void cutv(char *u,char *v, char*t, char occ) */
1427: /* { */
1428: /* /\* cuts string t into u and v where u ends before last occurence of char 'occ' */
1429: /* and v starts after last occurence of char 'occ' : ex cutv(u,v,"abcdef2ghi2j",'2') */
1430: /* gives u="abcdef2ghi" and v="j" *\/ */
1431: /* int i,lg,j,p=0; */
1432: /* i=0; */
1433: /* lg=strlen(t); */
1434: /* for(j=0; j<=lg-1; j++) { */
1435: /* if((t[j]!= occ) && (t[j+1]== occ)) p=j+1; */
1436: /* } */
1.126 brouard 1437:
1.137 brouard 1438: /* for(j=0; j<p; j++) { */
1439: /* (u[j] = t[j]); */
1440: /* } */
1441: /* u[p]='\0'; */
1.126 brouard 1442:
1.137 brouard 1443: /* for(j=0; j<= lg; j++) { */
1444: /* if (j>=(p+1))(v[j-p-1] = t[j]); */
1445: /* } */
1446: /* } */
1.126 brouard 1447:
1.160 brouard 1448: #ifdef _WIN32
1449: char * strsep(char **pp, const char *delim)
1450: {
1451: char *p, *q;
1452:
1453: if ((p = *pp) == NULL)
1454: return 0;
1455: if ((q = strpbrk (p, delim)) != NULL)
1456: {
1457: *pp = q + 1;
1458: *q = '\0';
1459: }
1460: else
1461: *pp = 0;
1462: return p;
1463: }
1464: #endif
1465:
1.126 brouard 1466: /********************** nrerror ********************/
1467:
1468: void nrerror(char error_text[])
1469: {
1470: fprintf(stderr,"ERREUR ...\n");
1471: fprintf(stderr,"%s\n",error_text);
1472: exit(EXIT_FAILURE);
1473: }
1474: /*********************** vector *******************/
1475: double *vector(int nl, int nh)
1476: {
1477: double *v;
1478: v=(double *) malloc((size_t)((nh-nl+1+NR_END)*sizeof(double)));
1479: if (!v) nrerror("allocation failure in vector");
1480: return v-nl+NR_END;
1481: }
1482:
1483: /************************ free vector ******************/
1484: void free_vector(double*v, int nl, int nh)
1485: {
1486: free((FREE_ARG)(v+nl-NR_END));
1487: }
1488:
1489: /************************ivector *******************************/
1490: int *ivector(long nl,long nh)
1491: {
1492: int *v;
1493: v=(int *) malloc((size_t)((nh-nl+1+NR_END)*sizeof(int)));
1494: if (!v) nrerror("allocation failure in ivector");
1495: return v-nl+NR_END;
1496: }
1497:
1498: /******************free ivector **************************/
1499: void free_ivector(int *v, long nl, long nh)
1500: {
1501: free((FREE_ARG)(v+nl-NR_END));
1502: }
1503:
1504: /************************lvector *******************************/
1505: long *lvector(long nl,long nh)
1506: {
1507: long *v;
1508: v=(long *) malloc((size_t)((nh-nl+1+NR_END)*sizeof(long)));
1509: if (!v) nrerror("allocation failure in ivector");
1510: return v-nl+NR_END;
1511: }
1512:
1513: /******************free lvector **************************/
1514: void free_lvector(long *v, long nl, long nh)
1515: {
1516: free((FREE_ARG)(v+nl-NR_END));
1517: }
1518:
1519: /******************* imatrix *******************************/
1520: int **imatrix(long nrl, long nrh, long ncl, long nch)
1521: /* allocate a int matrix with subscript range m[nrl..nrh][ncl..nch] */
1522: {
1523: long i, nrow=nrh-nrl+1,ncol=nch-ncl+1;
1524: int **m;
1525:
1526: /* allocate pointers to rows */
1527: m=(int **) malloc((size_t)((nrow+NR_END)*sizeof(int*)));
1528: if (!m) nrerror("allocation failure 1 in matrix()");
1529: m += NR_END;
1530: m -= nrl;
1531:
1532:
1533: /* allocate rows and set pointers to them */
1534: m[nrl]=(int *) malloc((size_t)((nrow*ncol+NR_END)*sizeof(int)));
1535: if (!m[nrl]) nrerror("allocation failure 2 in matrix()");
1536: m[nrl] += NR_END;
1537: m[nrl] -= ncl;
1538:
1539: for(i=nrl+1;i<=nrh;i++) m[i]=m[i-1]+ncol;
1540:
1541: /* return pointer to array of pointers to rows */
1542: return m;
1543: }
1544:
1545: /****************** free_imatrix *************************/
1546: void free_imatrix(m,nrl,nrh,ncl,nch)
1547: int **m;
1548: long nch,ncl,nrh,nrl;
1549: /* free an int matrix allocated by imatrix() */
1550: {
1551: free((FREE_ARG) (m[nrl]+ncl-NR_END));
1552: free((FREE_ARG) (m+nrl-NR_END));
1553: }
1554:
1555: /******************* matrix *******************************/
1556: double **matrix(long nrl, long nrh, long ncl, long nch)
1557: {
1558: long i, nrow=nrh-nrl+1, ncol=nch-ncl+1;
1559: double **m;
1560:
1561: m=(double **) malloc((size_t)((nrow+NR_END)*sizeof(double*)));
1562: if (!m) nrerror("allocation failure 1 in matrix()");
1563: m += NR_END;
1564: m -= nrl;
1565:
1566: m[nrl]=(double *) malloc((size_t)((nrow*ncol+NR_END)*sizeof(double)));
1567: if (!m[nrl]) nrerror("allocation failure 2 in matrix()");
1568: m[nrl] += NR_END;
1569: m[nrl] -= ncl;
1570:
1571: for (i=nrl+1; i<=nrh; i++) m[i]=m[i-1]+ncol;
1572: return m;
1.145 brouard 1573: /* print *(*(m+1)+70) or print m[1][70]; print m+1 or print &(m[1]) or &(m[1][0])
1574: m[i] = address of ith row of the table. &(m[i]) is its value which is another adress
1575: that of m[i][0]. In order to get the value p m[i][0] but it is unitialized.
1.126 brouard 1576: */
1577: }
1578:
1579: /*************************free matrix ************************/
1580: void free_matrix(double **m, long nrl, long nrh, long ncl, long nch)
1581: {
1582: free((FREE_ARG)(m[nrl]+ncl-NR_END));
1583: free((FREE_ARG)(m+nrl-NR_END));
1584: }
1585:
1586: /******************* ma3x *******************************/
1587: double ***ma3x(long nrl, long nrh, long ncl, long nch, long nll, long nlh)
1588: {
1589: long i, j, nrow=nrh-nrl+1, ncol=nch-ncl+1, nlay=nlh-nll+1;
1590: double ***m;
1591:
1592: m=(double ***) malloc((size_t)((nrow+NR_END)*sizeof(double*)));
1593: if (!m) nrerror("allocation failure 1 in matrix()");
1594: m += NR_END;
1595: m -= nrl;
1596:
1597: m[nrl]=(double **) malloc((size_t)((nrow*ncol+NR_END)*sizeof(double)));
1598: if (!m[nrl]) nrerror("allocation failure 2 in matrix()");
1599: m[nrl] += NR_END;
1600: m[nrl] -= ncl;
1601:
1602: for (i=nrl+1; i<=nrh; i++) m[i]=m[i-1]+ncol;
1603:
1604: m[nrl][ncl]=(double *) malloc((size_t)((nrow*ncol*nlay+NR_END)*sizeof(double)));
1605: if (!m[nrl][ncl]) nrerror("allocation failure 3 in matrix()");
1606: m[nrl][ncl] += NR_END;
1607: m[nrl][ncl] -= nll;
1608: for (j=ncl+1; j<=nch; j++)
1609: m[nrl][j]=m[nrl][j-1]+nlay;
1610:
1611: for (i=nrl+1; i<=nrh; i++) {
1612: m[i][ncl]=m[i-1l][ncl]+ncol*nlay;
1613: for (j=ncl+1; j<=nch; j++)
1614: m[i][j]=m[i][j-1]+nlay;
1615: }
1616: return m;
1617: /* gdb: p *(m+1) <=> p m[1] and p (m+1) <=> p (m+1) <=> p &(m[1])
1618: &(m[i][j][k]) <=> *((*(m+i) + j)+k)
1619: */
1620: }
1621:
1622: /*************************free ma3x ************************/
1623: void free_ma3x(double ***m, long nrl, long nrh, long ncl, long nch,long nll, long nlh)
1624: {
1625: free((FREE_ARG)(m[nrl][ncl]+ nll-NR_END));
1626: free((FREE_ARG)(m[nrl]+ncl-NR_END));
1627: free((FREE_ARG)(m+nrl-NR_END));
1628: }
1629:
1630: /*************** function subdirf ***********/
1631: char *subdirf(char fileres[])
1632: {
1633: /* Caution optionfilefiname is hidden */
1634: strcpy(tmpout,optionfilefiname);
1635: strcat(tmpout,"/"); /* Add to the right */
1636: strcat(tmpout,fileres);
1637: return tmpout;
1638: }
1639:
1640: /*************** function subdirf2 ***********/
1641: char *subdirf2(char fileres[], char *preop)
1642: {
1643:
1644: /* Caution optionfilefiname is hidden */
1645: strcpy(tmpout,optionfilefiname);
1646: strcat(tmpout,"/");
1647: strcat(tmpout,preop);
1648: strcat(tmpout,fileres);
1649: return tmpout;
1650: }
1651:
1652: /*************** function subdirf3 ***********/
1653: char *subdirf3(char fileres[], char *preop, char *preop2)
1654: {
1655:
1656: /* Caution optionfilefiname is hidden */
1657: strcpy(tmpout,optionfilefiname);
1658: strcat(tmpout,"/");
1659: strcat(tmpout,preop);
1660: strcat(tmpout,preop2);
1661: strcat(tmpout,fileres);
1662: return tmpout;
1663: }
1.213 brouard 1664:
1665: /*************** function subdirfext ***********/
1666: char *subdirfext(char fileres[], char *preop, char *postop)
1667: {
1668:
1669: strcpy(tmpout,preop);
1670: strcat(tmpout,fileres);
1671: strcat(tmpout,postop);
1672: return tmpout;
1673: }
1.126 brouard 1674:
1.213 brouard 1675: /*************** function subdirfext3 ***********/
1676: char *subdirfext3(char fileres[], char *preop, char *postop)
1677: {
1678:
1679: /* Caution optionfilefiname is hidden */
1680: strcpy(tmpout,optionfilefiname);
1681: strcat(tmpout,"/");
1682: strcat(tmpout,preop);
1683: strcat(tmpout,fileres);
1684: strcat(tmpout,postop);
1685: return tmpout;
1686: }
1687:
1.162 brouard 1688: char *asc_diff_time(long time_sec, char ascdiff[])
1689: {
1690: long sec_left, days, hours, minutes;
1691: days = (time_sec) / (60*60*24);
1692: sec_left = (time_sec) % (60*60*24);
1693: hours = (sec_left) / (60*60) ;
1694: sec_left = (sec_left) %(60*60);
1695: minutes = (sec_left) /60;
1696: sec_left = (sec_left) % (60);
1697: sprintf(ascdiff,"%ld day(s) %ld hour(s) %ld minute(s) %ld second(s)",days, hours, minutes, sec_left);
1698: return ascdiff;
1699: }
1700:
1.126 brouard 1701: /***************** f1dim *************************/
1702: extern int ncom;
1703: extern double *pcom,*xicom;
1704: extern double (*nrfunc)(double []);
1705:
1706: double f1dim(double x)
1707: {
1708: int j;
1709: double f;
1710: double *xt;
1711:
1712: xt=vector(1,ncom);
1713: for (j=1;j<=ncom;j++) xt[j]=pcom[j]+x*xicom[j];
1714: f=(*nrfunc)(xt);
1715: free_vector(xt,1,ncom);
1716: return f;
1717: }
1718:
1719: /*****************brent *************************/
1720: double brent(double ax, double bx, double cx, double (*f)(double), double tol, double *xmin)
1.187 brouard 1721: {
1722: /* Given a function f, and given a bracketing triplet of abscissas ax, bx, cx (such that bx is
1723: * between ax and cx, and f(bx) is less than both f(ax) and f(cx) ), this routine isolates
1724: * the minimum to a fractional precision of about tol using Brent’s method. The abscissa of
1725: * the minimum is returned as xmin, and the minimum function value is returned as brent , the
1726: * returned function value.
1727: */
1.126 brouard 1728: int iter;
1729: double a,b,d,etemp;
1.159 brouard 1730: double fu=0,fv,fw,fx;
1.164 brouard 1731: double ftemp=0.;
1.126 brouard 1732: double p,q,r,tol1,tol2,u,v,w,x,xm;
1733: double e=0.0;
1734:
1735: a=(ax < cx ? ax : cx);
1736: b=(ax > cx ? ax : cx);
1737: x=w=v=bx;
1738: fw=fv=fx=(*f)(x);
1739: for (iter=1;iter<=ITMAX;iter++) {
1740: xm=0.5*(a+b);
1741: tol2=2.0*(tol1=tol*fabs(x)+ZEPS);
1742: /* if (2.0*fabs(fp-(*fret)) <= ftol*(fabs(fp)+fabs(*fret)))*/
1743: printf(".");fflush(stdout);
1744: fprintf(ficlog,".");fflush(ficlog);
1.162 brouard 1745: #ifdef DEBUGBRENT
1.126 brouard 1746: 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);
1747: 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);
1748: /* if ((fabs(x-xm) <= (tol2-0.5*(b-a)))||(2.0*fabs(fu-ftemp) <= ftol*1.e-2*(fabs(fu)+fabs(ftemp)))) { */
1749: #endif
1750: if (fabs(x-xm) <= (tol2-0.5*(b-a))){
1751: *xmin=x;
1752: return fx;
1753: }
1754: ftemp=fu;
1755: if (fabs(e) > tol1) {
1756: r=(x-w)*(fx-fv);
1757: q=(x-v)*(fx-fw);
1758: p=(x-v)*q-(x-w)*r;
1759: q=2.0*(q-r);
1760: if (q > 0.0) p = -p;
1761: q=fabs(q);
1762: etemp=e;
1763: e=d;
1764: if (fabs(p) >= fabs(0.5*q*etemp) || p <= q*(a-x) || p >= q*(b-x))
1.224 brouard 1765: d=CGOLD*(e=(x >= xm ? a-x : b-x));
1.126 brouard 1766: else {
1.224 brouard 1767: d=p/q;
1768: u=x+d;
1769: if (u-a < tol2 || b-u < tol2)
1770: d=SIGN(tol1,xm-x);
1.126 brouard 1771: }
1772: } else {
1773: d=CGOLD*(e=(x >= xm ? a-x : b-x));
1774: }
1775: u=(fabs(d) >= tol1 ? x+d : x+SIGN(tol1,d));
1776: fu=(*f)(u);
1777: if (fu <= fx) {
1778: if (u >= x) a=x; else b=x;
1779: SHFT(v,w,x,u)
1.183 brouard 1780: SHFT(fv,fw,fx,fu)
1781: } else {
1782: if (u < x) a=u; else b=u;
1783: if (fu <= fw || w == x) {
1.224 brouard 1784: v=w;
1785: w=u;
1786: fv=fw;
1787: fw=fu;
1.183 brouard 1788: } else if (fu <= fv || v == x || v == w) {
1.224 brouard 1789: v=u;
1790: fv=fu;
1.183 brouard 1791: }
1792: }
1.126 brouard 1793: }
1794: nrerror("Too many iterations in brent");
1795: *xmin=x;
1796: return fx;
1797: }
1798:
1799: /****************** mnbrak ***********************/
1800:
1801: void mnbrak(double *ax, double *bx, double *cx, double *fa, double *fb, double *fc,
1802: double (*func)(double))
1.183 brouard 1803: { /* Given a function func , and given distinct initial points ax and bx , this routine searches in
1804: the downhill direction (defined by the function as evaluated at the initial points) and returns
1805: new points ax , bx , cx that bracket a minimum of the function. Also returned are the function
1806: values at the three points, fa, fb , and fc such that fa > fb and fb < fc.
1807: */
1.126 brouard 1808: double ulim,u,r,q, dum;
1809: double fu;
1.187 brouard 1810:
1811: double scale=10.;
1812: int iterscale=0;
1813:
1814: *fa=(*func)(*ax); /* xta[j]=pcom[j]+(*ax)*xicom[j]; fa=f(xta[j])*/
1815: *fb=(*func)(*bx); /* xtb[j]=pcom[j]+(*bx)*xicom[j]; fb=f(xtb[j]) */
1816:
1817:
1818: /* while(*fb != *fb){ /\* *ax should be ok, reducing distance to *ax *\/ */
1819: /* printf("Warning mnbrak *fb = %lf, *bx=%lf *ax=%lf *fa==%lf iter=%d\n",*fb, *bx, *ax, *fa, iterscale++); */
1820: /* *bx = *ax - (*ax - *bx)/scale; */
1821: /* *fb=(*func)(*bx); /\* xtb[j]=pcom[j]+(*bx)*xicom[j]; fb=f(xtb[j]) *\/ */
1822: /* } */
1823:
1.126 brouard 1824: if (*fb > *fa) {
1825: SHFT(dum,*ax,*bx,dum)
1.183 brouard 1826: SHFT(dum,*fb,*fa,dum)
1827: }
1.126 brouard 1828: *cx=(*bx)+GOLD*(*bx-*ax);
1829: *fc=(*func)(*cx);
1.183 brouard 1830: #ifdef DEBUG
1.224 brouard 1831: printf("mnbrak0 a=%lf *fa=%lf, b=%lf *fb=%lf, c=%lf *fc=%lf\n",*ax,*fa,*bx,*fb,*cx, *fc);
1832: 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 1833: #endif
1.224 brouard 1834: 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 1835: r=(*bx-*ax)*(*fb-*fc);
1.224 brouard 1836: q=(*bx-*cx)*(*fb-*fa); /* What if fa=inf */
1.126 brouard 1837: u=(*bx)-((*bx-*cx)*q-(*bx-*ax)*r)/
1.183 brouard 1838: (2.0*SIGN(FMAX(fabs(q-r),TINY),q-r)); /* Minimum abscissa of a parabolic estimated from (a,fa), (b,fb) and (c,fc). */
1839: ulim=(*bx)+GLIMIT*(*cx-*bx); /* Maximum abscissa where function should be evaluated */
1840: if ((*bx-u)*(u-*cx) > 0.0) { /* if u_p is between b and c */
1.126 brouard 1841: fu=(*func)(u);
1.163 brouard 1842: #ifdef DEBUG
1843: /* f(x)=A(x-u)**2+f(u) */
1844: double A, fparabu;
1845: A= (*fb - *fa)/(*bx-*ax)/(*bx+*ax-2*u);
1846: fparabu= *fa - A*(*ax-u)*(*ax-u);
1.224 brouard 1847: 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);
1848: 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 1849: /* And thus,it can be that fu > *fc even if fparabu < *fc */
1850: /* mnbrak (*ax=7.666299858533, *fa=299039.693133272231), (*bx=8.595447774979, *fb=298976.598289369489),
1851: (*cx=10.098840694817, *fc=298946.631474258087), (*u=9.852501168332, fu=298948.773013752128, fparabu=298945.434711494134) */
1852: /* In that case, there is no bracket in the output! Routine is wrong with many consequences.*/
1.163 brouard 1853: #endif
1.184 brouard 1854: #ifdef MNBRAKORIGINAL
1.183 brouard 1855: #else
1.191 brouard 1856: /* if (fu > *fc) { */
1857: /* #ifdef DEBUG */
1858: /* printf("mnbrak4 fu > fc \n"); */
1859: /* fprintf(ficlog, "mnbrak4 fu > fc\n"); */
1860: /* #endif */
1861: /* /\* 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 *\\/ *\/ */
1862: /* /\* SHFT(*fa,*fc,fu,*fc) /\\* (b, u, c) is a bracket while test fb > fc will be fu > fc will exit *\\/ *\/ */
1863: /* dum=u; /\* Shifting c and u *\/ */
1864: /* u = *cx; */
1865: /* *cx = dum; */
1866: /* dum = fu; */
1867: /* fu = *fc; */
1868: /* *fc =dum; */
1869: /* } else { /\* end *\/ */
1870: /* #ifdef DEBUG */
1871: /* printf("mnbrak3 fu < fc \n"); */
1872: /* fprintf(ficlog, "mnbrak3 fu < fc\n"); */
1873: /* #endif */
1874: /* dum=u; /\* Shifting c and u *\/ */
1875: /* u = *cx; */
1876: /* *cx = dum; */
1877: /* dum = fu; */
1878: /* fu = *fc; */
1879: /* *fc =dum; */
1880: /* } */
1.224 brouard 1881: #ifdef DEBUGMNBRAK
1882: double A, fparabu;
1883: A= (*fb - *fa)/(*bx-*ax)/(*bx+*ax-2*u);
1884: fparabu= *fa - A*(*ax-u)*(*ax-u);
1885: 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);
1886: 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 1887: #endif
1.191 brouard 1888: dum=u; /* Shifting c and u */
1889: u = *cx;
1890: *cx = dum;
1891: dum = fu;
1892: fu = *fc;
1893: *fc =dum;
1.183 brouard 1894: #endif
1.162 brouard 1895: } else if ((*cx-u)*(u-ulim) > 0.0) { /* u is after c but before ulim */
1.183 brouard 1896: #ifdef DEBUG
1.224 brouard 1897: printf("\nmnbrak2 u=%lf after c=%lf but before ulim\n",u,*cx);
1898: fprintf(ficlog,"\nmnbrak2 u=%lf after c=%lf but before ulim\n",u,*cx);
1.183 brouard 1899: #endif
1.126 brouard 1900: fu=(*func)(u);
1901: if (fu < *fc) {
1.183 brouard 1902: #ifdef DEBUG
1.224 brouard 1903: printf("\nmnbrak2 u=%lf after c=%lf but before ulim=%lf AND fu=%lf < %lf=fc\n",u,*cx,ulim,fu, *fc);
1904: fprintf(ficlog,"\nmnbrak2 u=%lf after c=%lf but before ulim=%lf AND fu=%lf < %lf=fc\n",u,*cx,ulim,fu, *fc);
1905: #endif
1906: SHFT(*bx,*cx,u,*cx+GOLD*(*cx-*bx))
1907: SHFT(*fb,*fc,fu,(*func)(u))
1908: #ifdef DEBUG
1909: printf("\nmnbrak2 shift GOLD c=%lf",*cx+GOLD*(*cx-*bx));
1.183 brouard 1910: #endif
1911: }
1.162 brouard 1912: } else if ((u-ulim)*(ulim-*cx) >= 0.0) { /* u outside ulim (verifying that ulim is beyond c) */
1.183 brouard 1913: #ifdef DEBUG
1.224 brouard 1914: printf("\nmnbrak2 u=%lf outside ulim=%lf (verifying that ulim is beyond c=%lf)\n",u,ulim,*cx);
1915: fprintf(ficlog,"\nmnbrak2 u=%lf outside ulim=%lf (verifying that ulim is beyond c=%lf)\n",u,ulim,*cx);
1.183 brouard 1916: #endif
1.126 brouard 1917: u=ulim;
1918: fu=(*func)(u);
1.183 brouard 1919: } else { /* u could be left to b (if r > q parabola has a maximum) */
1920: #ifdef DEBUG
1.224 brouard 1921: printf("\nmnbrak2 u=%lf could be left to b=%lf (if r=%lf > q=%lf parabola has a maximum)\n",u,*bx,r,q);
1922: 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 1923: #endif
1.126 brouard 1924: u=(*cx)+GOLD*(*cx-*bx);
1925: fu=(*func)(u);
1.224 brouard 1926: #ifdef DEBUG
1927: printf("\nmnbrak2 new u=%lf fu=%lf shifted gold left from c=%lf and b=%lf \n",u,fu,*cx,*bx);
1928: fprintf(ficlog,"\nmnbrak2 new u=%lf fu=%lf shifted gold left from c=%lf and b=%lf \n",u,fu,*cx,*bx);
1929: #endif
1.183 brouard 1930: } /* end tests */
1.126 brouard 1931: SHFT(*ax,*bx,*cx,u)
1.183 brouard 1932: SHFT(*fa,*fb,*fc,fu)
1933: #ifdef DEBUG
1.224 brouard 1934: printf("\nmnbrak2 shift (*ax=%.12f, *fa=%.12lf), (*bx=%.12f, *fb=%.12lf), (*cx=%.12f, *fc=%.12lf)\n",*ax,*fa,*bx,*fb,*cx,*fc);
1935: 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 1936: #endif
1937: } /* 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 1938: }
1939:
1940: /*************** linmin ************************/
1.162 brouard 1941: /* Given an n -dimensional point p[1..n] and an n -dimensional direction xi[1..n] , moves and
1942: resets p to where the function func(p) takes on a minimum along the direction xi from p ,
1943: and replaces xi by the actual vector displacement that p was moved. Also returns as fret
1944: the value of func at the returned location p . This is actually all accomplished by calling the
1945: routines mnbrak and brent .*/
1.126 brouard 1946: int ncom;
1947: double *pcom,*xicom;
1948: double (*nrfunc)(double []);
1949:
1.224 brouard 1950: #ifdef LINMINORIGINAL
1.126 brouard 1951: void linmin(double p[], double xi[], int n, double *fret,double (*func)(double []))
1.224 brouard 1952: #else
1953: void linmin(double p[], double xi[], int n, double *fret,double (*func)(double []), int *flat)
1954: #endif
1.126 brouard 1955: {
1956: double brent(double ax, double bx, double cx,
1957: double (*f)(double), double tol, double *xmin);
1958: double f1dim(double x);
1959: void mnbrak(double *ax, double *bx, double *cx, double *fa, double *fb,
1960: double *fc, double (*func)(double));
1961: int j;
1962: double xx,xmin,bx,ax;
1963: double fx,fb,fa;
1.187 brouard 1964:
1.203 brouard 1965: #ifdef LINMINORIGINAL
1966: #else
1967: double scale=10., axs, xxs; /* Scale added for infinity */
1968: #endif
1969:
1.126 brouard 1970: ncom=n;
1971: pcom=vector(1,n);
1972: xicom=vector(1,n);
1973: nrfunc=func;
1974: for (j=1;j<=n;j++) {
1975: pcom[j]=p[j];
1.202 brouard 1976: xicom[j]=xi[j]; /* Former scale xi[j] of currrent direction i */
1.126 brouard 1977: }
1.187 brouard 1978:
1.203 brouard 1979: #ifdef LINMINORIGINAL
1980: xx=1.;
1981: #else
1982: axs=0.0;
1983: xxs=1.;
1984: do{
1985: xx= xxs;
1986: #endif
1.187 brouard 1987: ax=0.;
1988: mnbrak(&ax,&xx,&bx,&fa,&fx,&fb,f1dim); /* Outputs: xtx[j]=pcom[j]+(*xx)*xicom[j]; fx=f(xtx[j]) */
1989: /* brackets with inputs ax=0 and xx=1, but points, pcom=p, and directions values, xicom=xi, are sent via f1dim(x) */
1990: /* 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)) */
1991: /* Outputs: fa=f(p(j)) and fx=f(p(j) + xxs * xi(j) ) and f(bx)= f(p(j)+ bx* xi(j)) */
1992: /* Given input ax=axs and xx=xxs, xx might be too far from ax to get a finite f(xx) */
1993: /* Searches on line, outputs (ax, xx, bx) such that fx < min(fa and fb) */
1994: /* 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 1995: #ifdef LINMINORIGINAL
1996: #else
1997: if (fx != fx){
1.224 brouard 1998: xxs=xxs/scale; /* Trying a smaller xx, closer to initial ax=0 */
1999: printf("|");
2000: fprintf(ficlog,"|");
1.203 brouard 2001: #ifdef DEBUGLINMIN
1.224 brouard 2002: 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 2003: #endif
2004: }
1.224 brouard 2005: }while(fx != fx && xxs > 1.e-5);
1.203 brouard 2006: #endif
2007:
1.191 brouard 2008: #ifdef DEBUGLINMIN
2009: 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 2010: 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 2011: #endif
1.224 brouard 2012: #ifdef LINMINORIGINAL
2013: #else
2014: if(fb == fx){ /* Flat function in the direction */
2015: xmin=xx;
2016: *flat=1;
2017: }else{
2018: *flat=0;
2019: #endif
2020: /*Flat mnbrak2 shift (*ax=0.000000000000, *fa=51626.272983130431), (*bx=-1.618034000000, *fb=51590.149499362531), (*cx=-4.236068025156, *fc=51590.149499362531) */
1.187 brouard 2021: *fret=brent(ax,xx,bx,f1dim,TOL,&xmin); /* Giving a bracketting triplet (ax, xx, bx), find a minimum, xmin, according to f1dim, *fret(xmin),*/
2022: /* fa = f(p[j] + ax * xi[j]), fx = f(p[j] + xx * xi[j]), fb = f(p[j] + bx * xi[j]) */
2023: /* fmin = f(p[j] + xmin * xi[j]) */
2024: /* P+lambda n in that direction (lambdamin), with TOL between abscisses */
2025: /* f1dim(xmin): for (j=1;j<=ncom;j++) xt[j]=pcom[j]+xmin*xicom[j]; */
1.126 brouard 2026: #ifdef DEBUG
1.224 brouard 2027: 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);
2028: 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);
2029: #endif
2030: #ifdef LINMINORIGINAL
2031: #else
2032: }
1.126 brouard 2033: #endif
1.191 brouard 2034: #ifdef DEBUGLINMIN
2035: printf("linmin end ");
1.202 brouard 2036: fprintf(ficlog,"linmin end ");
1.191 brouard 2037: #endif
1.126 brouard 2038: for (j=1;j<=n;j++) {
1.203 brouard 2039: #ifdef LINMINORIGINAL
2040: xi[j] *= xmin;
2041: #else
2042: #ifdef DEBUGLINMIN
2043: if(xxs <1.0)
2044: printf(" before xi[%d]=%12.8f", j,xi[j]);
2045: #endif
2046: 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) */
2047: #ifdef DEBUGLINMIN
2048: if(xxs <1.0)
2049: 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 );
2050: #endif
2051: #endif
1.187 brouard 2052: p[j] += xi[j]; /* Parameters values are updated accordingly */
1.126 brouard 2053: }
1.191 brouard 2054: #ifdef DEBUGLINMIN
1.203 brouard 2055: printf("\n");
1.191 brouard 2056: printf("Comparing last *frec(xmin=%12.8f)=%12.8f from Brent and frec(0.)=%12.8f \n", xmin, *fret, (*func)(p));
1.202 brouard 2057: 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 2058: for (j=1;j<=n;j++) {
1.202 brouard 2059: printf(" xi[%d]= %14.10f p[%d]= %12.7f",j,xi[j],j,p[j]);
2060: fprintf(ficlog," xi[%d]= %14.10f p[%d]= %12.7f",j,xi[j],j,p[j]);
2061: if(j % ncovmodel == 0){
1.191 brouard 2062: printf("\n");
1.202 brouard 2063: fprintf(ficlog,"\n");
2064: }
1.191 brouard 2065: }
1.203 brouard 2066: #else
1.191 brouard 2067: #endif
1.126 brouard 2068: free_vector(xicom,1,n);
2069: free_vector(pcom,1,n);
2070: }
2071:
2072:
2073: /*************** powell ************************/
1.162 brouard 2074: /*
2075: Minimization of a function func of n variables. Input consists of an initial starting point
2076: p[1..n] ; an initial matrix xi[1..n][1..n] , whose columns contain the initial set of di-
2077: rections (usually the n unit vectors); and ftol , the fractional tolerance in the function value
2078: such that failure to decrease by more than this amount on one iteration signals doneness. On
2079: output, p is set to the best point found, xi is the then-current direction set, fret is the returned
2080: function value at p , and iter is the number of iterations taken. The routine linmin is used.
2081: */
1.224 brouard 2082: #ifdef LINMINORIGINAL
2083: #else
2084: int *flatdir; /* Function is vanishing in that direction */
1.225 brouard 2085: int flat=0, flatd=0; /* Function is vanishing in that direction */
1.224 brouard 2086: #endif
1.126 brouard 2087: void powell(double p[], double **xi, int n, double ftol, int *iter, double *fret,
2088: double (*func)(double []))
2089: {
1.224 brouard 2090: #ifdef LINMINORIGINAL
2091: void linmin(double p[], double xi[], int n, double *fret,
1.126 brouard 2092: double (*func)(double []));
1.224 brouard 2093: #else
1.241 brouard 2094: void linmin(double p[], double xi[], int n, double *fret,
2095: double (*func)(double []),int *flat);
1.224 brouard 2096: #endif
1.239 brouard 2097: int i,ibig,j,jk,k;
1.126 brouard 2098: double del,t,*pt,*ptt,*xit;
1.181 brouard 2099: double directest;
1.126 brouard 2100: double fp,fptt;
2101: double *xits;
2102: int niterf, itmp;
1.224 brouard 2103: #ifdef LINMINORIGINAL
2104: #else
2105:
2106: flatdir=ivector(1,n);
2107: for (j=1;j<=n;j++) flatdir[j]=0;
2108: #endif
1.126 brouard 2109:
2110: pt=vector(1,n);
2111: ptt=vector(1,n);
2112: xit=vector(1,n);
2113: xits=vector(1,n);
2114: *fret=(*func)(p);
2115: for (j=1;j<=n;j++) pt[j]=p[j];
1.202 brouard 2116: rcurr_time = time(NULL);
1.126 brouard 2117: for (*iter=1;;++(*iter)) {
1.187 brouard 2118: fp=(*fret); /* From former iteration or initial value */
1.126 brouard 2119: ibig=0;
2120: del=0.0;
1.157 brouard 2121: rlast_time=rcurr_time;
2122: /* (void) gettimeofday(&curr_time,&tzp); */
2123: rcurr_time = time(NULL);
2124: curr_time = *localtime(&rcurr_time);
2125: printf("\nPowell iter=%d -2*LL=%.12f %ld sec. %ld sec.",*iter,*fret, rcurr_time-rlast_time, rcurr_time-rstart_time);fflush(stdout);
2126: fprintf(ficlog,"\nPowell iter=%d -2*LL=%.12f %ld sec. %ld sec.",*iter,*fret,rcurr_time-rlast_time, rcurr_time-rstart_time); fflush(ficlog);
2127: /* fprintf(ficrespow,"%d %.12f %ld",*iter,*fret,curr_time.tm_sec-start_time.tm_sec); */
1.192 brouard 2128: for (i=1;i<=n;i++) {
1.126 brouard 2129: fprintf(ficrespow," %.12lf", p[i]);
2130: }
1.239 brouard 2131: fprintf(ficrespow,"\n");fflush(ficrespow);
2132: printf("\n#model= 1 + age ");
2133: fprintf(ficlog,"\n#model= 1 + age ");
2134: if(nagesqr==1){
1.241 brouard 2135: printf(" + age*age ");
2136: fprintf(ficlog," + age*age ");
1.239 brouard 2137: }
2138: for(j=1;j <=ncovmodel-2;j++){
2139: if(Typevar[j]==0) {
2140: printf(" + V%d ",Tvar[j]);
2141: fprintf(ficlog," + V%d ",Tvar[j]);
2142: }else if(Typevar[j]==1) {
2143: printf(" + V%d*age ",Tvar[j]);
2144: fprintf(ficlog," + V%d*age ",Tvar[j]);
2145: }else if(Typevar[j]==2) {
2146: printf(" + V%d*V%d ",Tvard[Tposprod[j]][1],Tvard[Tposprod[j]][2]);
2147: fprintf(ficlog," + V%d*V%d ",Tvard[Tposprod[j]][1],Tvard[Tposprod[j]][2]);
2148: }
2149: }
1.126 brouard 2150: printf("\n");
1.239 brouard 2151: /* printf("12 47.0114589 0.0154322 33.2424412 0.3279905 2.3731903 */
2152: /* 13 -21.5392400 0.1118147 1.2680506 1.2973408 -1.0663662 */
1.126 brouard 2153: fprintf(ficlog,"\n");
1.239 brouard 2154: for(i=1,jk=1; i <=nlstate; i++){
2155: for(k=1; k <=(nlstate+ndeath); k++){
2156: if (k != i) {
2157: printf("%d%d ",i,k);
2158: fprintf(ficlog,"%d%d ",i,k);
2159: for(j=1; j <=ncovmodel; j++){
2160: printf("%12.7f ",p[jk]);
2161: fprintf(ficlog,"%12.7f ",p[jk]);
2162: jk++;
2163: }
2164: printf("\n");
2165: fprintf(ficlog,"\n");
2166: }
2167: }
2168: }
1.241 brouard 2169: if(*iter <=3 && *iter >1){
1.157 brouard 2170: tml = *localtime(&rcurr_time);
2171: strcpy(strcurr,asctime(&tml));
2172: rforecast_time=rcurr_time;
1.126 brouard 2173: itmp = strlen(strcurr);
2174: if(strcurr[itmp-1]=='\n') /* Windows outputs with a new line */
1.241 brouard 2175: strcurr[itmp-1]='\0';
1.162 brouard 2176: printf("\nConsidering the time needed for the last iteration #%d: %ld seconds,\n",*iter,rcurr_time-rlast_time);
1.157 brouard 2177: fprintf(ficlog,"\nConsidering the time needed for this last iteration #%d: %ld seconds,\n",*iter,rcurr_time-rlast_time);
1.126 brouard 2178: for(niterf=10;niterf<=30;niterf+=10){
1.241 brouard 2179: rforecast_time=rcurr_time+(niterf-*iter)*(rcurr_time-rlast_time);
2180: forecast_time = *localtime(&rforecast_time);
2181: strcpy(strfor,asctime(&forecast_time));
2182: itmp = strlen(strfor);
2183: if(strfor[itmp-1]=='\n')
2184: strfor[itmp-1]='\0';
2185: 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);
2186: 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 2187: }
2188: }
1.187 brouard 2189: for (i=1;i<=n;i++) { /* For each direction i */
2190: for (j=1;j<=n;j++) xit[j]=xi[j][i]; /* Directions stored from previous iteration with previous scales */
1.126 brouard 2191: fptt=(*fret);
2192: #ifdef DEBUG
1.203 brouard 2193: printf("fret=%lf, %lf, %lf \n", *fret, *fret, *fret);
2194: fprintf(ficlog, "fret=%lf, %lf, %lf \n", *fret, *fret, *fret);
1.126 brouard 2195: #endif
1.203 brouard 2196: printf("%d",i);fflush(stdout); /* print direction (parameter) i */
1.126 brouard 2197: fprintf(ficlog,"%d",i);fflush(ficlog);
1.224 brouard 2198: #ifdef LINMINORIGINAL
1.188 brouard 2199: linmin(p,xit,n,fret,func); /* Point p[n]. xit[n] has been loaded for direction i as input.*/
1.224 brouard 2200: #else
2201: linmin(p,xit,n,fret,func,&flat); /* Point p[n]. xit[n] has been loaded for direction i as input.*/
2202: flatdir[i]=flat; /* Function is vanishing in that direction i */
2203: #endif
2204: /* Outputs are fret(new point p) p is updated and xit rescaled */
1.188 brouard 2205: if (fabs(fptt-(*fret)) > del) { /* We are keeping the max gain on each of the n directions */
1.224 brouard 2206: /* because that direction will be replaced unless the gain del is small */
2207: /* in comparison with the 'probable' gain, mu^2, with the last average direction. */
2208: /* Unless the n directions are conjugate some gain in the determinant may be obtained */
2209: /* with the new direction. */
2210: del=fabs(fptt-(*fret));
2211: ibig=i;
1.126 brouard 2212: }
2213: #ifdef DEBUG
2214: printf("%d %.12e",i,(*fret));
2215: fprintf(ficlog,"%d %.12e",i,(*fret));
2216: for (j=1;j<=n;j++) {
1.224 brouard 2217: xits[j]=FMAX(fabs(p[j]-pt[j]),1.e-5);
2218: printf(" x(%d)=%.12e",j,xit[j]);
2219: fprintf(ficlog," x(%d)=%.12e",j,xit[j]);
1.126 brouard 2220: }
2221: for(j=1;j<=n;j++) {
1.225 brouard 2222: printf(" p(%d)=%.12e",j,p[j]);
2223: fprintf(ficlog," p(%d)=%.12e",j,p[j]);
1.126 brouard 2224: }
2225: printf("\n");
2226: fprintf(ficlog,"\n");
2227: #endif
1.187 brouard 2228: } /* end loop on each direction i */
2229: /* Convergence test will use last linmin estimation (fret) and compare former iteration (fp) */
1.188 brouard 2230: /* But p and xit have been updated at the end of linmin, *fret corresponds to new p, xit */
1.187 brouard 2231: /* New value of last point Pn is not computed, P(n-1) */
1.224 brouard 2232: for(j=1;j<=n;j++) {
1.225 brouard 2233: if(flatdir[j] >0){
2234: printf(" p(%d)=%lf flat=%d ",j,p[j],flatdir[j]);
2235: fprintf(ficlog," p(%d)=%lf flat=%d ",j,p[j],flatdir[j]);
2236: }
2237: /* printf("\n"); */
2238: /* fprintf(ficlog,"\n"); */
2239: }
1.243 brouard 2240: /* if (2.0*fabs(fp-(*fret)) <= ftol*(fabs(fp)+fabs(*fret))) { /\* Did we reach enough precision? *\/ */
2241: if (2.0*fabs(fp-(*fret)) <= ftol) { /* Did we reach enough precision? */
1.188 brouard 2242: /* We could compare with a chi^2. chisquare(0.95,ddl=1)=3.84 */
2243: /* By adding age*age in a model, the new -2LL should be lower and the difference follows a */
2244: /* a chisquare statistics with 1 degree. To be significant at the 95% level, it should have */
2245: /* decreased of more than 3.84 */
2246: /* By adding age*age and V1*age the gain (-2LL) should be more than 5.99 (ddl=2) */
2247: /* By using V1+V2+V3, the gain should be 7.82, compared with basic 1+age. */
2248: /* By adding 10 parameters more the gain should be 18.31 */
1.224 brouard 2249:
1.188 brouard 2250: /* Starting the program with initial values given by a former maximization will simply change */
2251: /* the scales of the directions and the directions, because the are reset to canonical directions */
2252: /* Thus the first calls to linmin will give new points and better maximizations until fp-(*fret) is */
2253: /* under the tolerance value. If the tolerance is very small 1.e-9, it could last long. */
1.126 brouard 2254: #ifdef DEBUG
2255: int k[2],l;
2256: k[0]=1;
2257: k[1]=-1;
2258: printf("Max: %.12e",(*func)(p));
2259: fprintf(ficlog,"Max: %.12e",(*func)(p));
2260: for (j=1;j<=n;j++) {
2261: printf(" %.12e",p[j]);
2262: fprintf(ficlog," %.12e",p[j]);
2263: }
2264: printf("\n");
2265: fprintf(ficlog,"\n");
2266: for(l=0;l<=1;l++) {
2267: for (j=1;j<=n;j++) {
2268: ptt[j]=p[j]+(p[j]-pt[j])*k[l];
2269: printf("l=%d j=%d ptt=%.12e, xits=%.12e, p=%.12e, xit=%.12e", l,j,ptt[j],xits[j],p[j],xit[j]);
2270: fprintf(ficlog,"l=%d j=%d ptt=%.12e, xits=%.12e, p=%.12e, xit=%.12e", l,j,ptt[j],xits[j],p[j],xit[j]);
2271: }
2272: printf("func(ptt)=%.12e, deriv=%.12e\n",(*func)(ptt),(ptt[j]-p[j])/((*func)(ptt)-(*func)(p)));
2273: fprintf(ficlog,"func(ptt)=%.12e, deriv=%.12e\n",(*func)(ptt),(ptt[j]-p[j])/((*func)(ptt)-(*func)(p)));
2274: }
2275: #endif
2276:
1.224 brouard 2277: #ifdef LINMINORIGINAL
2278: #else
2279: free_ivector(flatdir,1,n);
2280: #endif
1.126 brouard 2281: free_vector(xit,1,n);
2282: free_vector(xits,1,n);
2283: free_vector(ptt,1,n);
2284: free_vector(pt,1,n);
2285: return;
1.192 brouard 2286: } /* enough precision */
1.240 brouard 2287: if (*iter == ITMAX*n) nrerror("powell exceeding maximum iterations.");
1.181 brouard 2288: for (j=1;j<=n;j++) { /* Computes the extrapolated point P_0 + 2 (P_n-P_0) */
1.126 brouard 2289: ptt[j]=2.0*p[j]-pt[j];
2290: xit[j]=p[j]-pt[j];
2291: pt[j]=p[j];
2292: }
1.181 brouard 2293: fptt=(*func)(ptt); /* f_3 */
1.224 brouard 2294: #ifdef NODIRECTIONCHANGEDUNTILNITER /* No change in drections until some iterations are done */
2295: if (*iter <=4) {
1.225 brouard 2296: #else
2297: #endif
1.224 brouard 2298: #ifdef POWELLNOF3INFF1TEST /* skips test F3 <F1 */
1.192 brouard 2299: #else
1.161 brouard 2300: if (fptt < fp) { /* If extrapolated point is better, decide if we keep that new direction or not */
1.192 brouard 2301: #endif
1.162 brouard 2302: /* (x1 f1=fp), (x2 f2=*fret), (x3 f3=fptt), (xm fm) */
1.161 brouard 2303: /* From x1 (P0) distance of x2 is at h and x3 is 2h */
1.162 brouard 2304: /* Let f"(x2) be the 2nd derivative equal everywhere. */
2305: /* Then the parabolic through (x1,f1), (x2,f2) and (x3,f3) */
2306: /* will reach at f3 = fm + h^2/2 f"m ; f" = (f1 -2f2 +f3 ) / h**2 */
1.224 brouard 2307: /* Conditional for using this new direction is that mu^2 = (f1-2f2+f3)^2 /2 < del or directest <0 */
2308: /* also lamda^2=(f1-f2)^2/mu² is a parasite solution of powell */
2309: /* For powell, inclusion of this average direction is only if t(del)<0 or del inbetween mu^2 and lambda^2 */
1.161 brouard 2310: /* t=2.0*(fp-2.0*(*fret)+fptt)*SQR(fp-(*fret)-del)-del*SQR(fp-fptt); */
1.224 brouard 2311: /* Even if f3 <f1, directest can be negative and t >0 */
2312: /* mu² and del² are equal when f3=f1 */
2313: /* f3 < f1 : mu² < del <= lambda^2 both test are equivalent */
2314: /* f3 < f1 : mu² < lambda^2 < del then directtest is negative and powell t is positive */
2315: /* f3 > f1 : lambda² < mu^2 < del then t is negative and directest >0 */
2316: /* f3 > f1 : lambda² < del < mu^2 then t is positive and directest >0 */
1.183 brouard 2317: #ifdef NRCORIGINAL
2318: t=2.0*(fp-2.0*(*fret)+fptt)*SQR(fp-(*fret)-del)- del*SQR(fp-fptt); /* Original Numerical Recipes in C*/
2319: #else
2320: 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 2321: t= t- del*SQR(fp-fptt);
1.183 brouard 2322: #endif
1.202 brouard 2323: directest = fp-2.0*(*fret)+fptt - 2.0 * del; /* If delta was big enough we change it for a new direction */
1.161 brouard 2324: #ifdef DEBUG
1.181 brouard 2325: 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);
2326: 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 2327: printf("t3= %.12lf, t4= %.12lf, t3*= %.12lf, t4*= %.12lf\n",SQR(fp-(*fret)-del),SQR(fp-fptt),
2328: (fp-(*fret)-del)*(fp-(*fret)-del),(fp-fptt)*(fp-fptt));
2329: fprintf(ficlog,"t3= %.12lf, t4= %.12lf, t3*= %.12lf, t4*= %.12lf\n",SQR(fp-(*fret)-del),SQR(fp-fptt),
2330: (fp-(*fret)-del)*(fp-(*fret)-del),(fp-fptt)*(fp-fptt));
2331: 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);
2332: 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);
2333: #endif
1.183 brouard 2334: #ifdef POWELLORIGINAL
2335: if (t < 0.0) { /* Then we use it for new direction */
2336: #else
1.182 brouard 2337: if (directest*t < 0.0) { /* Contradiction between both tests */
1.224 brouard 2338: 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 2339: 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 2340: 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 2341: fprintf(ficlog,"f1-2f2+f3= %.12lf, f1-f2-del= %.12lf, f1-f3= %.12lf\n",fp-2.0*(*fret)+fptt, fp -(*fret) -del, fp-fptt);
2342: }
1.181 brouard 2343: if (directest < 0.0) { /* Then we use it for new direction */
2344: #endif
1.191 brouard 2345: #ifdef DEBUGLINMIN
1.234 brouard 2346: printf("Before linmin in direction P%d-P0\n",n);
2347: for (j=1;j<=n;j++) {
2348: printf(" Before xit[%d]= %12.7f p[%d]= %12.7f",j,xit[j],j,p[j]);
2349: fprintf(ficlog," Before xit[%d]= %12.7f p[%d]= %12.7f",j,xit[j],j,p[j]);
2350: if(j % ncovmodel == 0){
2351: printf("\n");
2352: fprintf(ficlog,"\n");
2353: }
2354: }
1.224 brouard 2355: #endif
2356: #ifdef LINMINORIGINAL
1.234 brouard 2357: linmin(p,xit,n,fret,func); /* computes minimum on the extrapolated direction: changes p and rescales xit.*/
1.224 brouard 2358: #else
1.234 brouard 2359: linmin(p,xit,n,fret,func,&flat); /* computes minimum on the extrapolated direction: changes p and rescales xit.*/
2360: flatdir[i]=flat; /* Function is vanishing in that direction i */
1.191 brouard 2361: #endif
1.234 brouard 2362:
1.191 brouard 2363: #ifdef DEBUGLINMIN
1.234 brouard 2364: for (j=1;j<=n;j++) {
2365: printf("After xit[%d]= %12.7f p[%d]= %12.7f",j,xit[j],j,p[j]);
2366: fprintf(ficlog,"After xit[%d]= %12.7f p[%d]= %12.7f",j,xit[j],j,p[j]);
2367: if(j % ncovmodel == 0){
2368: printf("\n");
2369: fprintf(ficlog,"\n");
2370: }
2371: }
1.224 brouard 2372: #endif
1.234 brouard 2373: for (j=1;j<=n;j++) {
2374: xi[j][ibig]=xi[j][n]; /* Replace direction with biggest decrease by last direction n */
2375: xi[j][n]=xit[j]; /* and this nth direction by the by the average p_0 p_n */
2376: }
1.224 brouard 2377: #ifdef LINMINORIGINAL
2378: #else
1.234 brouard 2379: for (j=1, flatd=0;j<=n;j++) {
2380: if(flatdir[j]>0)
2381: flatd++;
2382: }
2383: if(flatd >0){
2384: printf("%d flat directions\n",flatd);
2385: fprintf(ficlog,"%d flat directions\n",flatd);
2386: for (j=1;j<=n;j++) {
2387: if(flatdir[j]>0){
2388: printf("%d ",j);
2389: fprintf(ficlog,"%d ",j);
2390: }
2391: }
2392: printf("\n");
2393: fprintf(ficlog,"\n");
2394: }
1.191 brouard 2395: #endif
1.234 brouard 2396: printf("Gaining to use new average direction of P0 P%d instead of biggest increase direction %d :\n",n,ibig);
2397: fprintf(ficlog,"Gaining to use new average direction of P0 P%d instead of biggest increase direction %d :\n",n,ibig);
2398:
1.126 brouard 2399: #ifdef DEBUG
1.234 brouard 2400: printf("Direction changed last moved %d in place of ibig=%d, new last is the average:\n",n,ibig);
2401: fprintf(ficlog,"Direction changed last moved %d in place of ibig=%d, new last is the average:\n",n,ibig);
2402: for(j=1;j<=n;j++){
2403: printf(" %lf",xit[j]);
2404: fprintf(ficlog," %lf",xit[j]);
2405: }
2406: printf("\n");
2407: fprintf(ficlog,"\n");
1.126 brouard 2408: #endif
1.192 brouard 2409: } /* end of t or directest negative */
1.224 brouard 2410: #ifdef POWELLNOF3INFF1TEST
1.192 brouard 2411: #else
1.234 brouard 2412: } /* end if (fptt < fp) */
1.192 brouard 2413: #endif
1.225 brouard 2414: #ifdef NODIRECTIONCHANGEDUNTILNITER /* No change in drections until some iterations are done */
1.234 brouard 2415: } /*NODIRECTIONCHANGEDUNTILNITER No change in drections until some iterations are done */
1.225 brouard 2416: #else
1.224 brouard 2417: #endif
1.234 brouard 2418: } /* loop iteration */
1.126 brouard 2419: }
1.234 brouard 2420:
1.126 brouard 2421: /**** Prevalence limit (stable or period prevalence) ****************/
1.234 brouard 2422:
1.235 brouard 2423: 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 2424: {
1.235 brouard 2425: /* Computes the prevalence limit in each live state at age x and for covariate combination ij
2426: (and selected quantitative values in nres)
2427: by left multiplying the unit
1.234 brouard 2428: matrix by transitions matrix until convergence is reached with precision ftolpl */
1.206 brouard 2429: /* Wx= Wx-1 Px-1= Wx-2 Px-2 Px-1 = Wx-n Px-n ... Px-2 Px-1 I */
2430: /* Wx is row vector: population in state 1, population in state 2, population dead */
2431: /* or prevalence in state 1, prevalence in state 2, 0 */
2432: /* newm is the matrix after multiplications, its rows are identical at a factor */
2433: /* Initial matrix pimij */
2434: /* {0.85204250825084937, 0.13044499163996345, 0.017512500109187184, */
2435: /* 0.090851990222114765, 0.88271245433047185, 0.026435555447413338, */
2436: /* 0, 0 , 1} */
2437: /*
2438: * and after some iteration: */
2439: /* {0.45504275246439968, 0.42731458730878791, 0.11764266022681241, */
2440: /* 0.45201005341706885, 0.42865420071559901, 0.11933574586733192, */
2441: /* 0, 0 , 1} */
2442: /* And prevalence by suppressing the deaths are close to identical rows in prlim: */
2443: /* {0.51571254859325999, 0.4842874514067399, */
2444: /* 0.51326036147820708, 0.48673963852179264} */
2445: /* If we start from prlim again, prlim tends to a constant matrix */
1.234 brouard 2446:
1.126 brouard 2447: int i, ii,j,k;
1.209 brouard 2448: double *min, *max, *meandiff, maxmax,sumnew=0.;
1.145 brouard 2449: /* double **matprod2(); */ /* test */
1.218 brouard 2450: double **out, cov[NCOVMAX+1], **pmij(); /* **pmmij is a global variable feeded with oldms etc */
1.126 brouard 2451: double **newm;
1.209 brouard 2452: double agefin, delaymax=200. ; /* 100 Max number of years to converge */
1.203 brouard 2453: int ncvloop=0;
1.169 brouard 2454:
1.209 brouard 2455: min=vector(1,nlstate);
2456: max=vector(1,nlstate);
2457: meandiff=vector(1,nlstate);
2458:
1.218 brouard 2459: /* Starting with matrix unity */
1.126 brouard 2460: for (ii=1;ii<=nlstate+ndeath;ii++)
2461: for (j=1;j<=nlstate+ndeath;j++){
2462: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
2463: }
1.169 brouard 2464:
2465: cov[1]=1.;
2466:
2467: /* Even if hstepm = 1, at least one multiplication by the unit matrix */
1.202 brouard 2468: /* Start at agefin= age, computes the matrix of passage and loops decreasing agefin until convergence is reached */
1.126 brouard 2469: for(agefin=age-stepm/YEARM; agefin>=age-delaymax; agefin=agefin-stepm/YEARM){
1.202 brouard 2470: ncvloop++;
1.126 brouard 2471: newm=savm;
2472: /* Covariates have to be included here again */
1.138 brouard 2473: cov[2]=agefin;
1.187 brouard 2474: if(nagesqr==1)
2475: cov[3]= agefin*agefin;;
1.234 brouard 2476: for (k=1; k<=nsd;k++) { /* For single dummy covariates only */
2477: /* Here comes the value of the covariate 'ij' after renumbering k with single dummy covariates */
2478: cov[2+nagesqr+TvarsDind[k]]=nbcode[TvarsD[k]][codtabm(ij,k)];
1.235 brouard 2479: /* 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 2480: }
2481: for (k=1; k<=nsq;k++) { /* For single varying covariates only */
2482: /* Here comes the value of quantitative after renumbering k with single quantitative covariates */
1.235 brouard 2483: cov[2+nagesqr+TvarsQind[k]]=Tqresult[nres][k];
2484: /* 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 2485: }
1.237 brouard 2486: for (k=1; k<=cptcovage;k++){ /* For product with age */
1.234 brouard 2487: if(Dummy[Tvar[Tage[k]]]){
2488: cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2];
2489: } else{
1.235 brouard 2490: cov[2+nagesqr+Tage[k]]=Tqresult[nres][k];
1.234 brouard 2491: }
1.235 brouard 2492: /* 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 2493: }
1.237 brouard 2494: for (k=1; k<=cptcovprod;k++){ /* For product without age */
1.235 brouard 2495: /* 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 2496: if(Dummy[Tvard[k][1]==0]){
2497: if(Dummy[Tvard[k][2]==0]){
2498: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * nbcode[Tvard[k][2]][codtabm(ij,k)];
2499: }else{
2500: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * Tqresult[nres][k];
2501: }
2502: }else{
2503: if(Dummy[Tvard[k][2]==0]){
2504: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][2]][codtabm(ij,k)] * Tqinvresult[nres][Tvard[k][1]];
2505: }else{
2506: cov[2+nagesqr+Tprod[k]]=Tqinvresult[nres][Tvard[k][1]]* Tqinvresult[nres][Tvard[k][2]];
2507: }
2508: }
1.234 brouard 2509: }
1.138 brouard 2510: /*printf("ij=%d cptcovprod=%d tvar=%d ", ij, cptcovprod, Tvar[1]);*/
2511: /*printf("ij=%d cov[3]=%lf cov[4]=%lf \n",ij, cov[3],cov[4]);*/
2512: /*printf("ij=%d cov[3]=%lf \n",ij, cov[3]);*/
1.145 brouard 2513: /* savm=pmij(pmmij,cov,ncovmodel,x,nlstate); */
2514: /* out=matprod2(newm, pmij(pmmij,cov,ncovmodel,x,nlstate),1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, oldm); /\* Bug Valgrind *\/ */
1.218 brouard 2515: /* age and covariate values of ij are in 'cov' */
1.142 brouard 2516: out=matprod2(newm, pmij(pmmij,cov,ncovmodel,x,nlstate),1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, oldm); /* Bug Valgrind */
1.138 brouard 2517:
1.126 brouard 2518: savm=oldm;
2519: oldm=newm;
1.209 brouard 2520:
2521: for(j=1; j<=nlstate; j++){
2522: max[j]=0.;
2523: min[j]=1.;
2524: }
2525: for(i=1;i<=nlstate;i++){
2526: sumnew=0;
2527: for(k=1; k<=ndeath; k++) sumnew+=newm[i][nlstate+k];
2528: for(j=1; j<=nlstate; j++){
2529: prlim[i][j]= newm[i][j]/(1-sumnew);
2530: max[j]=FMAX(max[j],prlim[i][j]);
2531: min[j]=FMIN(min[j],prlim[i][j]);
2532: }
2533: }
2534:
1.126 brouard 2535: maxmax=0.;
1.209 brouard 2536: for(j=1; j<=nlstate; j++){
2537: meandiff[j]=(max[j]-min[j])/(max[j]+min[j])*2.; /* mean difference for each column */
2538: maxmax=FMAX(maxmax,meandiff[j]);
2539: /* 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 2540: } /* j loop */
1.203 brouard 2541: *ncvyear= (int)age- (int)agefin;
1.208 brouard 2542: /* 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 2543: if(maxmax < ftolpl){
1.209 brouard 2544: /* printf("maxmax=%lf ncvloop=%ld, age=%d, agefin=%d ncvyear=%d \n", maxmax, ncvloop, (int)age, (int)agefin, *ncvyear); */
2545: free_vector(min,1,nlstate);
2546: free_vector(max,1,nlstate);
2547: free_vector(meandiff,1,nlstate);
1.126 brouard 2548: return prlim;
2549: }
1.169 brouard 2550: } /* age loop */
1.208 brouard 2551: /* After some age loop it doesn't converge */
1.209 brouard 2552: 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 2553: 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 2554: /* 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); */
2555: free_vector(min,1,nlstate);
2556: free_vector(max,1,nlstate);
2557: free_vector(meandiff,1,nlstate);
1.208 brouard 2558:
1.169 brouard 2559: return prlim; /* should not reach here */
1.126 brouard 2560: }
2561:
1.217 brouard 2562:
2563: /**** Back Prevalence limit (stable or period prevalence) ****************/
2564:
1.218 brouard 2565: /* 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) */
2566: /* 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 2567: double **bprevalim(double **bprlim, double ***prevacurrent, int nlstate, double x[], double age, double ftolpl, int *ncvyear, int ij, int nres)
1.217 brouard 2568: {
1.218 brouard 2569: /* Computes the prevalence limit in each live state at age x and covariate ij by left multiplying the unit
1.217 brouard 2570: matrix by transitions matrix until convergence is reached with precision ftolpl */
2571: /* Wx= Wx-1 Px-1= Wx-2 Px-2 Px-1 = Wx-n Px-n ... Px-2 Px-1 I */
2572: /* Wx is row vector: population in state 1, population in state 2, population dead */
2573: /* or prevalence in state 1, prevalence in state 2, 0 */
2574: /* newm is the matrix after multiplications, its rows are identical at a factor */
2575: /* Initial matrix pimij */
2576: /* {0.85204250825084937, 0.13044499163996345, 0.017512500109187184, */
2577: /* 0.090851990222114765, 0.88271245433047185, 0.026435555447413338, */
2578: /* 0, 0 , 1} */
2579: /*
2580: * and after some iteration: */
2581: /* {0.45504275246439968, 0.42731458730878791, 0.11764266022681241, */
2582: /* 0.45201005341706885, 0.42865420071559901, 0.11933574586733192, */
2583: /* 0, 0 , 1} */
2584: /* And prevalence by suppressing the deaths are close to identical rows in prlim: */
2585: /* {0.51571254859325999, 0.4842874514067399, */
2586: /* 0.51326036147820708, 0.48673963852179264} */
2587: /* If we start from prlim again, prlim tends to a constant matrix */
2588:
2589: int i, ii,j,k;
1.247 brouard 2590: int first=0;
1.217 brouard 2591: double *min, *max, *meandiff, maxmax,sumnew=0.;
2592: /* double **matprod2(); */ /* test */
2593: double **out, cov[NCOVMAX+1], **bmij();
2594: double **newm;
1.218 brouard 2595: double **dnewm, **doldm, **dsavm; /* for use */
2596: double **oldm, **savm; /* for use */
2597:
1.217 brouard 2598: double agefin, delaymax=200. ; /* 100 Max number of years to converge */
2599: int ncvloop=0;
2600:
2601: min=vector(1,nlstate);
2602: max=vector(1,nlstate);
2603: meandiff=vector(1,nlstate);
2604:
1.218 brouard 2605: dnewm=ddnewms; doldm=ddoldms; dsavm=ddsavms;
2606: oldm=oldms; savm=savms;
2607:
2608: /* Starting with matrix unity */
2609: for (ii=1;ii<=nlstate+ndeath;ii++)
2610: for (j=1;j<=nlstate+ndeath;j++){
1.217 brouard 2611: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
2612: }
2613:
2614: cov[1]=1.;
2615:
2616: /* Even if hstepm = 1, at least one multiplication by the unit matrix */
2617: /* Start at agefin= age, computes the matrix of passage and loops decreasing agefin until convergence is reached */
1.218 brouard 2618: /* for(agefin=age+stepm/YEARM; agefin<=age+delaymax; agefin=agefin+stepm/YEARM){ /\* A changer en age *\/ */
2619: for(agefin=age; agefin<AGESUP; agefin=agefin+stepm/YEARM){ /* A changer en age */
1.217 brouard 2620: ncvloop++;
1.218 brouard 2621: newm=savm; /* oldm should be kept from previous iteration or unity at start */
2622: /* newm points to the allocated table savm passed by the function it can be written, savm could be reallocated */
1.217 brouard 2623: /* Covariates have to be included here again */
2624: cov[2]=agefin;
2625: if(nagesqr==1)
2626: cov[3]= agefin*agefin;;
1.242 brouard 2627: for (k=1; k<=nsd;k++) { /* For single dummy covariates only */
2628: /* Here comes the value of the covariate 'ij' after renumbering k with single dummy covariates */
2629: cov[2+nagesqr+TvarsDind[k]]=nbcode[TvarsD[k]][codtabm(ij,k)];
2630: /* printf("bprevalim 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)); */
2631: }
2632: /* for (k=1; k<=cptcovn;k++) { */
2633: /* /\* cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(ij,Tvar[k])]; *\/ */
2634: /* cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(ij,k)]; */
2635: /* /\* 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])]); *\/ */
2636: /* } */
2637: for (k=1; k<=nsq;k++) { /* For single varying covariates only */
2638: /* Here comes the value of quantitative after renumbering k with single quantitative covariates */
2639: cov[2+nagesqr+TvarsQind[k]]=Tqresult[nres][k];
2640: /* 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]); */
2641: }
2642: /* for (k=1; k<=cptcovage;k++) cov[2+nagesqr+Tage[k]]=nbcode[Tvar[k]][codtabm(ij,k)]*cov[2]; */
2643: /* for (k=1; k<=cptcovprod;k++) /\* Useless *\/ */
2644: /* /\* cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,Tvard[k][1])] * nbcode[Tvard[k][2]][codtabm(ij,Tvard[k][2])]; *\/ */
2645: /* cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * nbcode[Tvard[k][2]][codtabm(ij,k)]; */
2646: for (k=1; k<=cptcovage;k++){ /* For product with age */
2647: if(Dummy[Tvar[Tage[k]]]){
2648: cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2];
2649: } else{
2650: cov[2+nagesqr+Tage[k]]=Tqresult[nres][k];
2651: }
2652: /* 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]); */
2653: }
2654: for (k=1; k<=cptcovprod;k++){ /* For product without age */
2655: /* 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]); */
2656: if(Dummy[Tvard[k][1]==0]){
2657: if(Dummy[Tvard[k][2]==0]){
2658: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * nbcode[Tvard[k][2]][codtabm(ij,k)];
2659: }else{
2660: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * Tqresult[nres][k];
2661: }
2662: }else{
2663: if(Dummy[Tvard[k][2]==0]){
2664: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][2]][codtabm(ij,k)] * Tqinvresult[nres][Tvard[k][1]];
2665: }else{
2666: cov[2+nagesqr+Tprod[k]]=Tqinvresult[nres][Tvard[k][1]]* Tqinvresult[nres][Tvard[k][2]];
2667: }
2668: }
1.217 brouard 2669: }
2670:
2671: /*printf("ij=%d cptcovprod=%d tvar=%d ", ij, cptcovprod, Tvar[1]);*/
2672: /*printf("ij=%d cov[3]=%lf cov[4]=%lf \n",ij, cov[3],cov[4]);*/
2673: /*printf("ij=%d cov[3]=%lf \n",ij, cov[3]);*/
2674: /* savm=pmij(pmmij,cov,ncovmodel,x,nlstate); */
2675: /* out=matprod2(newm, pmij(pmmij,cov,ncovmodel,x,nlstate),1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, oldm); /\* Bug Valgrind *\/ */
1.218 brouard 2676: /* ij should be linked to the correct index of cov */
2677: /* age and covariate values ij are in 'cov', but we need to pass
2678: * ij for the observed prevalence at age and status and covariate
2679: * number: prevacurrent[(int)agefin][ii][ij]
2680: */
2681: /* 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 *\/ */
2682: /* 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 *\/ */
2683: 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 2684: savm=oldm;
2685: oldm=newm;
2686: for(j=1; j<=nlstate; j++){
2687: max[j]=0.;
2688: min[j]=1.;
2689: }
2690: for(j=1; j<=nlstate; j++){
2691: for(i=1;i<=nlstate;i++){
1.234 brouard 2692: /* bprlim[i][j]= newm[i][j]/(1-sumnew); */
2693: bprlim[i][j]= newm[i][j];
2694: max[i]=FMAX(max[i],bprlim[i][j]); /* Max in line */
2695: min[i]=FMIN(min[i],bprlim[i][j]);
1.217 brouard 2696: }
2697: }
1.218 brouard 2698:
1.217 brouard 2699: maxmax=0.;
2700: for(i=1; i<=nlstate; i++){
2701: meandiff[i]=(max[i]-min[i])/(max[i]+min[i])*2.; /* mean difference for each column */
2702: maxmax=FMAX(maxmax,meandiff[i]);
2703: /* 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); */
2704: } /* j loop */
2705: *ncvyear= -( (int)age- (int)agefin);
1.218 brouard 2706: /* printf("Back maxmax=%lf ncvloop=%d, age=%d, agefin=%d ncvyear=%d \n", maxmax, ncvloop, (int)age, (int)agefin, *ncvyear);*/
1.217 brouard 2707: if(maxmax < ftolpl){
1.220 brouard 2708: /* printf("OK Back maxmax=%lf ncvloop=%d, age=%d, agefin=%d ncvyear=%d \n", maxmax, ncvloop, (int)age, (int)agefin, *ncvyear); */
1.217 brouard 2709: free_vector(min,1,nlstate);
2710: free_vector(max,1,nlstate);
2711: free_vector(meandiff,1,nlstate);
2712: return bprlim;
2713: }
2714: } /* age loop */
2715: /* After some age loop it doesn't converge */
1.247 brouard 2716: if(first){
2717: first=1;
2718: 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\
2719: 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);
2720: }
2721: 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 2722: 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);
2723: /* 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); */
2724: free_vector(min,1,nlstate);
2725: free_vector(max,1,nlstate);
2726: free_vector(meandiff,1,nlstate);
2727:
2728: return bprlim; /* should not reach here */
2729: }
2730:
1.126 brouard 2731: /*************** transition probabilities ***************/
2732:
2733: double **pmij(double **ps, double *cov, int ncovmodel, double *x, int nlstate )
2734: {
1.138 brouard 2735: /* According to parameters values stored in x and the covariate's values stored in cov,
2736: computes the probability to be observed in state j being in state i by appying the
2737: model to the ncovmodel covariates (including constant and age).
2738: lnpijopii=ln(pij/pii)= aij+bij*age+cij*v1+dij*v2+... = sum_nc=1^ncovmodel xij(nc)*cov[nc]
2739: and, according on how parameters are entered, the position of the coefficient xij(nc) of the
2740: ncth covariate in the global vector x is given by the formula:
2741: j<i nc+((i-1)*(nlstate+ndeath-1)+j-1)*ncovmodel
2742: j>=i nc + ((i-1)*(nlstate+ndeath-1)+(j-2))*ncovmodel
2743: Computes ln(pij/pii) (lnpijopii), deduces pij/pii by exponentiation,
2744: sums on j different of i to get 1-pii/pii, deduces pii, and then all pij.
2745: Outputs ps[i][j] the probability to be observed in j being in j according to
2746: the values of the covariates cov[nc] and corresponding parameter values x[nc+shiftij]
2747: */
2748: double s1, lnpijopii;
1.126 brouard 2749: /*double t34;*/
1.164 brouard 2750: int i,j, nc, ii, jj;
1.126 brouard 2751:
1.223 brouard 2752: for(i=1; i<= nlstate; i++){
2753: for(j=1; j<i;j++){
2754: for (nc=1, lnpijopii=0.;nc <=ncovmodel; nc++){
2755: /*lnpijopii += param[i][j][nc]*cov[nc];*/
2756: lnpijopii += x[nc+((i-1)*(nlstate+ndeath-1)+j-1)*ncovmodel]*cov[nc];
2757: /* printf("Int j<i s1=%.17e, lnpijopii=%.17e\n",s1,lnpijopii); */
2758: }
2759: ps[i][j]=lnpijopii; /* In fact ln(pij/pii) */
2760: /* printf("s1=%.17e, lnpijopii=%.17e\n",s1,lnpijopii); */
2761: }
2762: for(j=i+1; j<=nlstate+ndeath;j++){
2763: for (nc=1, lnpijopii=0.;nc <=ncovmodel; nc++){
2764: /*lnpijopii += x[(i-1)*nlstate*ncovmodel+(j-2)*ncovmodel+nc+(i-1)*(ndeath-1)*ncovmodel]*cov[nc];*/
2765: lnpijopii += x[nc + ((i-1)*(nlstate+ndeath-1)+(j-2))*ncovmodel]*cov[nc];
2766: /* printf("Int j>i s1=%.17e, lnpijopii=%.17e %lx %lx\n",s1,lnpijopii,s1,lnpijopii); */
2767: }
2768: ps[i][j]=lnpijopii; /* In fact ln(pij/pii) */
2769: }
2770: }
1.218 brouard 2771:
1.223 brouard 2772: for(i=1; i<= nlstate; i++){
2773: s1=0;
2774: for(j=1; j<i; j++){
2775: s1+=exp(ps[i][j]); /* In fact sums pij/pii */
2776: /*printf("debug1 %d %d ps=%lf exp(ps)=%lf s1+=%lf\n",i,j,ps[i][j],exp(ps[i][j]),s1); */
2777: }
2778: for(j=i+1; j<=nlstate+ndeath; j++){
2779: s1+=exp(ps[i][j]); /* In fact sums pij/pii */
2780: /*printf("debug2 %d %d ps=%lf exp(ps)=%lf s1+=%lf\n",i,j,ps[i][j],exp(ps[i][j]),s1); */
2781: }
2782: /* s1= sum_{j<>i} pij/pii=(1-pii)/pii and thus pii is known from s1 */
2783: ps[i][i]=1./(s1+1.);
2784: /* Computing other pijs */
2785: for(j=1; j<i; j++)
2786: ps[i][j]= exp(ps[i][j])*ps[i][i];
2787: for(j=i+1; j<=nlstate+ndeath; j++)
2788: ps[i][j]= exp(ps[i][j])*ps[i][i];
2789: /* ps[i][nlstate+1]=1.-s1- ps[i][i];*/ /* Sum should be 1 */
2790: } /* end i */
1.218 brouard 2791:
1.223 brouard 2792: for(ii=nlstate+1; ii<= nlstate+ndeath; ii++){
2793: for(jj=1; jj<= nlstate+ndeath; jj++){
2794: ps[ii][jj]=0;
2795: ps[ii][ii]=1;
2796: }
2797: }
1.218 brouard 2798:
2799:
1.223 brouard 2800: /* for(ii=1; ii<= nlstate+ndeath; ii++){ */
2801: /* for(jj=1; jj<= nlstate+ndeath; jj++){ */
2802: /* printf(" pmij ps[%d][%d]=%lf ",ii,jj,ps[ii][jj]); */
2803: /* } */
2804: /* printf("\n "); */
2805: /* } */
2806: /* printf("\n ");printf("%lf ",cov[2]);*/
2807: /*
2808: for(i=1; i<= npar; i++) printf("%f ",x[i]);
1.218 brouard 2809: goto end;*/
1.223 brouard 2810: return ps;
1.126 brouard 2811: }
2812:
1.218 brouard 2813: /*************** backward transition probabilities ***************/
2814:
2815: /* 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 ) */
2816: /* double **bmij(double **ps, double *cov, int ncovmodel, double *x, int nlstate, double ***prevacurrent, double ***dnewm, double **doldm, double **dsavm, int ij ) */
2817: double **bmij(double **ps, double *cov, int ncovmodel, double *x, int nlstate, double ***prevacurrent, int ij )
2818: {
1.222 brouard 2819: /* Computes the backward probability at age agefin and covariate ij
2820: * and returns in **ps as well as **bmij.
2821: */
1.218 brouard 2822: int i, ii, j,k;
1.222 brouard 2823:
2824: double **out, **pmij();
2825: double sumnew=0.;
1.218 brouard 2826: double agefin;
1.222 brouard 2827:
2828: double **dnewm, **dsavm, **doldm;
2829: double **bbmij;
2830:
1.218 brouard 2831: doldm=ddoldms; /* global pointers */
1.222 brouard 2832: dnewm=ddnewms;
2833: dsavm=ddsavms;
2834:
2835: agefin=cov[2];
2836: /* bmij *//* age is cov[2], ij is included in cov, but we need for
2837: the observed prevalence (with this covariate ij) */
2838: dsavm=pmij(pmmij,cov,ncovmodel,x,nlstate);
2839: /* We do have the matrix Px in savm and we need pij */
2840: for (j=1;j<=nlstate+ndeath;j++){
2841: sumnew=0.; /* w1 p11 + w2 p21 only on live states */
2842: for (ii=1;ii<=nlstate;ii++){
2843: sumnew+=dsavm[ii][j]*prevacurrent[(int)agefin][ii][ij];
2844: } /* sumnew is (N11+N21)/N..= N.1/N.. = sum on i of w_i pij */
2845: for (ii=1;ii<=nlstate+ndeath;ii++){
2846: if(sumnew >= 1.e-10){
2847: /* if(agefin >= agemaxpar && agefin <= agemaxpar+stepm/YEARM){ */
2848: /* doldm[ii][j]=(ii==j ? 1./sumnew : 0.0); */
2849: /* }else if(agefin >= agemaxpar+stepm/YEARM){ */
2850: /* doldm[ii][j]=(ii==j ? 1./sumnew : 0.0); */
2851: /* }else */
2852: doldm[ii][j]=(ii==j ? 1./sumnew : 0.0);
2853: }else{
1.242 brouard 2854: ;
2855: /* 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 2856: }
2857: } /*End ii */
2858: } /* End j, At the end doldm is diag[1/(w_1p1i+w_2 p2i)] */
2859: /* left Product of this diag matrix by dsavm=Px (newm=dsavm*doldm) */
2860: bbmij=matprod2(dnewm, dsavm,1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, doldm); /* Bug Valgrind */
2861: /* dsavm=doldm; /\* dsavm is now diag [1/(w_1p1i+w_2 p2i)] but can be overwritten*\/ */
2862: /* doldm=dnewm; /\* doldm is now Px * diag [1/(w_1p1i+w_2 p2i)] *\/ */
2863: /* dnewm=dsavm; /\* doldm is now Px * diag [1/(w_1p1i+w_2 p2i)] *\/ */
2864: /* left Product of this matrix by diag matrix of prevalences (savm) */
2865: for (j=1;j<=nlstate+ndeath;j++){
2866: for (ii=1;ii<=nlstate+ndeath;ii++){
2867: dsavm[ii][j]=(ii==j ? prevacurrent[(int)agefin][ii][ij] : 0.0);
2868: }
2869: } /* End j, At the end oldm is diag[1/(w_1p1i+w_2 p2i)] */
2870: ps=matprod2(doldm, dsavm,1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, dnewm); /* Bug Valgrind */
2871: /* newm or out is now diag[w_i] * Px * diag [1/(w_1p1i+w_2 p2i)] */
2872: /* end bmij */
2873: return ps;
1.218 brouard 2874: }
1.217 brouard 2875: /*************** transition probabilities ***************/
2876:
1.218 brouard 2877: double **bpmij(double **ps, double *cov, int ncovmodel, double *x, int nlstate )
1.217 brouard 2878: {
2879: /* According to parameters values stored in x and the covariate's values stored in cov,
2880: computes the probability to be observed in state j being in state i by appying the
2881: model to the ncovmodel covariates (including constant and age).
2882: lnpijopii=ln(pij/pii)= aij+bij*age+cij*v1+dij*v2+... = sum_nc=1^ncovmodel xij(nc)*cov[nc]
2883: and, according on how parameters are entered, the position of the coefficient xij(nc) of the
2884: ncth covariate in the global vector x is given by the formula:
2885: j<i nc+((i-1)*(nlstate+ndeath-1)+j-1)*ncovmodel
2886: j>=i nc + ((i-1)*(nlstate+ndeath-1)+(j-2))*ncovmodel
2887: Computes ln(pij/pii) (lnpijopii), deduces pij/pii by exponentiation,
2888: sums on j different of i to get 1-pii/pii, deduces pii, and then all pij.
2889: Outputs ps[i][j] the probability to be observed in j being in j according to
2890: the values of the covariates cov[nc] and corresponding parameter values x[nc+shiftij]
2891: */
2892: double s1, lnpijopii;
2893: /*double t34;*/
2894: int i,j, nc, ii, jj;
2895:
1.234 brouard 2896: for(i=1; i<= nlstate; i++){
2897: for(j=1; j<i;j++){
2898: for (nc=1, lnpijopii=0.;nc <=ncovmodel; nc++){
2899: /*lnpijopii += param[i][j][nc]*cov[nc];*/
2900: lnpijopii += x[nc+((i-1)*(nlstate+ndeath-1)+j-1)*ncovmodel]*cov[nc];
2901: /* printf("Int j<i s1=%.17e, lnpijopii=%.17e\n",s1,lnpijopii); */
2902: }
2903: ps[i][j]=lnpijopii; /* In fact ln(pij/pii) */
2904: /* printf("s1=%.17e, lnpijopii=%.17e\n",s1,lnpijopii); */
2905: }
2906: for(j=i+1; j<=nlstate+ndeath;j++){
2907: for (nc=1, lnpijopii=0.;nc <=ncovmodel; nc++){
2908: /*lnpijopii += x[(i-1)*nlstate*ncovmodel+(j-2)*ncovmodel+nc+(i-1)*(ndeath-1)*ncovmodel]*cov[nc];*/
2909: lnpijopii += x[nc + ((i-1)*(nlstate+ndeath-1)+(j-2))*ncovmodel]*cov[nc];
2910: /* printf("Int j>i s1=%.17e, lnpijopii=%.17e %lx %lx\n",s1,lnpijopii,s1,lnpijopii); */
2911: }
2912: ps[i][j]=lnpijopii; /* In fact ln(pij/pii) */
2913: }
2914: }
2915:
2916: for(i=1; i<= nlstate; i++){
2917: s1=0;
2918: for(j=1; j<i; j++){
2919: s1+=exp(ps[i][j]); /* In fact sums pij/pii */
2920: /*printf("debug1 %d %d ps=%lf exp(ps)=%lf s1+=%lf\n",i,j,ps[i][j],exp(ps[i][j]),s1); */
2921: }
2922: for(j=i+1; j<=nlstate+ndeath; j++){
2923: s1+=exp(ps[i][j]); /* In fact sums pij/pii */
2924: /*printf("debug2 %d %d ps=%lf exp(ps)=%lf s1+=%lf\n",i,j,ps[i][j],exp(ps[i][j]),s1); */
2925: }
2926: /* s1= sum_{j<>i} pij/pii=(1-pii)/pii and thus pii is known from s1 */
2927: ps[i][i]=1./(s1+1.);
2928: /* Computing other pijs */
2929: for(j=1; j<i; j++)
2930: ps[i][j]= exp(ps[i][j])*ps[i][i];
2931: for(j=i+1; j<=nlstate+ndeath; j++)
2932: ps[i][j]= exp(ps[i][j])*ps[i][i];
2933: /* ps[i][nlstate+1]=1.-s1- ps[i][i];*/ /* Sum should be 1 */
2934: } /* end i */
2935:
2936: for(ii=nlstate+1; ii<= nlstate+ndeath; ii++){
2937: for(jj=1; jj<= nlstate+ndeath; jj++){
2938: ps[ii][jj]=0;
2939: ps[ii][ii]=1;
2940: }
2941: }
2942: /* Added for backcast */ /* Transposed matrix too */
2943: for(jj=1; jj<= nlstate+ndeath; jj++){
2944: s1=0.;
2945: for(ii=1; ii<= nlstate+ndeath; ii++){
2946: s1+=ps[ii][jj];
2947: }
2948: for(ii=1; ii<= nlstate; ii++){
2949: ps[ii][jj]=ps[ii][jj]/s1;
2950: }
2951: }
2952: /* Transposition */
2953: for(jj=1; jj<= nlstate+ndeath; jj++){
2954: for(ii=jj; ii<= nlstate+ndeath; ii++){
2955: s1=ps[ii][jj];
2956: ps[ii][jj]=ps[jj][ii];
2957: ps[jj][ii]=s1;
2958: }
2959: }
2960: /* for(ii=1; ii<= nlstate+ndeath; ii++){ */
2961: /* for(jj=1; jj<= nlstate+ndeath; jj++){ */
2962: /* printf(" pmij ps[%d][%d]=%lf ",ii,jj,ps[ii][jj]); */
2963: /* } */
2964: /* printf("\n "); */
2965: /* } */
2966: /* printf("\n ");printf("%lf ",cov[2]);*/
2967: /*
2968: for(i=1; i<= npar; i++) printf("%f ",x[i]);
2969: goto end;*/
2970: return ps;
1.217 brouard 2971: }
2972:
2973:
1.126 brouard 2974: /**************** Product of 2 matrices ******************/
2975:
1.145 brouard 2976: double **matprod2(double **out, double **in,int nrl, int nrh, int ncl, int nch, int ncolol, int ncoloh, double **b)
1.126 brouard 2977: {
2978: /* Computes the matrix product of in(1,nrh-nrl+1)(1,nch-ncl+1) times
2979: b(1,nch-ncl+1)(1,ncoloh-ncolol+1) into out(...) */
2980: /* in, b, out are matrice of pointers which should have been initialized
2981: before: only the contents of out is modified. The function returns
2982: a pointer to pointers identical to out */
1.145 brouard 2983: int i, j, k;
1.126 brouard 2984: for(i=nrl; i<= nrh; i++)
1.145 brouard 2985: for(k=ncolol; k<=ncoloh; k++){
2986: out[i][k]=0.;
2987: for(j=ncl; j<=nch; j++)
2988: out[i][k] +=in[i][j]*b[j][k];
2989: }
1.126 brouard 2990: return out;
2991: }
2992:
2993:
2994: /************* Higher Matrix Product ***************/
2995:
1.235 brouard 2996: 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 2997: {
1.218 brouard 2998: /* Computes the transition matrix starting at age 'age' and combination of covariate values corresponding to ij over
1.126 brouard 2999: 'nhstepm*hstepm*stepm' months (i.e. until
3000: age (in years) age+nhstepm*hstepm*stepm/12) by multiplying
3001: nhstepm*hstepm matrices.
3002: Output is stored in matrix po[i][j][h] for h every 'hstepm' step
3003: (typically every 2 years instead of every month which is too big
3004: for the memory).
3005: Model is determined by parameters x and covariates have to be
3006: included manually here.
3007:
3008: */
3009:
3010: int i, j, d, h, k;
1.131 brouard 3011: double **out, cov[NCOVMAX+1];
1.126 brouard 3012: double **newm;
1.187 brouard 3013: double agexact;
1.214 brouard 3014: double agebegin, ageend;
1.126 brouard 3015:
3016: /* Hstepm could be zero and should return the unit matrix */
3017: for (i=1;i<=nlstate+ndeath;i++)
3018: for (j=1;j<=nlstate+ndeath;j++){
3019: oldm[i][j]=(i==j ? 1.0 : 0.0);
3020: po[i][j][0]=(i==j ? 1.0 : 0.0);
3021: }
3022: /* Even if hstepm = 1, at least one multiplication by the unit matrix */
3023: for(h=1; h <=nhstepm; h++){
3024: for(d=1; d <=hstepm; d++){
3025: newm=savm;
3026: /* Covariates have to be included here again */
3027: cov[1]=1.;
1.214 brouard 3028: agexact=age+((h-1)*hstepm + (d-1))*stepm/YEARM; /* age just before transition */
1.187 brouard 3029: cov[2]=agexact;
3030: if(nagesqr==1)
1.227 brouard 3031: cov[3]= agexact*agexact;
1.235 brouard 3032: for (k=1; k<=nsd;k++) { /* For single dummy covariates only */
3033: /* Here comes the value of the covariate 'ij' after renumbering k with single dummy covariates */
3034: cov[2+nagesqr+TvarsDind[k]]=nbcode[TvarsD[k]][codtabm(ij,k)];
3035: /* 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)); */
3036: }
3037: for (k=1; k<=nsq;k++) { /* For single varying covariates only */
3038: /* Here comes the value of quantitative after renumbering k with single quantitative covariates */
3039: cov[2+nagesqr+TvarsQind[k]]=Tqresult[nres][k];
3040: /* 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]); */
3041: }
3042: for (k=1; k<=cptcovage;k++){
3043: if(Dummy[Tvar[Tage[k]]]){
3044: cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2];
3045: } else{
3046: cov[2+nagesqr+Tage[k]]=Tqresult[nres][k];
3047: }
3048: /* 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]); */
3049: }
3050: for (k=1; k<=cptcovprod;k++){ /* */
3051: /* 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]); */
3052: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)] * nbcode[Tvard[k][2]][codtabm(ij,k)];
3053: }
3054: /* for (k=1; k<=cptcovn;k++) */
3055: /* cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(ij,k)]; */
3056: /* for (k=1; k<=cptcovage;k++) /\* Should start at cptcovn+1 *\/ */
3057: /* cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2]; */
3058: /* for (k=1; k<=cptcovprod;k++) /\* Useless because included in cptcovn *\/ */
3059: /* cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)]*nbcode[Tvard[k][2]][codtabm(ij,k)]; */
1.227 brouard 3060:
3061:
1.126 brouard 3062: /*printf("hxi cptcov=%d cptcode=%d\n",cptcov,cptcode);*/
3063: /*printf("h=%d d=%d age=%f cov=%f\n",h,d,age,cov[2]);*/
1.218 brouard 3064: /* right multiplication of oldm by the current matrix */
1.126 brouard 3065: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath,
3066: pmij(pmmij,cov,ncovmodel,x,nlstate));
1.217 brouard 3067: /* if((int)age == 70){ */
3068: /* printf(" Forward hpxij age=%d agexact=%f d=%d nhstepm=%d hstepm=%d\n", (int) age, agexact, d, nhstepm, hstepm); */
3069: /* for(i=1; i<=nlstate+ndeath; i++) { */
3070: /* printf("%d pmmij ",i); */
3071: /* for(j=1;j<=nlstate+ndeath;j++) { */
3072: /* printf("%f ",pmmij[i][j]); */
3073: /* } */
3074: /* printf(" oldm "); */
3075: /* for(j=1;j<=nlstate+ndeath;j++) { */
3076: /* printf("%f ",oldm[i][j]); */
3077: /* } */
3078: /* printf("\n"); */
3079: /* } */
3080: /* } */
1.126 brouard 3081: savm=oldm;
3082: oldm=newm;
3083: }
3084: for(i=1; i<=nlstate+ndeath; i++)
3085: for(j=1;j<=nlstate+ndeath;j++) {
1.218 brouard 3086: po[i][j][h]=newm[i][j];
3087: /*if(h==nhstepm) printf("po[%d][%d][%d]=%f ",i,j,h,po[i][j][h]);*/
1.126 brouard 3088: }
1.128 brouard 3089: /*printf("h=%d ",h);*/
1.126 brouard 3090: } /* end h */
1.218 brouard 3091: /* printf("\n H=%d \n",h); */
1.126 brouard 3092: return po;
3093: }
3094:
1.217 brouard 3095: /************* Higher Back Matrix Product ***************/
1.218 brouard 3096: /* 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 3097: double ***hbxij(double ***po, int nhstepm, double age, int hstepm, double *x, double ***prevacurrent, int nlstate, int stepm, int ij )
1.217 brouard 3098: {
1.218 brouard 3099: /* Computes the transition matrix starting at age 'age' over
1.217 brouard 3100: 'nhstepm*hstepm*stepm' months (i.e. until
1.218 brouard 3101: age (in years) age+nhstepm*hstepm*stepm/12) by multiplying
3102: nhstepm*hstepm matrices.
3103: Output is stored in matrix po[i][j][h] for h every 'hstepm' step
3104: (typically every 2 years instead of every month which is too big
1.217 brouard 3105: for the memory).
1.218 brouard 3106: Model is determined by parameters x and covariates have to be
3107: included manually here.
1.217 brouard 3108:
1.222 brouard 3109: */
1.217 brouard 3110:
3111: int i, j, d, h, k;
3112: double **out, cov[NCOVMAX+1];
3113: double **newm;
3114: double agexact;
3115: double agebegin, ageend;
1.222 brouard 3116: double **oldm, **savm;
1.217 brouard 3117:
1.222 brouard 3118: oldm=oldms;savm=savms;
1.217 brouard 3119: /* Hstepm could be zero and should return the unit matrix */
3120: for (i=1;i<=nlstate+ndeath;i++)
3121: for (j=1;j<=nlstate+ndeath;j++){
3122: oldm[i][j]=(i==j ? 1.0 : 0.0);
3123: po[i][j][0]=(i==j ? 1.0 : 0.0);
3124: }
3125: /* Even if hstepm = 1, at least one multiplication by the unit matrix */
3126: for(h=1; h <=nhstepm; h++){
3127: for(d=1; d <=hstepm; d++){
3128: newm=savm;
3129: /* Covariates have to be included here again */
3130: cov[1]=1.;
3131: agexact=age-((h-1)*hstepm + (d-1))*stepm/YEARM; /* age just before transition */
3132: /* agexact=age+((h-1)*hstepm + (d-1))*stepm/YEARM; /\* age just before transition *\/ */
3133: cov[2]=agexact;
3134: if(nagesqr==1)
1.222 brouard 3135: cov[3]= agexact*agexact;
1.218 brouard 3136: for (k=1; k<=cptcovn;k++)
1.222 brouard 3137: cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(ij,k)];
3138: /* cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(ij,Tvar[k])]; */
1.217 brouard 3139: for (k=1; k<=cptcovage;k++) /* Should start at cptcovn+1 */
1.222 brouard 3140: /* cov[2+Tage[k]]=cov[2+Tage[k]]*cov[2]; */
3141: cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2];
3142: /* cov[2+nagesqr+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,Tvar[Tage[k]])]*cov[2]; */
1.217 brouard 3143: for (k=1; k<=cptcovprod;k++) /* Useless because included in cptcovn */
1.222 brouard 3144: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)]*nbcode[Tvard[k][2]][codtabm(ij,k)];
3145: /* 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 3146:
3147:
1.217 brouard 3148: /*printf("hxi cptcov=%d cptcode=%d\n",cptcov,cptcode);*/
3149: /*printf("h=%d d=%d age=%f cov=%f\n",h,d,age,cov[2]);*/
1.218 brouard 3150: /* Careful transposed matrix */
1.222 brouard 3151: /* age is in cov[2] */
1.218 brouard 3152: /* out=matprod2(newm, bmij(pmmij,cov,ncovmodel,x,nlstate,prevacurrent, dnewm, doldm, dsavm,ij),\ */
1.222 brouard 3153: /* 1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, oldm); */
1.218 brouard 3154: out=matprod2(newm, bmij(pmmij,cov,ncovmodel,x,nlstate,prevacurrent,ij),\
1.222 brouard 3155: 1,nlstate+ndeath,1,nlstate+ndeath,1,nlstate+ndeath, oldm);
1.217 brouard 3156: /* if((int)age == 70){ */
3157: /* printf(" Backward hbxij age=%d agexact=%f d=%d nhstepm=%d hstepm=%d\n", (int) age, agexact, d, nhstepm, hstepm); */
3158: /* for(i=1; i<=nlstate+ndeath; i++) { */
3159: /* printf("%d pmmij ",i); */
3160: /* for(j=1;j<=nlstate+ndeath;j++) { */
3161: /* printf("%f ",pmmij[i][j]); */
3162: /* } */
3163: /* printf(" oldm "); */
3164: /* for(j=1;j<=nlstate+ndeath;j++) { */
3165: /* printf("%f ",oldm[i][j]); */
3166: /* } */
3167: /* printf("\n"); */
3168: /* } */
3169: /* } */
3170: savm=oldm;
3171: oldm=newm;
3172: }
3173: for(i=1; i<=nlstate+ndeath; i++)
3174: for(j=1;j<=nlstate+ndeath;j++) {
1.222 brouard 3175: po[i][j][h]=newm[i][j];
3176: /*if(h==nhstepm) printf("po[%d][%d][%d]=%f ",i,j,h,po[i][j][h]);*/
1.217 brouard 3177: }
3178: /*printf("h=%d ",h);*/
3179: } /* end h */
1.222 brouard 3180: /* printf("\n H=%d \n",h); */
1.217 brouard 3181: return po;
3182: }
3183:
3184:
1.162 brouard 3185: #ifdef NLOPT
3186: double myfunc(unsigned n, const double *p1, double *grad, void *pd){
3187: double fret;
3188: double *xt;
3189: int j;
3190: myfunc_data *d2 = (myfunc_data *) pd;
3191: /* xt = (p1-1); */
3192: xt=vector(1,n);
3193: for (j=1;j<=n;j++) xt[j]=p1[j-1]; /* xt[1]=p1[0] */
3194:
3195: fret=(d2->function)(xt); /* p xt[1]@8 is fine */
3196: /* fret=(*func)(xt); /\* p xt[1]@8 is fine *\/ */
3197: printf("Function = %.12lf ",fret);
3198: for (j=1;j<=n;j++) printf(" %d %.8lf", j, xt[j]);
3199: printf("\n");
3200: free_vector(xt,1,n);
3201: return fret;
3202: }
3203: #endif
1.126 brouard 3204:
3205: /*************** log-likelihood *************/
3206: double func( double *x)
3207: {
1.226 brouard 3208: int i, ii, j, k, mi, d, kk;
3209: int ioffset=0;
3210: double l, ll[NLSTATEMAX+1], cov[NCOVMAX+1];
3211: double **out;
3212: double lli; /* Individual log likelihood */
3213: int s1, s2;
1.228 brouard 3214: 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 3215: double bbh, survp;
3216: long ipmx;
3217: double agexact;
3218: /*extern weight */
3219: /* We are differentiating ll according to initial status */
3220: /* for (i=1;i<=npar;i++) printf("%f ", x[i]);*/
3221: /*for(i=1;i<imx;i++)
3222: printf(" %d\n",s[4][i]);
3223: */
1.162 brouard 3224:
1.226 brouard 3225: ++countcallfunc;
1.162 brouard 3226:
1.226 brouard 3227: cov[1]=1.;
1.126 brouard 3228:
1.226 brouard 3229: for(k=1; k<=nlstate; k++) ll[k]=0.;
1.224 brouard 3230: ioffset=0;
1.226 brouard 3231: if(mle==1){
3232: for (i=1,ipmx=0, sw=0.; i<=imx; i++){
3233: /* Computes the values of the ncovmodel covariates of the model
3234: depending if the covariates are fixed or varying (age dependent) and stores them in cov[]
3235: Then computes with function pmij which return a matrix p[i][j] giving the elementary probability
3236: to be observed in j being in i according to the model.
3237: */
1.243 brouard 3238: ioffset=2+nagesqr ;
1.233 brouard 3239: /* Fixed */
1.234 brouard 3240: for (k=1; k<=ncovf;k++){ /* Simple and product fixed covariates without age* products */
3241: 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)*/
3242: }
1.226 brouard 3243: /* In model V2+V1*V4+age*V3+V3*V2 Tvar[1] is V2, Tvar[2=V1*V4]
3244: is 6, Tvar[3=age*V3] should not be computed because of age Tvar[4=V3*V2]
3245: has been calculated etc */
3246: /* For an individual i, wav[i] gives the number of effective waves */
3247: /* We compute the contribution to Likelihood of each effective transition
3248: mw[mi][i] is real wave of the mi th effectve wave */
3249: /* Then statuses are computed at each begin and end of an effective wave s1=s[ mw[mi][i] ][i];
3250: s2=s[mw[mi+1][i]][i];
3251: And the iv th varying covariate is the cotvar[mw[mi+1][i]][iv][i]
3252: But if the variable is not in the model TTvar[iv] is the real variable effective in the model:
3253: meaning that decodemodel should be used cotvar[mw[mi+1][i]][TTvar[iv]][i]
3254: */
3255: for(mi=1; mi<= wav[i]-1; mi++){
1.234 brouard 3256: for(k=1; k <= ncovv ; k++){ /* Varying covariates (single and product but no age )*/
1.242 brouard 3257: /* cov[ioffset+TvarVind[k]]=cotvar[mw[mi][i]][Tvar[TvarVind[k]]][i]; */
3258: cov[ioffset+TvarVind[k]]=cotvar[mw[mi][i]][Tvar[TvarVind[k]]-ncovcol-nqv][i];
1.234 brouard 3259: }
3260: for (ii=1;ii<=nlstate+ndeath;ii++)
3261: for (j=1;j<=nlstate+ndeath;j++){
3262: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
3263: savm[ii][j]=(ii==j ? 1.0 : 0.0);
3264: }
3265: for(d=0; d<dh[mi][i]; d++){
3266: newm=savm;
3267: agexact=agev[mw[mi][i]][i]+d*stepm/YEARM;
3268: cov[2]=agexact;
3269: if(nagesqr==1)
3270: cov[3]= agexact*agexact; /* Should be changed here */
3271: for (kk=1; kk<=cptcovage;kk++) {
1.242 brouard 3272: if(!FixedV[Tvar[Tage[kk]]])
1.234 brouard 3273: cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact; /* Tage[kk] gives the data-covariate associated with age */
1.242 brouard 3274: else
3275: cov[Tage[kk]+2+nagesqr]=cotvar[mw[mi][i]][Tvar[Tage[kk]]-ncovcol-nqv][i]*agexact;
1.234 brouard 3276: }
3277: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,
3278: 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate));
3279: savm=oldm;
3280: oldm=newm;
3281: } /* end mult */
3282:
3283: /*lli=log(out[s[mw[mi][i]][i]][s[mw[mi+1][i]][i]]);*/ /* Original formula */
3284: /* But now since version 0.9 we anticipate for bias at large stepm.
3285: * If stepm is larger than one month (smallest stepm) and if the exact delay
3286: * (in months) between two waves is not a multiple of stepm, we rounded to
3287: * the nearest (and in case of equal distance, to the lowest) interval but now
3288: * we keep into memory the bias bh[mi][i] and also the previous matrix product
3289: * (i.e to dh[mi][i]-1) saved in 'savm'. Then we inter(extra)polate the
3290: * probability in order to take into account the bias as a fraction of the way
1.231 brouard 3291: * from savm to out if bh is negative or even beyond if bh is positive. bh varies
3292: * -stepm/2 to stepm/2 .
3293: * For stepm=1 the results are the same as for previous versions of Imach.
3294: * For stepm > 1 the results are less biased than in previous versions.
3295: */
1.234 brouard 3296: s1=s[mw[mi][i]][i];
3297: s2=s[mw[mi+1][i]][i];
3298: bbh=(double)bh[mi][i]/(double)stepm;
3299: /* bias bh is positive if real duration
3300: * is higher than the multiple of stepm and negative otherwise.
3301: */
3302: /* lli= (savm[s1][s2]>1.e-8 ?(1.+bbh)*log(out[s1][s2])- bbh*log(savm[s1][s2]):log((1.+bbh)*out[s1][s2]));*/
3303: if( s2 > nlstate){
3304: /* i.e. if s2 is a death state and if the date of death is known
3305: then the contribution to the likelihood is the probability to
3306: die between last step unit time and current step unit time,
3307: which is also equal to probability to die before dh
3308: minus probability to die before dh-stepm .
3309: In version up to 0.92 likelihood was computed
3310: as if date of death was unknown. Death was treated as any other
3311: health state: the date of the interview describes the actual state
3312: and not the date of a change in health state. The former idea was
3313: to consider that at each interview the state was recorded
3314: (healthy, disable or death) and IMaCh was corrected; but when we
3315: introduced the exact date of death then we should have modified
3316: the contribution of an exact death to the likelihood. This new
3317: contribution is smaller and very dependent of the step unit
3318: stepm. It is no more the probability to die between last interview
3319: and month of death but the probability to survive from last
3320: interview up to one month before death multiplied by the
3321: probability to die within a month. Thanks to Chris
3322: Jackson for correcting this bug. Former versions increased
3323: mortality artificially. The bad side is that we add another loop
3324: which slows down the processing. The difference can be up to 10%
3325: lower mortality.
3326: */
3327: /* If, at the beginning of the maximization mostly, the
3328: cumulative probability or probability to be dead is
3329: constant (ie = 1) over time d, the difference is equal to
3330: 0. out[s1][3] = savm[s1][3]: probability, being at state
3331: s1 at precedent wave, to be dead a month before current
3332: wave is equal to probability, being at state s1 at
3333: precedent wave, to be dead at mont of the current
3334: wave. Then the observed probability (that this person died)
3335: is null according to current estimated parameter. In fact,
3336: it should be very low but not zero otherwise the log go to
3337: infinity.
3338: */
1.183 brouard 3339: /* #ifdef INFINITYORIGINAL */
3340: /* lli=log(out[s1][s2] - savm[s1][s2]); */
3341: /* #else */
3342: /* if ((out[s1][s2] - savm[s1][s2]) < mytinydouble) */
3343: /* lli=log(mytinydouble); */
3344: /* else */
3345: /* lli=log(out[s1][s2] - savm[s1][s2]); */
3346: /* #endif */
1.226 brouard 3347: lli=log(out[s1][s2] - savm[s1][s2]);
1.216 brouard 3348:
1.226 brouard 3349: } else if ( s2==-1 ) { /* alive */
3350: for (j=1,survp=0. ; j<=nlstate; j++)
3351: survp += (1.+bbh)*out[s1][j]- bbh*savm[s1][j];
3352: /*survp += out[s1][j]; */
3353: lli= log(survp);
3354: }
3355: else if (s2==-4) {
3356: for (j=3,survp=0. ; j<=nlstate; j++)
3357: survp += (1.+bbh)*out[s1][j]- bbh*savm[s1][j];
3358: lli= log(survp);
3359: }
3360: else if (s2==-5) {
3361: for (j=1,survp=0. ; j<=2; j++)
3362: survp += (1.+bbh)*out[s1][j]- bbh*savm[s1][j];
3363: lli= log(survp);
3364: }
3365: else{
3366: lli= log((1.+bbh)*out[s1][s2]- bbh*savm[s1][s2]); /* linear interpolation */
3367: /* 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 */
3368: }
3369: /*lli=(1.+bbh)*log(out[s1][s2])- bbh*log(savm[s1][s2]);*/
3370: /*if(lli ==000.0)*/
3371: /*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); */
3372: ipmx +=1;
3373: sw += weight[i];
3374: ll[s[mw[mi][i]][i]] += 2*weight[i]*lli;
3375: /* if (lli < log(mytinydouble)){ */
3376: /* 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); */
3377: /* 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]); */
3378: /* } */
3379: } /* end of wave */
3380: } /* end of individual */
3381: } else if(mle==2){
3382: for (i=1,ipmx=0, sw=0.; i<=imx; i++){
3383: for (k=1; k<=cptcovn;k++) cov[2+nagesqr+k]=covar[Tvar[k]][i];
3384: for(mi=1; mi<= wav[i]-1; mi++){
3385: for (ii=1;ii<=nlstate+ndeath;ii++)
3386: for (j=1;j<=nlstate+ndeath;j++){
3387: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
3388: savm[ii][j]=(ii==j ? 1.0 : 0.0);
3389: }
3390: for(d=0; d<=dh[mi][i]; d++){
3391: newm=savm;
3392: agexact=agev[mw[mi][i]][i]+d*stepm/YEARM;
3393: cov[2]=agexact;
3394: if(nagesqr==1)
3395: cov[3]= agexact*agexact;
3396: for (kk=1; kk<=cptcovage;kk++) {
3397: cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact;
3398: }
3399: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,
3400: 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate));
3401: savm=oldm;
3402: oldm=newm;
3403: } /* end mult */
3404:
3405: s1=s[mw[mi][i]][i];
3406: s2=s[mw[mi+1][i]][i];
3407: bbh=(double)bh[mi][i]/(double)stepm;
3408: 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 */
3409: ipmx +=1;
3410: sw += weight[i];
3411: ll[s[mw[mi][i]][i]] += 2*weight[i]*lli;
3412: } /* end of wave */
3413: } /* end of individual */
3414: } else if(mle==3){ /* exponential inter-extrapolation */
3415: for (i=1,ipmx=0, sw=0.; i<=imx; i++){
3416: for (k=1; k<=cptcovn;k++) cov[2+nagesqr+k]=covar[Tvar[k]][i];
3417: for(mi=1; mi<= wav[i]-1; mi++){
3418: for (ii=1;ii<=nlstate+ndeath;ii++)
3419: for (j=1;j<=nlstate+ndeath;j++){
3420: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
3421: savm[ii][j]=(ii==j ? 1.0 : 0.0);
3422: }
3423: for(d=0; d<dh[mi][i]; d++){
3424: newm=savm;
3425: agexact=agev[mw[mi][i]][i]+d*stepm/YEARM;
3426: cov[2]=agexact;
3427: if(nagesqr==1)
3428: cov[3]= agexact*agexact;
3429: for (kk=1; kk<=cptcovage;kk++) {
3430: cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact;
3431: }
3432: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,
3433: 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate));
3434: savm=oldm;
3435: oldm=newm;
3436: } /* end mult */
3437:
3438: s1=s[mw[mi][i]][i];
3439: s2=s[mw[mi+1][i]][i];
3440: bbh=(double)bh[mi][i]/(double)stepm;
3441: 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 */
3442: ipmx +=1;
3443: sw += weight[i];
3444: ll[s[mw[mi][i]][i]] += 2*weight[i]*lli;
3445: } /* end of wave */
3446: } /* end of individual */
3447: }else if (mle==4){ /* ml=4 no inter-extrapolation */
3448: for (i=1,ipmx=0, sw=0.; i<=imx; i++){
3449: for (k=1; k<=cptcovn;k++) cov[2+nagesqr+k]=covar[Tvar[k]][i];
3450: for(mi=1; mi<= wav[i]-1; mi++){
3451: for (ii=1;ii<=nlstate+ndeath;ii++)
3452: for (j=1;j<=nlstate+ndeath;j++){
3453: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
3454: savm[ii][j]=(ii==j ? 1.0 : 0.0);
3455: }
3456: for(d=0; d<dh[mi][i]; d++){
3457: newm=savm;
3458: agexact=agev[mw[mi][i]][i]+d*stepm/YEARM;
3459: cov[2]=agexact;
3460: if(nagesqr==1)
3461: cov[3]= agexact*agexact;
3462: for (kk=1; kk<=cptcovage;kk++) {
3463: cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact;
3464: }
1.126 brouard 3465:
1.226 brouard 3466: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,
3467: 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate));
3468: savm=oldm;
3469: oldm=newm;
3470: } /* end mult */
3471:
3472: s1=s[mw[mi][i]][i];
3473: s2=s[mw[mi+1][i]][i];
3474: if( s2 > nlstate){
3475: lli=log(out[s1][s2] - savm[s1][s2]);
3476: } else if ( s2==-1 ) { /* alive */
3477: for (j=1,survp=0. ; j<=nlstate; j++)
3478: survp += out[s1][j];
3479: lli= log(survp);
3480: }else{
3481: lli=log(out[s[mw[mi][i]][i]][s[mw[mi+1][i]][i]]); /* Original formula */
3482: }
3483: ipmx +=1;
3484: sw += weight[i];
3485: ll[s[mw[mi][i]][i]] += 2*weight[i]*lli;
1.126 brouard 3486: /* 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 3487: } /* end of wave */
3488: } /* end of individual */
3489: }else{ /* ml=5 no inter-extrapolation no jackson =0.8a */
3490: for (i=1,ipmx=0, sw=0.; i<=imx; i++){
3491: for (k=1; k<=cptcovn;k++) cov[2+nagesqr+k]=covar[Tvar[k]][i];
3492: for(mi=1; mi<= wav[i]-1; mi++){
3493: for (ii=1;ii<=nlstate+ndeath;ii++)
3494: for (j=1;j<=nlstate+ndeath;j++){
3495: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
3496: savm[ii][j]=(ii==j ? 1.0 : 0.0);
3497: }
3498: for(d=0; d<dh[mi][i]; d++){
3499: newm=savm;
3500: agexact=agev[mw[mi][i]][i]+d*stepm/YEARM;
3501: cov[2]=agexact;
3502: if(nagesqr==1)
3503: cov[3]= agexact*agexact;
3504: for (kk=1; kk<=cptcovage;kk++) {
3505: cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact;
3506: }
1.126 brouard 3507:
1.226 brouard 3508: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,
3509: 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate));
3510: savm=oldm;
3511: oldm=newm;
3512: } /* end mult */
3513:
3514: s1=s[mw[mi][i]][i];
3515: s2=s[mw[mi+1][i]][i];
3516: lli=log(out[s[mw[mi][i]][i]][s[mw[mi+1][i]][i]]); /* Original formula */
3517: ipmx +=1;
3518: sw += weight[i];
3519: ll[s[mw[mi][i]][i]] += 2*weight[i]*lli;
3520: /*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]);*/
3521: } /* end of wave */
3522: } /* end of individual */
3523: } /* End of if */
3524: for(k=1,l=0.; k<=nlstate; k++) l += ll[k];
3525: /* printf("l1=%f l2=%f ",ll[1],ll[2]); */
3526: l= l*ipmx/sw; /* To get the same order of magnitude as if weight=1 for every body */
3527: return -l;
1.126 brouard 3528: }
3529:
3530: /*************** log-likelihood *************/
3531: double funcone( double *x)
3532: {
1.228 brouard 3533: /* Same as func but slower because of a lot of printf and if */
1.126 brouard 3534: int i, ii, j, k, mi, d, kk;
1.228 brouard 3535: int ioffset=0;
1.131 brouard 3536: double l, ll[NLSTATEMAX+1], cov[NCOVMAX+1];
1.126 brouard 3537: double **out;
3538: double lli; /* Individual log likelihood */
3539: double llt;
3540: int s1, s2;
1.228 brouard 3541: int iv=0, iqv=0, itv=0, iqtv=0 ; /* Index of varying covariate, fixed quantitative cov, time varying covariate, quantitative time varying covariate */
3542:
1.126 brouard 3543: double bbh, survp;
1.187 brouard 3544: double agexact;
1.214 brouard 3545: double agebegin, ageend;
1.126 brouard 3546: /*extern weight */
3547: /* We are differentiating ll according to initial status */
3548: /* for (i=1;i<=npar;i++) printf("%f ", x[i]);*/
3549: /*for(i=1;i<imx;i++)
3550: printf(" %d\n",s[4][i]);
3551: */
3552: cov[1]=1.;
3553:
3554: for(k=1; k<=nlstate; k++) ll[k]=0.;
1.224 brouard 3555: ioffset=0;
3556: for (i=1,ipmx=0, sw=0.; i<=imx; i++){
1.243 brouard 3557: /* ioffset=2+nagesqr+cptcovage; */
3558: ioffset=2+nagesqr;
1.232 brouard 3559: /* Fixed */
1.224 brouard 3560: /* for (k=1; k<=cptcovn;k++) cov[2+nagesqr+k]=covar[Tvar[k]][i]; */
1.232 brouard 3561: /* for (k=1; k<=ncoveff;k++){ /\* Simple and product fixed Dummy covariates without age* products *\/ */
3562: for (k=1; k<=ncovf;k++){ /* Simple and product fixed covariates without age* products */
3563: 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)*/
3564: /* cov[ioffset+TvarFind[1]]=covar[Tvar[TvarFind[1]]][i]; */
3565: /* cov[2+6]=covar[Tvar[6]][i]; */
3566: /* cov[2+6]=covar[2][i]; V2 */
3567: /* cov[TvarFind[2]]=covar[Tvar[TvarFind[2]]][i]; */
3568: /* cov[2+7]=covar[Tvar[7]][i]; */
3569: /* cov[2+7]=covar[7][i]; V7=V1*V2 */
3570: /* cov[TvarFind[3]]=covar[Tvar[TvarFind[3]]][i]; */
3571: /* cov[2+9]=covar[Tvar[9]][i]; */
3572: /* cov[2+9]=covar[1][i]; V1 */
1.225 brouard 3573: }
1.232 brouard 3574: /* for (k=1; k<=nqfveff;k++){ /\* Simple and product fixed Quantitative covariates without age* products *\/ */
3575: /* 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?)*\/ */
3576: /* } */
1.231 brouard 3577: /* for(iqv=1; iqv <= nqfveff; iqv++){ /\* Quantitative fixed covariates *\/ */
3578: /* cov[++ioffset]=coqvar[Tvar[iqv]][i]; /\* Only V2 k=6 and V1*V2 7 *\/ */
3579: /* } */
1.225 brouard 3580:
1.233 brouard 3581:
3582: for(mi=1; mi<= wav[i]-1; mi++){ /* Varying with waves */
1.232 brouard 3583: /* Wave varying (but not age varying) */
3584: for(k=1; k <= ncovv ; k++){ /* Varying covariates (single and product but no age )*/
1.242 brouard 3585: /* cov[ioffset+TvarVind[k]]=cotvar[mw[mi][i]][Tvar[TvarVind[k]]][i]; */
3586: cov[ioffset+TvarVind[k]]=cotvar[mw[mi][i]][Tvar[TvarVind[k]]-ncovcol-nqv][i];
3587: }
1.232 brouard 3588: /* for(itv=1; itv <= ntveff; itv++){ /\* Varying dummy covariates (single??)*\/ */
1.242 brouard 3589: /* iv= Tvar[Tmodelind[ioffset-2-nagesqr-cptcovage+itv]]-ncovcol-nqv; /\* Counting the # varying covariate from 1 to ntveff *\/ */
3590: /* cov[ioffset+iv]=cotvar[mw[mi][i]][iv][i]; */
3591: /* k=ioffset-2-nagesqr-cptcovage+itv; /\* position in simple model *\/ */
3592: /* cov[ioffset+itv]=cotvar[mw[mi][i]][TmodelInvind[itv]][i]; */
3593: /* 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 3594: /* for(iqtv=1; iqtv <= nqtveff; iqtv++){ /\* Varying quantitatives covariates *\/ */
1.242 brouard 3595: /* iv=TmodelInvQind[iqtv]; /\* Counting the # varying covariate from 1 to ntveff *\/ */
3596: /* /\* 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]); *\/ */
3597: /* cov[ioffset+ntveff+iqtv]=cotqvar[mw[mi][i]][TmodelInvQind[iqtv]][i]; */
1.232 brouard 3598: /* } */
1.126 brouard 3599: for (ii=1;ii<=nlstate+ndeath;ii++)
1.242 brouard 3600: for (j=1;j<=nlstate+ndeath;j++){
3601: oldm[ii][j]=(ii==j ? 1.0 : 0.0);
3602: savm[ii][j]=(ii==j ? 1.0 : 0.0);
3603: }
1.214 brouard 3604:
3605: agebegin=agev[mw[mi][i]][i]; /* Age at beginning of effective wave */
3606: ageend=agev[mw[mi][i]][i] + (dh[mi][i])*stepm/YEARM; /* Age at end of effective wave and at the end of transition */
3607: for(d=0; d<dh[mi][i]; d++){ /* Delay between two effective waves */
1.247 brouard 3608: /* for(d=0; d<=0; d++){ /\* Delay between two effective waves Only one matrix to speed up*\/ */
1.242 brouard 3609: /*dh[m][i] or dh[mw[mi][i]][i] is the delay between two effective waves m=mw[mi][i]
3610: and mw[mi+1][i]. dh depends on stepm.*/
3611: newm=savm;
1.247 brouard 3612: agexact=agev[mw[mi][i]][i]+d*stepm/YEARM; /* Here d is needed */
1.242 brouard 3613: cov[2]=agexact;
3614: if(nagesqr==1)
3615: cov[3]= agexact*agexact;
3616: for (kk=1; kk<=cptcovage;kk++) {
3617: if(!FixedV[Tvar[Tage[kk]]])
3618: cov[Tage[kk]+2+nagesqr]=covar[Tvar[Tage[kk]]][i]*agexact;
3619: else
3620: cov[Tage[kk]+2+nagesqr]=cotvar[mw[mi][i]][Tvar[Tage[kk]]-ncovcol-nqv][i]*agexact;
3621: }
3622: /* printf("i=%d,mi=%d,d=%d,mw[mi][i]=%d\n",i, mi,d,mw[mi][i]); */
3623: /* savm=pmij(pmmij,cov,ncovmodel,x,nlstate); */
3624: out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath,
3625: 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate));
3626: /* out=matprod2(newm,oldm,1,nlstate+ndeath,1,nlstate+ndeath, */
3627: /* 1,nlstate+ndeath,pmij(pmmij,cov,ncovmodel,x,nlstate)); */
3628: savm=oldm;
3629: oldm=newm;
1.126 brouard 3630: } /* end mult */
3631:
3632: s1=s[mw[mi][i]][i];
3633: s2=s[mw[mi+1][i]][i];
1.217 brouard 3634: /* if(s2==-1){ */
3635: /* printf(" s1=%d, s2=%d i=%d \n", s1, s2, i); */
3636: /* /\* exit(1); *\/ */
3637: /* } */
1.126 brouard 3638: bbh=(double)bh[mi][i]/(double)stepm;
3639: /* bias is positive if real duration
3640: * is higher than the multiple of stepm and negative otherwise.
3641: */
3642: if( s2 > nlstate && (mle <5) ){ /* Jackson */
1.242 brouard 3643: lli=log(out[s1][s2] - savm[s1][s2]);
1.216 brouard 3644: } else if ( s2==-1 ) { /* alive */
1.242 brouard 3645: for (j=1,survp=0. ; j<=nlstate; j++)
3646: survp += (1.+bbh)*out[s1][j]- bbh*savm[s1][j];
3647: lli= log(survp);
1.126 brouard 3648: }else if (mle==1){
1.242 brouard 3649: lli= log((1.+bbh)*out[s1][s2]- bbh*savm[s1][s2]); /* linear interpolation */
1.126 brouard 3650: } else if(mle==2){
1.242 brouard 3651: 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 3652: } else if(mle==3){ /* exponential inter-extrapolation */
1.242 brouard 3653: 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 3654: } else if (mle==4){ /* mle=4 no inter-extrapolation */
1.242 brouard 3655: lli=log(out[s1][s2]); /* Original formula */
1.136 brouard 3656: } else{ /* mle=0 back to 1 */
1.242 brouard 3657: lli= log((1.+bbh)*out[s1][s2]- bbh*savm[s1][s2]); /* linear interpolation */
3658: /*lli=log(out[s1][s2]); */ /* Original formula */
1.126 brouard 3659: } /* End of if */
3660: ipmx +=1;
3661: sw += weight[i];
3662: ll[s[mw[mi][i]][i]] += 2*weight[i]*lli;
1.132 brouard 3663: /*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 3664: if(globpr){
1.246 brouard 3665: fprintf(ficresilk,"%09ld %6.1f %6.1f %6d %2d %2d %2d %2d %3d %15.6f %8.4f %8.3f\
1.126 brouard 3666: %11.6f %11.6f %11.6f ", \
1.242 brouard 3667: num[i], agebegin, ageend, i,s1,s2,mi,mw[mi][i],dh[mi][i],exp(lli),weight[i],weight[i]*gipmx/gsw,
3668: 2*weight[i]*lli,out[s1][s2],savm[s1][s2]);
3669: for(k=1,llt=0.,l=0.; k<=nlstate; k++){
3670: llt +=ll[k]*gipmx/gsw;
3671: fprintf(ficresilk," %10.6f",-ll[k]*gipmx/gsw);
3672: }
3673: fprintf(ficresilk," %10.6f\n", -llt);
1.126 brouard 3674: }
1.232 brouard 3675: } /* end of wave */
3676: } /* end of individual */
3677: for(k=1,l=0.; k<=nlstate; k++) l += ll[k];
3678: /* printf("l1=%f l2=%f ",ll[1],ll[2]); */
3679: l= l*ipmx/sw; /* To get the same order of magnitude as if weight=1 for every body */
3680: if(globpr==0){ /* First time we count the contributions and weights */
3681: gipmx=ipmx;
3682: gsw=sw;
3683: }
3684: return -l;
1.126 brouard 3685: }
3686:
3687:
3688: /*************** function likelione ***********/
3689: void likelione(FILE *ficres,double p[], int npar, int nlstate, int *globpri, long *ipmx, double *sw, double *fretone, double (*funcone)(double []))
3690: {
3691: /* This routine should help understanding what is done with
3692: the selection of individuals/waves and
3693: to check the exact contribution to the likelihood.
3694: Plotting could be done.
3695: */
3696: int k;
3697:
3698: if(*globpri !=0){ /* Just counts and sums, no printings */
1.201 brouard 3699: strcpy(fileresilk,"ILK_");
1.202 brouard 3700: strcat(fileresilk,fileresu);
1.126 brouard 3701: if((ficresilk=fopen(fileresilk,"w"))==NULL) {
3702: printf("Problem with resultfile: %s\n", fileresilk);
3703: fprintf(ficlog,"Problem with resultfile: %s\n", fileresilk);
3704: }
1.214 brouard 3705: 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");
3706: fprintf(ficresilk, "#num_i ageb agend i s1 s2 mi mw dh likeli weight %%weight 2wlli out sav ");
1.126 brouard 3707: /* i,s1,s2,mi,mw[mi][i],dh[mi][i],exp(lli),weight[i],2*weight[i]*lli,out[s1][s2],savm[s1][s2]); */
3708: for(k=1; k<=nlstate; k++)
3709: fprintf(ficresilk," -2*gipw/gsw*weight*ll[%d]++",k);
3710: fprintf(ficresilk," -2*gipw/gsw*weight*ll(total)\n");
3711: }
3712:
3713: *fretone=(*funcone)(p);
3714: if(*globpri !=0){
3715: fclose(ficresilk);
1.205 brouard 3716: if (mle ==0)
3717: fprintf(fichtm,"\n<br>File of contributions to the likelihood computed with initial parameters and mle = %d.",mle);
3718: else if(mle >=1)
3719: fprintf(fichtm,"\n<br>File of contributions to the likelihood computed with optimized parameters mle = %d.",mle);
3720: 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 3721:
1.208 brouard 3722:
3723: for (k=1; k<= nlstate ; k++) {
1.211 brouard 3724: 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 3725: <img src=\"%s-p%dj.png\">",k,k,subdirf2(optionfilefiname,"ILK_"),k,subdirf2(optionfilefiname,"ILK_"),k,subdirf2(optionfilefiname,"ILK_"),k);
3726: }
1.207 brouard 3727: 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 3728: <img src=\"%s-ori.png\">",subdirf2(optionfilefiname,"ILK_"),subdirf2(optionfilefiname,"ILK_"),subdirf2(optionfilefiname,"ILK_"));
1.207 brouard 3729: fprintf(fichtm,"<br>- and by state of destination <a href=\"%s-dest.png\">%s-dest.png</a><br> \
1.204 brouard 3730: <img src=\"%s-dest.png\">",subdirf2(optionfilefiname,"ILK_"),subdirf2(optionfilefiname,"ILK_"),subdirf2(optionfilefiname,"ILK_"));
1.207 brouard 3731: fflush(fichtm);
1.205 brouard 3732: }
1.126 brouard 3733: return;
3734: }
3735:
3736:
3737: /*********** Maximum Likelihood Estimation ***************/
3738:
3739: void mlikeli(FILE *ficres,double p[], int npar, int ncovmodel, int nlstate, double ftol, double (*func)(double []))
3740: {
1.165 brouard 3741: int i,j, iter=0;
1.126 brouard 3742: double **xi;
3743: double fret;
3744: double fretone; /* Only one call to likelihood */
3745: /* char filerespow[FILENAMELENGTH];*/
1.162 brouard 3746:
3747: #ifdef NLOPT
3748: int creturn;
3749: nlopt_opt opt;
3750: /* double lb[9] = { -HUGE_VAL, -HUGE_VAL, -HUGE_VAL, -HUGE_VAL, -HUGE_VAL, -HUGE_VAL, -HUGE_VAL, -HUGE_VAL, -HUGE_VAL }; /\* lower bounds *\/ */
3751: double *lb;
3752: double minf; /* the minimum objective value, upon return */
3753: double * p1; /* Shifted parameters from 0 instead of 1 */
3754: myfunc_data dinst, *d = &dinst;
3755: #endif
3756:
3757:
1.126 brouard 3758: xi=matrix(1,npar,1,npar);
3759: for (i=1;i<=npar;i++)
3760: for (j=1;j<=npar;j++)
3761: xi[i][j]=(i==j ? 1.0 : 0.0);
3762: printf("Powell\n"); fprintf(ficlog,"Powell\n");
1.201 brouard 3763: strcpy(filerespow,"POW_");
1.126 brouard 3764: strcat(filerespow,fileres);
3765: if((ficrespow=fopen(filerespow,"w"))==NULL) {
3766: printf("Problem with resultfile: %s\n", filerespow);
3767: fprintf(ficlog,"Problem with resultfile: %s\n", filerespow);
3768: }
3769: fprintf(ficrespow,"# Powell\n# iter -2*LL");
3770: for (i=1;i<=nlstate;i++)
3771: for(j=1;j<=nlstate+ndeath;j++)
3772: if(j!=i)fprintf(ficrespow," p%1d%1d",i,j);
3773: fprintf(ficrespow,"\n");
1.162 brouard 3774: #ifdef POWELL
1.126 brouard 3775: powell(p,xi,npar,ftol,&iter,&fret,func);
1.162 brouard 3776: #endif
1.126 brouard 3777:
1.162 brouard 3778: #ifdef NLOPT
3779: #ifdef NEWUOA
3780: opt = nlopt_create(NLOPT_LN_NEWUOA,npar);
3781: #else
3782: opt = nlopt_create(NLOPT_LN_BOBYQA,npar);
3783: #endif
3784: lb=vector(0,npar-1);
3785: for (i=0;i<npar;i++) lb[i]= -HUGE_VAL;
3786: nlopt_set_lower_bounds(opt, lb);
3787: nlopt_set_initial_step1(opt, 0.1);
3788:
3789: p1= (p+1); /* p *(p+1)@8 and p *(p1)@8 are equal p1[0]=p[1] */
3790: d->function = func;
3791: printf(" Func %.12lf \n",myfunc(npar,p1,NULL,d));
3792: nlopt_set_min_objective(opt, myfunc, d);
3793: nlopt_set_xtol_rel(opt, ftol);
3794: if ((creturn=nlopt_optimize(opt, p1, &minf)) < 0) {
3795: printf("nlopt failed! %d\n",creturn);
3796: }
3797: else {
3798: printf("found minimum after %d evaluations (NLOPT=%d)\n", countcallfunc ,NLOPT);
3799: printf("found minimum at f(%g,%g) = %0.10g\n", p[0], p[1], minf);
3800: iter=1; /* not equal */
3801: }
3802: nlopt_destroy(opt);
3803: #endif
1.126 brouard 3804: free_matrix(xi,1,npar,1,npar);
3805: fclose(ficrespow);
1.203 brouard 3806: printf("\n#Number of iterations & function calls = %d & %d, -2 Log likelihood = %.12f\n",iter, countcallfunc,func(p));
3807: fprintf(ficlog,"\n#Number of iterations & function calls = %d & %d, -2 Log likelihood = %.12f\n",iter, countcallfunc,func(p));
1.180 brouard 3808: fprintf(ficres,"#Number of iterations & function calls = %d & %d, -2 Log likelihood = %.12f\n",iter, countcallfunc,func(p));
1.126 brouard 3809:
3810: }
3811:
3812: /**** Computes Hessian and covariance matrix ***/
1.203 brouard 3813: void hesscov(double **matcov, double **hess, double p[], int npar, double delti[], double ftolhess, double (*func)(double []))
1.126 brouard 3814: {
3815: double **a,**y,*x,pd;
1.203 brouard 3816: /* double **hess; */
1.164 brouard 3817: int i, j;
1.126 brouard 3818: int *indx;
3819:
3820: double hessii(double p[], double delta, int theta, double delti[],double (*func)(double []),int npar);
1.203 brouard 3821: double hessij(double p[], double **hess, double delti[], int i, int j,double (*func)(double []),int npar);
1.126 brouard 3822: void lubksb(double **a, int npar, int *indx, double b[]) ;
3823: void ludcmp(double **a, int npar, int *indx, double *d) ;
3824: double gompertz(double p[]);
1.203 brouard 3825: /* hess=matrix(1,npar,1,npar); */
1.126 brouard 3826:
3827: printf("\nCalculation of the hessian matrix. Wait...\n");
3828: fprintf(ficlog,"\nCalculation of the hessian matrix. Wait...\n");
3829: for (i=1;i<=npar;i++){
1.203 brouard 3830: printf("%d-",i);fflush(stdout);
3831: fprintf(ficlog,"%d-",i);fflush(ficlog);
1.126 brouard 3832:
3833: hess[i][i]=hessii(p,ftolhess,i,delti,func,npar);
3834:
3835: /* printf(" %f ",p[i]);
3836: printf(" %lf %lf %lf",hess[i][i],ftolhess,delti[i]);*/
3837: }
3838:
3839: for (i=1;i<=npar;i++) {
3840: for (j=1;j<=npar;j++) {
3841: if (j>i) {
1.203 brouard 3842: printf(".%d-%d",i,j);fflush(stdout);
3843: fprintf(ficlog,".%d-%d",i,j);fflush(ficlog);
3844: hess[i][j]=hessij(p,hess, delti,i,j,func,npar);
1.126 brouard 3845:
3846: hess[j][i]=hess[i][j];
3847: /*printf(" %lf ",hess[i][j]);*/
3848: }
3849: }
3850: }
3851: printf("\n");
3852: fprintf(ficlog,"\n");
3853:
3854: printf("\nInverting the hessian to get the covariance matrix. Wait...\n");
3855: fprintf(ficlog,"\nInverting the hessian to get the covariance matrix. Wait...\n");
3856:
3857: a=matrix(1,npar,1,npar);
3858: y=matrix(1,npar,1,npar);
3859: x=vector(1,npar);
3860: indx=ivector(1,npar);
3861: for (i=1;i<=npar;i++)
3862: for (j=1;j<=npar;j++) a[i][j]=hess[i][j];
3863: ludcmp(a,npar,indx,&pd);
3864:
3865: for (j=1;j<=npar;j++) {
3866: for (i=1;i<=npar;i++) x[i]=0;
3867: x[j]=1;
3868: lubksb(a,npar,indx,x);
3869: for (i=1;i<=npar;i++){
3870: matcov[i][j]=x[i];
3871: }
3872: }
3873:
3874: printf("\n#Hessian matrix#\n");
3875: fprintf(ficlog,"\n#Hessian matrix#\n");
3876: for (i=1;i<=npar;i++) {
3877: for (j=1;j<=npar;j++) {
1.203 brouard 3878: printf("%.6e ",hess[i][j]);
3879: fprintf(ficlog,"%.6e ",hess[i][j]);
1.126 brouard 3880: }
3881: printf("\n");
3882: fprintf(ficlog,"\n");
3883: }
3884:
1.203 brouard 3885: /* printf("\n#Covariance matrix#\n"); */
3886: /* fprintf(ficlog,"\n#Covariance matrix#\n"); */
3887: /* for (i=1;i<=npar;i++) { */
3888: /* for (j=1;j<=npar;j++) { */
3889: /* printf("%.6e ",matcov[i][j]); */
3890: /* fprintf(ficlog,"%.6e ",matcov[i][j]); */
3891: /* } */
3892: /* printf("\n"); */
3893: /* fprintf(ficlog,"\n"); */
3894: /* } */
3895:
1.126 brouard 3896: /* Recompute Inverse */
1.203 brouard 3897: /* for (i=1;i<=npar;i++) */
3898: /* for (j=1;j<=npar;j++) a[i][j]=matcov[i][j]; */
3899: /* ludcmp(a,npar,indx,&pd); */
3900:
3901: /* printf("\n#Hessian matrix recomputed#\n"); */
3902:
3903: /* for (j=1;j<=npar;j++) { */
3904: /* for (i=1;i<=npar;i++) x[i]=0; */
3905: /* x[j]=1; */
3906: /* lubksb(a,npar,indx,x); */
3907: /* for (i=1;i<=npar;i++){ */
3908: /* y[i][j]=x[i]; */
3909: /* printf("%.3e ",y[i][j]); */
3910: /* fprintf(ficlog,"%.3e ",y[i][j]); */
3911: /* } */
3912: /* printf("\n"); */
3913: /* fprintf(ficlog,"\n"); */
3914: /* } */
3915:
3916: /* Verifying the inverse matrix */
3917: #ifdef DEBUGHESS
3918: y=matprod2(y,hess,1,npar,1,npar,1,npar,matcov);
1.126 brouard 3919:
1.203 brouard 3920: printf("\n#Verification: multiplying the matrix of covariance by the Hessian matrix, should be unity:#\n");
3921: fprintf(ficlog,"\n#Verification: multiplying the matrix of covariance by the Hessian matrix. Should be unity:#\n");
1.126 brouard 3922:
3923: for (j=1;j<=npar;j++) {
3924: for (i=1;i<=npar;i++){
1.203 brouard 3925: printf("%.2f ",y[i][j]);
3926: fprintf(ficlog,"%.2f ",y[i][j]);
1.126 brouard 3927: }
3928: printf("\n");
3929: fprintf(ficlog,"\n");
3930: }
1.203 brouard 3931: #endif
1.126 brouard 3932:
3933: free_matrix(a,1,npar,1,npar);
3934: free_matrix(y,1,npar,1,npar);
3935: free_vector(x,1,npar);
3936: free_ivector(indx,1,npar);
1.203 brouard 3937: /* free_matrix(hess,1,npar,1,npar); */
1.126 brouard 3938:
3939:
3940: }
3941:
3942: /*************** hessian matrix ****************/
3943: double hessii(double x[], double delta, int theta, double delti[], double (*func)(double []), int npar)
1.203 brouard 3944: { /* Around values of x, computes the function func and returns the scales delti and hessian */
1.126 brouard 3945: int i;
3946: int l=1, lmax=20;
1.203 brouard 3947: double k1,k2, res, fx;
1.132 brouard 3948: double p2[MAXPARM+1]; /* identical to x */
1.126 brouard 3949: double delt=0.0001, delts, nkhi=10.,nkhif=1., khi=1.e-4;
3950: int k=0,kmax=10;
3951: double l1;
3952:
3953: fx=func(x);
3954: for (i=1;i<=npar;i++) p2[i]=x[i];
1.145 brouard 3955: for(l=0 ; l <=lmax; l++){ /* Enlarging the zone around the Maximum */
1.126 brouard 3956: l1=pow(10,l);
3957: delts=delt;
3958: for(k=1 ; k <kmax; k=k+1){
3959: delt = delta*(l1*k);
3960: p2[theta]=x[theta] +delt;
1.145 brouard 3961: k1=func(p2)-fx; /* Might be negative if too close to the theoretical maximum */
1.126 brouard 3962: p2[theta]=x[theta]-delt;
3963: k2=func(p2)-fx;
3964: /*res= (k1-2.0*fx+k2)/delt/delt; */
1.203 brouard 3965: res= (k1+k2)/delt/delt/2.; /* Divided by 2 because L and not 2*L */
1.126 brouard 3966:
1.203 brouard 3967: #ifdef DEBUGHESSII
1.126 brouard 3968: 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);
3969: 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);
3970: #endif
3971: /*if(fabs(k1-2.0*fx+k2) <1.e-13){ */
3972: if((k1 <khi/nkhi/2.) || (k2 <khi/nkhi/2.)){
3973: k=kmax;
3974: }
3975: else if((k1 >khi/nkhif) || (k2 >khi/nkhif)){ /* Keeps lastvalue before 3.84/2 KHI2 5% 1d.f. */
1.164 brouard 3976: k=kmax; l=lmax*10;
1.126 brouard 3977: }
3978: else if((k1 >khi/nkhi) || (k2 >khi/nkhi)){
3979: delts=delt;
3980: }
1.203 brouard 3981: } /* End loop k */
1.126 brouard 3982: }
3983: delti[theta]=delts;
3984: return res;
3985:
3986: }
3987:
1.203 brouard 3988: double hessij( double x[], double **hess, double delti[], int thetai,int thetaj,double (*func)(double []),int npar)
1.126 brouard 3989: {
3990: int i;
1.164 brouard 3991: int l=1, lmax=20;
1.126 brouard 3992: double k1,k2,k3,k4,res,fx;
1.132 brouard 3993: double p2[MAXPARM+1];
1.203 brouard 3994: int k, kmax=1;
3995: double v1, v2, cv12, lc1, lc2;
1.208 brouard 3996:
3997: int firstime=0;
1.203 brouard 3998:
1.126 brouard 3999: fx=func(x);
1.203 brouard 4000: for (k=1; k<=kmax; k=k+10) {
1.126 brouard 4001: for (i=1;i<=npar;i++) p2[i]=x[i];
1.203 brouard 4002: p2[thetai]=x[thetai]+delti[thetai]*k;
4003: p2[thetaj]=x[thetaj]+delti[thetaj]*k;
1.126 brouard 4004: k1=func(p2)-fx;
4005:
1.203 brouard 4006: p2[thetai]=x[thetai]+delti[thetai]*k;
4007: p2[thetaj]=x[thetaj]-delti[thetaj]*k;
1.126 brouard 4008: k2=func(p2)-fx;
4009:
1.203 brouard 4010: p2[thetai]=x[thetai]-delti[thetai]*k;
4011: p2[thetaj]=x[thetaj]+delti[thetaj]*k;
1.126 brouard 4012: k3=func(p2)-fx;
4013:
1.203 brouard 4014: p2[thetai]=x[thetai]-delti[thetai]*k;
4015: p2[thetaj]=x[thetaj]-delti[thetaj]*k;
1.126 brouard 4016: k4=func(p2)-fx;
1.203 brouard 4017: res=(k1-k2-k3+k4)/4.0/delti[thetai]/k/delti[thetaj]/k/2.; /* Because of L not 2*L */
4018: if(k1*k2*k3*k4 <0.){
1.208 brouard 4019: firstime=1;
1.203 brouard 4020: kmax=kmax+10;
1.208 brouard 4021: }
4022: if(kmax >=10 || firstime ==1){
1.246 brouard 4023: 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);
4024: 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 4025: 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);
4026: 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);
4027: }
4028: #ifdef DEBUGHESSIJ
4029: v1=hess[thetai][thetai];
4030: v2=hess[thetaj][thetaj];
4031: cv12=res;
4032: /* Computing eigen value of Hessian matrix */
4033: lc1=((v1+v2)+sqrt((v1+v2)*(v1+v2) - 4*(v1*v2-cv12*cv12)))/2.;
4034: lc2=((v1+v2)-sqrt((v1+v2)*(v1+v2) - 4*(v1*v2-cv12*cv12)))/2.;
4035: if ((lc2 <0) || (lc1 <0) ){
4036: printf("Warning: sub Hessian matrix '%d%d' does not have positive eigen values \n",thetai,thetaj);
4037: fprintf(ficlog, "Warning: sub Hessian matrix '%d%d' does not have positive eigen values \n",thetai,thetaj);
4038: 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);
4039: 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);
4040: }
1.126 brouard 4041: #endif
4042: }
4043: return res;
4044: }
4045:
1.203 brouard 4046: /* Not done yet: Was supposed to fix if not exactly at the maximum */
4047: /* double hessij( double x[], double delti[], int thetai,int thetaj,double (*func)(double []),int npar) */
4048: /* { */
4049: /* int i; */
4050: /* int l=1, lmax=20; */
4051: /* double k1,k2,k3,k4,res,fx; */
4052: /* double p2[MAXPARM+1]; */
4053: /* double delt=0.0001, delts, nkhi=10.,nkhif=1., khi=1.e-4; */
4054: /* int k=0,kmax=10; */
4055: /* double l1; */
4056:
4057: /* fx=func(x); */
4058: /* for(l=0 ; l <=lmax; l++){ /\* Enlarging the zone around the Maximum *\/ */
4059: /* l1=pow(10,l); */
4060: /* delts=delt; */
4061: /* for(k=1 ; k <kmax; k=k+1){ */
4062: /* delt = delti*(l1*k); */
4063: /* for (i=1;i<=npar;i++) p2[i]=x[i]; */
4064: /* p2[thetai]=x[thetai]+delti[thetai]/k; */
4065: /* p2[thetaj]=x[thetaj]+delti[thetaj]/k; */
4066: /* k1=func(p2)-fx; */
4067:
4068: /* p2[thetai]=x[thetai]+delti[thetai]/k; */
4069: /* p2[thetaj]=x[thetaj]-delti[thetaj]/k; */
4070: /* k2=func(p2)-fx; */
4071:
4072: /* p2[thetai]=x[thetai]-delti[thetai]/k; */
4073: /* p2[thetaj]=x[thetaj]+delti[thetaj]/k; */
4074: /* k3=func(p2)-fx; */
4075:
4076: /* p2[thetai]=x[thetai]-delti[thetai]/k; */
4077: /* p2[thetaj]=x[thetaj]-delti[thetaj]/k; */
4078: /* k4=func(p2)-fx; */
4079: /* res=(k1-k2-k3+k4)/4.0/delti[thetai]*k/delti[thetaj]*k/2.; /\* Because of L not 2*L *\/ */
4080: /* #ifdef DEBUGHESSIJ */
4081: /* 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); */
4082: /* 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); */
4083: /* #endif */
4084: /* if((k1 <khi/nkhi/2.) || (k2 <khi/nkhi/2.)|| (k4 <khi/nkhi/2.)|| (k4 <khi/nkhi/2.)){ */
4085: /* k=kmax; */
4086: /* } */
4087: /* else if((k1 >khi/nkhif) || (k2 >khi/nkhif) || (k4 >khi/nkhif) || (k4 >khi/nkhif)){ /\* Keeps lastvalue before 3.84/2 KHI2 5% 1d.f. *\/ */
4088: /* k=kmax; l=lmax*10; */
4089: /* } */
4090: /* else if((k1 >khi/nkhi) || (k2 >khi/nkhi)){ */
4091: /* delts=delt; */
4092: /* } */
4093: /* } /\* End loop k *\/ */
4094: /* } */
4095: /* delti[theta]=delts; */
4096: /* return res; */
4097: /* } */
4098:
4099:
1.126 brouard 4100: /************** Inverse of matrix **************/
4101: void ludcmp(double **a, int n, int *indx, double *d)
4102: {
4103: int i,imax,j,k;
4104: double big,dum,sum,temp;
4105: double *vv;
4106:
4107: vv=vector(1,n);
4108: *d=1.0;
4109: for (i=1;i<=n;i++) {
4110: big=0.0;
4111: for (j=1;j<=n;j++)
4112: if ((temp=fabs(a[i][j])) > big) big=temp;
4113: if (big == 0.0) nrerror("Singular matrix in routine ludcmp");
4114: vv[i]=1.0/big;
4115: }
4116: for (j=1;j<=n;j++) {
4117: for (i=1;i<j;i++) {
4118: sum=a[i][j];
4119: for (k=1;k<i;k++) sum -= a[i][k]*a[k][j];
4120: a[i][j]=sum;
4121: }
4122: big=0.0;
4123: for (i=j;i<=n;i++) {
4124: sum=a[i][j];
4125: for (k=1;k<j;k++)
4126: sum -= a[i][k]*a[k][j];
4127: a[i][j]=sum;
4128: if ( (dum=vv[i]*fabs(sum)) >= big) {
4129: big=dum;
4130: imax=i;
4131: }
4132: }
4133: if (j != imax) {
4134: for (k=1;k<=n;k++) {
4135: dum=a[imax][k];
4136: a[imax][k]=a[j][k];
4137: a[j][k]=dum;
4138: }
4139: *d = -(*d);
4140: vv[imax]=vv[j];
4141: }
4142: indx[j]=imax;
4143: if (a[j][j] == 0.0) a[j][j]=TINY;
4144: if (j != n) {
4145: dum=1.0/(a[j][j]);
4146: for (i=j+1;i<=n;i++) a[i][j] *= dum;
4147: }
4148: }
4149: free_vector(vv,1,n); /* Doesn't work */
4150: ;
4151: }
4152:
4153: void lubksb(double **a, int n, int *indx, double b[])
4154: {
4155: int i,ii=0,ip,j;
4156: double sum;
4157:
4158: for (i=1;i<=n;i++) {
4159: ip=indx[i];
4160: sum=b[ip];
4161: b[ip]=b[i];
4162: if (ii)
4163: for (j=ii;j<=i-1;j++) sum -= a[i][j]*b[j];
4164: else if (sum) ii=i;
4165: b[i]=sum;
4166: }
4167: for (i=n;i>=1;i--) {
4168: sum=b[i];
4169: for (j=i+1;j<=n;j++) sum -= a[i][j]*b[j];
4170: b[i]=sum/a[i][i];
4171: }
4172: }
4173:
4174: void pstamp(FILE *fichier)
4175: {
1.196 brouard 4176: fprintf(fichier,"# %s.%s\n#IMaCh version %s, %s\n#%s\n# %s", optionfilefiname,optionfilext,version,copyright, fullversion, strstart);
1.126 brouard 4177: }
4178:
4179: /************ Frequencies ********************/
1.226 brouard 4180: void freqsummary(char fileres[], int iagemin, int iagemax, int **s, double **agev, int nlstate, int imx, \
4181: int *Tvaraff, int *invalidvarcomb, int **nbcode, int *ncodemax,double **mint,double **anint, char strstart[], \
4182: int firstpass, int lastpass, int stepm, int weightopt, char model[])
4183: { /* Some frequencies */
4184:
1.227 brouard 4185: int i, m, jk, j1, bool, z1,j, k, iv;
1.226 brouard 4186: int iind=0, iage=0;
4187: int mi; /* Effective wave */
4188: int first;
4189: double ***freq; /* Frequencies */
4190: double *meanq;
4191: double **meanqt;
4192: double *pp, **prop, *posprop, *pospropt;
4193: double pos=0., posproptt=0., pospropta=0., k2, dateintsum=0,k2cpt=0;
4194: char fileresp[FILENAMELENGTH], fileresphtm[FILENAMELENGTH], fileresphtmfr[FILENAMELENGTH];
4195: double agebegin, ageend;
4196:
4197: pp=vector(1,nlstate);
4198: prop=matrix(1,nlstate,iagemin-AGEMARGE,iagemax+3+AGEMARGE);
4199: posprop=vector(1,nlstate); /* Counting the number of transition starting from a live state per age */
4200: pospropt=vector(1,nlstate); /* Counting the number of transition starting from a live state */
4201: /* prop=matrix(1,nlstate,iagemin,iagemax+3); */
4202: meanq=vector(1,nqfveff); /* Number of Quantitative Fixed Variables Effective */
4203: meanqt=matrix(1,lastpass,1,nqtveff);
4204: strcpy(fileresp,"P_");
4205: strcat(fileresp,fileresu);
4206: /*strcat(fileresphtm,fileresu);*/
4207: if((ficresp=fopen(fileresp,"w"))==NULL) {
4208: printf("Problem with prevalence resultfile: %s\n", fileresp);
4209: fprintf(ficlog,"Problem with prevalence resultfile: %s\n", fileresp);
4210: exit(0);
4211: }
1.240 brouard 4212:
1.226 brouard 4213: strcpy(fileresphtm,subdirfext(optionfilefiname,"PHTM_",".htm"));
4214: if((ficresphtm=fopen(fileresphtm,"w"))==NULL) {
4215: printf("Problem with prevalence HTM resultfile '%s' with errno='%s'\n",fileresphtm,strerror(errno));
4216: fprintf(ficlog,"Problem with prevalence HTM resultfile '%s' with errno='%s'\n",fileresphtm,strerror(errno));
4217: fflush(ficlog);
4218: exit(70);
4219: }
4220: else{
4221: fprintf(ficresphtm,"<html><head>\n<title>IMaCh PHTM_ %s</title></head>\n <body><font size=\"2\">%s <br> %s</font> \
1.240 brouard 4222: <hr size=\"2\" color=\"#EC5E5E\"> \n \
1.214 brouard 4223: Title=%s <br>Datafile=%s Firstpass=%d Lastpass=%d Stepm=%d Weight=%d Model=1+age+%s<br>\n",\
1.226 brouard 4224: fileresphtm,version,fullversion,title,datafile,firstpass,lastpass,stepm, weightopt, model);
4225: }
1.237 brouard 4226: 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 4227:
1.226 brouard 4228: strcpy(fileresphtmfr,subdirfext(optionfilefiname,"PHTMFR_",".htm"));
4229: if((ficresphtmfr=fopen(fileresphtmfr,"w"))==NULL) {
4230: printf("Problem with frequency table HTM resultfile '%s' with errno='%s'\n",fileresphtmfr,strerror(errno));
4231: fprintf(ficlog,"Problem with frequency table HTM resultfile '%s' with errno='%s'\n",fileresphtmfr,strerror(errno));
4232: fflush(ficlog);
4233: exit(70);
1.240 brouard 4234: } else{
1.226 brouard 4235: 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 4236: <hr size=\"2\" color=\"#EC5E5E\"> \n \
1.214 brouard 4237: Title=%s <br>Datafile=%s Firstpass=%d Lastpass=%d Stepm=%d Weight=%d Model=1+age+%s<br>\n",\
1.226 brouard 4238: fileresphtmfr,version,fullversion,title,datafile,firstpass,lastpass,stepm, weightopt, model);
4239: }
1.240 brouard 4240: 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);
4241:
1.226 brouard 4242: freq= ma3x(-5,nlstate+ndeath,-5,nlstate+ndeath,iagemin-AGEMARGE,iagemax+3+AGEMARGE);
4243: j1=0;
1.126 brouard 4244:
1.227 brouard 4245: /* j=ncoveff; /\* Only fixed dummy covariates *\/ */
4246: j=cptcoveff; /* Only dummy covariates of the model */
1.226 brouard 4247: if (cptcovn<1) {j=1;ncodemax[1]=1;}
1.240 brouard 4248:
1.226 brouard 4249: first=1;
1.240 brouard 4250:
1.226 brouard 4251: /* Detects if a combination j1 is empty: for a multinomial variable like 3 education levels:
4252: reference=low_education V1=0,V2=0
4253: med_educ V1=1 V2=0,
4254: high_educ V1=0 V2=1
4255: Then V1=1 and V2=1 is a noisy combination that we want to exclude for the list 2**cptcoveff
4256: */
1.240 brouard 4257:
1.227 brouard 4258: for (j1 = 1; j1 <= (int) pow(2,j); j1++){ /* Loop on covariates combination in order of model, excluding quantitatives V4=0, V3=0 for example, fixed or varying covariates */
1.226 brouard 4259: posproptt=0.;
4260: /*printf("cptcoveff=%d Tvaraff=%d", cptcoveff,Tvaraff[1]);
4261: scanf("%d", i);*/
4262: for (i=-5; i<=nlstate+ndeath; i++)
4263: for (jk=-5; jk<=nlstate+ndeath; jk++)
1.240 brouard 4264: for(m=iagemin; m <= iagemax+3; m++)
4265: freq[i][jk][m]=0;
4266:
1.226 brouard 4267: for (i=1; i<=nlstate; i++) {
4268: for(m=iagemin; m <= iagemax+3; m++)
1.240 brouard 4269: prop[i][m]=0;
1.226 brouard 4270: posprop[i]=0;
4271: pospropt[i]=0;
4272: }
1.227 brouard 4273: /* for (z1=1; z1<= nqfveff; z1++) { */
4274: /* meanq[z1]+=0.; */
4275: /* for(m=1;m<=lastpass;m++){ */
4276: /* meanqt[m][z1]=0.; */
4277: /* } */
4278: /* } */
1.240 brouard 4279:
1.226 brouard 4280: dateintsum=0;
4281: k2cpt=0;
1.227 brouard 4282: /* For that combination of covariate j1, we count and print the frequencies in one pass */
1.226 brouard 4283: for (iind=1; iind<=imx; iind++) { /* For each individual iind */
4284: bool=1;
1.227 brouard 4285: if(anyvaryingduminmodel==0){ /* If All fixed covariates */
1.234 brouard 4286: if (cptcoveff >0) { /* Filter is here: Must be looked at for model=V1+V2+V3+V4 */
1.227 brouard 4287: /* for (z1=1; z1<= nqfveff; z1++) { */
4288: /* meanq[z1]+=coqvar[Tvar[z1]][iind]; /\* Computes mean of quantitative with selected filter *\/ */
4289: /* } */
1.234 brouard 4290: for (z1=1; z1<=cptcoveff; z1++) {
4291: /* if(Tvaraff[z1] ==-20){ */
4292: /* /\* sumnew+=cotvar[mw[mi][iind]][z1][iind]; *\/ */
4293: /* }else if(Tvaraff[z1] ==-10){ */
4294: /* /\* sumnew+=coqvar[z1][iind]; *\/ */
4295: /* }else */
4296: if (covar[Tvaraff[z1]][iind]!= nbcode[Tvaraff[z1]][codtabm(j1,z1)]){
4297: /* Tests if this individual iind responded to j1 (V4=1 V3=0) */
4298: bool=0;
4299: /* 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",
4300: bool,i,z1, z1, Tvaraff[z1],i,covar[Tvaraff[z1]][i],j1,z1,codtabm(j1,z1),
4301: j1,z1,nbcode[Tvaraff[z1]][codtabm(j1,z1)],j1);*/
4302: /* For j1=7 in V1+V2+V3+V4 = 0 1 1 0 and codtabm(7,3)=1 and nbcde[3][?]=1*/
4303: } /* Onlyf fixed */
4304: } /* end z1 */
4305: } /* cptcovn > 0 */
1.227 brouard 4306: } /* end any */
4307: if (bool==1){ /* We selected an individual iind satisfying combination j1 or all fixed */
1.234 brouard 4308: /* for(m=firstpass; m<=lastpass; m++){ */
4309: for(mi=1; mi<wav[iind];mi++){ /* For that wave */
4310: m=mw[mi][iind];
4311: if(anyvaryingduminmodel==1){ /* Some are varying covariates */
4312: for (z1=1; z1<=cptcoveff; z1++) {
4313: if( Fixed[Tmodelind[z1]]==1){
4314: iv= Tvar[Tmodelind[z1]]-ncovcol-nqv;
4315: if (cotvar[m][iv][iind]!= nbcode[Tvaraff[z1]][codtabm(j1,z1)]) /* iv=1 to ntv, right modality */
4316: bool=0;
4317: }else if( Fixed[Tmodelind[z1]]== 0) { /* fixed */
4318: if (covar[Tvaraff[z1]][iind]!= nbcode[Tvaraff[z1]][codtabm(j1,z1)]) {
4319: bool=0;
4320: }
4321: }
4322: }
4323: }/* Some are varying covariates, we tried to speed up if all fixed covariates in the model, avoiding waves loop */
4324: /* bool =0 we keep that guy which corresponds to the combination of dummy values */
4325: if(bool==1){
4326: /* dh[m][iind] or dh[mw[mi][iind]][iind] is the delay between two effective (mi) waves m=mw[mi][iind]
4327: and mw[mi+1][iind]. dh depends on stepm. */
4328: agebegin=agev[m][iind]; /* Age at beginning of wave before transition*/
4329: ageend=agev[m][iind]+(dh[m][iind])*stepm/YEARM; /* Age at end of wave and transition */
4330: if(m >=firstpass && m <=lastpass){
4331: k2=anint[m][iind]+(mint[m][iind]/12.);
4332: /*if ((k2>=dateprev1) && (k2<=dateprev2)) {*/
4333: if(agev[m][iind]==0) agev[m][iind]=iagemax+1; /* All ages equal to 0 are in iagemax+1 */
4334: if(agev[m][iind]==1) agev[m][iind]=iagemax+2; /* All ages equal to 1 are in iagemax+2 */
4335: if (s[m][iind]>0 && s[m][iind]<=nlstate) /* If status at wave m is known and a live state */
4336: prop[s[m][iind]][(int)agev[m][iind]] += weight[iind]; /* At age of beginning of transition, where status is known */
4337: if (m<lastpass) {
4338: /* if(s[m][iind]==4 && s[m+1][iind]==4) */
4339: /* 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]); */
4340: if(s[m][iind]==-1)
4341: 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.));
4342: freq[s[m][iind]][s[m+1][iind]][(int)agev[m][iind]] += weight[iind]; /* At age of beginning of transition, where status is known */
4343: /* freq[s[m][iind]][s[m+1][iind]][(int)((agebegin+ageend)/2.)] += weight[iind]; */
4344: 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 */
4345: }
4346: } /* end if between passes */
4347: if ((agev[m][iind]>1) && (agev[m][iind]< (iagemax+3)) && (anint[m][iind]!=9999) && (mint[m][iind]!=99)) {
4348: dateintsum=dateintsum+k2;
4349: k2cpt++;
4350: /* printf("iind=%ld dateintmean = %lf dateintsum=%lf k2cpt=%lf k2=%lf\n",iind, dateintsum/k2cpt, dateintsum,k2cpt, k2); */
4351: }
4352: } /* end bool 2 */
4353: } /* end m */
1.226 brouard 4354: } /* end bool */
4355: } /* end iind = 1 to imx */
4356: /* prop[s][age] is feeded for any initial and valid live state as well as
4357: freq[s1][s2][age] at single age of beginning the transition, for a combination j1 */
1.240 brouard 4358:
4359:
1.226 brouard 4360: /* fprintf(ficresp, "#Count between %.lf/%.lf/%.lf and %.lf/%.lf/%.lf\n",jprev1, mprev1,anprev1,jprev2, mprev2,anprev2);*/
4361: pstamp(ficresp);
1.240 brouard 4362: if (cptcoveff>0){
1.226 brouard 4363: fprintf(ficresp, "\n#********** Variable ");
4364: fprintf(ficresphtm, "\n<br/><br/><h3>********** Variable ");
4365: fprintf(ficresphtmfr, "\n<br/><br/><h3>********** Variable ");
1.240 brouard 4366: fprintf(ficlog, "\n#********** Variable ");
1.227 brouard 4367: for (z1=1; z1<=cptcoveff; z1++){
1.240 brouard 4368: if(DummyV[z1]){
4369: fprintf(ficresp, "V%d (fixed)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4370: fprintf(ficresphtm, "V%d (fixed)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4371: fprintf(ficresphtmfr, "V%d (fixed)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4372: fprintf(ficlog, "V%d (fixed)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4373: }else{
4374: fprintf(ficresp, "V%d(varying)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4375: fprintf(ficresphtm, "V%d(varying)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4376: fprintf(ficresphtmfr, "V%d(varying)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4377: fprintf(ficlog, "V%d(varying)=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
4378: }
1.226 brouard 4379: }
4380: fprintf(ficresp, "**********\n#");
4381: fprintf(ficresphtm, "**********</h3>\n");
4382: fprintf(ficresphtmfr, "**********</h3>\n");
4383: fprintf(ficlog, "**********\n");
4384: }
4385: fprintf(ficresphtm,"<table style=\"text-align:center; border: 1px solid\">");
4386: for(i=1; i<=nlstate;i++) {
1.240 brouard 4387: fprintf(ficresp, " Age Prev(%d) N(%d) N ",i,i);
1.226 brouard 4388: fprintf(ficresphtm, "<th>Age</th><th>Prev(%d)</th><th>N(%d)</th><th>N</th>",i,i);
4389: }
4390: fprintf(ficresp, "\n");
4391: fprintf(ficresphtm, "\n");
1.240 brouard 4392:
1.226 brouard 4393: /* Header of frequency table by age */
4394: fprintf(ficresphtmfr,"<table style=\"text-align:center; border: 1px solid\">");
4395: fprintf(ficresphtmfr,"<th>Age</th> ");
4396: for(jk=-1; jk <=nlstate+ndeath; jk++){
4397: for(m=-1; m <=nlstate+ndeath; m++){
1.234 brouard 4398: if(jk!=0 && m!=0)
4399: fprintf(ficresphtmfr,"<th>%d%d</th> ",jk,m);
1.226 brouard 4400: }
4401: }
4402: fprintf(ficresphtmfr, "\n");
1.240 brouard 4403:
1.226 brouard 4404: /* For each age */
4405: for(iage=iagemin; iage <= iagemax+3; iage++){
4406: fprintf(ficresphtm,"<tr>");
4407: if(iage==iagemax+1){
1.240 brouard 4408: fprintf(ficlog,"1");
4409: fprintf(ficresphtmfr,"<tr><th>0</th> ");
1.226 brouard 4410: }else if(iage==iagemax+2){
1.240 brouard 4411: fprintf(ficlog,"0");
4412: fprintf(ficresphtmfr,"<tr><th>Unknown</th> ");
1.226 brouard 4413: }else if(iage==iagemax+3){
1.240 brouard 4414: fprintf(ficlog,"Total");
4415: fprintf(ficresphtmfr,"<tr><th>Total</th> ");
1.226 brouard 4416: }else{
1.240 brouard 4417: if(first==1){
4418: first=0;
4419: printf("See log file for details...\n");
4420: }
4421: fprintf(ficresphtmfr,"<tr><th>%d</th> ",iage);
4422: fprintf(ficlog,"Age %d", iage);
1.226 brouard 4423: }
4424: for(jk=1; jk <=nlstate ; jk++){
1.240 brouard 4425: for(m=-1, pp[jk]=0; m <=nlstate+ndeath ; m++)
4426: pp[jk] += freq[jk][m][iage];
1.226 brouard 4427: }
4428: for(jk=1; jk <=nlstate ; jk++){
1.240 brouard 4429: for(m=-1, pos=0; m <=0 ; m++)
4430: pos += freq[jk][m][iage];
4431: if(pp[jk]>=1.e-10){
4432: if(first==1){
4433: printf(" %d.=%.0f loss[%d]=%.1f%%",jk,pp[jk],jk,100*pos/pp[jk]);
4434: }
4435: fprintf(ficlog," %d.=%.0f loss[%d]=%.1f%%",jk,pp[jk],jk,100*pos/pp[jk]);
4436: }else{
4437: if(first==1)
4438: printf(" %d.=%.0f loss[%d]=NaNQ%%",jk,pp[jk],jk);
4439: fprintf(ficlog," %d.=%.0f loss[%d]=NaNQ%%",jk,pp[jk],jk);
4440: }
1.226 brouard 4441: }
1.240 brouard 4442:
1.226 brouard 4443: for(jk=1; jk <=nlstate ; jk++){
1.240 brouard 4444: /* posprop[jk]=0; */
4445: for(m=0, pp[jk]=0; m <=nlstate+ndeath; m++)/* Summing on all ages */
4446: pp[jk] += freq[jk][m][iage];
1.226 brouard 4447: } /* pp[jk] is the total number of transitions starting from state jk and any ending status until this age */
1.240 brouard 4448:
1.226 brouard 4449: for(jk=1,pos=0, pospropta=0.; jk <=nlstate ; jk++){
1.240 brouard 4450: pos += pp[jk]; /* pos is the total number of transitions until this age */
4451: posprop[jk] += prop[jk][iage]; /* prop is the number of transitions from a live state
4452: from jk at age iage prop[s[m][iind]][(int)agev[m][iind]] += weight[iind] */
4453: pospropta += prop[jk][iage]; /* prop is the number of transitions from a live state
4454: from jk at age iage prop[s[m][iind]][(int)agev[m][iind]] += weight[iind] */
1.226 brouard 4455: }
4456: for(jk=1; jk <=nlstate ; jk++){
1.240 brouard 4457: if(pos>=1.e-5){
4458: if(first==1)
4459: printf(" %d.=%.0f prev[%d]=%.1f%%",jk,pp[jk],jk,100*pp[jk]/pos);
4460: fprintf(ficlog," %d.=%.0f prev[%d]=%.1f%%",jk,pp[jk],jk,100*pp[jk]/pos);
4461: }else{
4462: if(first==1)
4463: printf(" %d.=%.0f prev[%d]=NaNQ%%",jk,pp[jk],jk);
4464: fprintf(ficlog," %d.=%.0f prev[%d]=NaNQ%%",jk,pp[jk],jk);
4465: }
4466: if( iage <= iagemax){
4467: if(pos>=1.e-5){
4468: fprintf(ficresp," %d %.5f %.0f %.0f",iage,prop[jk][iage]/pospropta, prop[jk][iage],pospropta);
4469: fprintf(ficresphtm,"<th>%d</th><td>%.5f</td><td>%.0f</td><td>%.0f</td>",iage,prop[jk][iage]/pospropta, prop[jk][iage],pospropta);
4470: /*probs[iage][jk][j1]= pp[jk]/pos;*/
4471: /*printf("\niage=%d jk=%d j1=%d %.5f %.0f %.0f %f",iage,jk,j1,pp[jk]/pos, pp[jk],pos,probs[iage][jk][j1]);*/
4472: }
4473: else{
4474: fprintf(ficresp," %d NaNq %.0f %.0f",iage,prop[jk][iage],pospropta);
4475: fprintf(ficresphtm,"<th>%d</th><td>NaNq</td><td>%.0f</td><td>%.0f</td>",iage, prop[jk][iage],pospropta);
4476: }
4477: }
4478: pospropt[jk] +=posprop[jk];
1.226 brouard 4479: } /* end loop jk */
4480: /* pospropt=0.; */
4481: for(jk=-1; jk <=nlstate+ndeath; jk++){
1.240 brouard 4482: for(m=-1; m <=nlstate+ndeath; m++){
4483: if(freq[jk][m][iage] !=0 ) { /* minimizing output */
4484: if(first==1){
4485: printf(" %d%d=%.0f",jk,m,freq[jk][m][iage]);
4486: }
4487: fprintf(ficlog," %d%d=%.0f",jk,m,freq[jk][m][iage]);
4488: }
4489: if(jk!=0 && m!=0)
4490: fprintf(ficresphtmfr,"<td>%.0f</td> ",freq[jk][m][iage]);
4491: }
1.226 brouard 4492: } /* end loop jk */
4493: posproptt=0.;
4494: for(jk=1; jk <=nlstate; jk++){
1.240 brouard 4495: posproptt += pospropt[jk];
1.226 brouard 4496: }
4497: fprintf(ficresphtmfr,"</tr>\n ");
4498: if(iage <= iagemax){
1.240 brouard 4499: fprintf(ficresp,"\n");
4500: fprintf(ficresphtm,"</tr>\n");
1.226 brouard 4501: }
4502: if(first==1)
1.240 brouard 4503: printf("Others in log...\n");
1.226 brouard 4504: fprintf(ficlog,"\n");
4505: } /* end loop age iage */
4506: fprintf(ficresphtm,"<tr><th>Tot</th>");
4507: for(jk=1; jk <=nlstate ; jk++){
4508: if(posproptt < 1.e-5){
1.240 brouard 4509: fprintf(ficresphtm,"<td>Nanq</td><td>%.0f</td><td>%.0f</td>",pospropt[jk],posproptt);
1.226 brouard 4510: }else{
1.240 brouard 4511: fprintf(ficresphtm,"<td>%.5f</td><td>%.0f</td><td>%.0f</td>",pospropt[jk]/posproptt,pospropt[jk],posproptt);
1.226 brouard 4512: }
4513: }
4514: fprintf(ficresphtm,"</tr>\n");
4515: fprintf(ficresphtm,"</table>\n");
4516: fprintf(ficresphtmfr,"</table>\n");
4517: if(posproptt < 1.e-5){
4518: fprintf(ficresphtm,"\n <p><b> This combination (%d) is not valid and no result will be produced</b></p>",j1);
4519: fprintf(ficresphtmfr,"\n <p><b> This combination (%d) is not valid and no result will be produced</b></p>",j1);
4520: fprintf(ficres,"\n This combination (%d) is not valid and no result will be produced\n\n",j1);
4521: invalidvarcomb[j1]=1;
4522: }else{
4523: fprintf(ficresphtm,"\n <p> This combination (%d) is valid and result will be produced.</p>",j1);
4524: invalidvarcomb[j1]=0;
4525: }
4526: fprintf(ficresphtmfr,"</table>\n");
4527: } /* end selected combination of covariate j1 */
4528: dateintmean=dateintsum/k2cpt;
1.240 brouard 4529:
1.226 brouard 4530: fclose(ficresp);
4531: fclose(ficresphtm);
4532: fclose(ficresphtmfr);
4533: free_vector(meanq,1,nqfveff);
4534: free_matrix(meanqt,1,lastpass,1,nqtveff);
4535: free_ma3x(freq,-5,nlstate+ndeath,-5,nlstate+ndeath, iagemin-AGEMARGE, iagemax+3+AGEMARGE);
4536: free_vector(pospropt,1,nlstate);
4537: free_vector(posprop,1,nlstate);
4538: free_matrix(prop,1,nlstate,iagemin-AGEMARGE, iagemax+3+AGEMARGE);
4539: free_vector(pp,1,nlstate);
4540: /* End of freqsummary */
4541: }
1.126 brouard 4542:
4543: /************ Prevalence ********************/
1.227 brouard 4544: 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)
4545: {
4546: /* Compute observed prevalence between dateprev1 and dateprev2 by counting the number of people
4547: in each health status at the date of interview (if between dateprev1 and dateprev2).
4548: We still use firstpass and lastpass as another selection.
4549: */
1.126 brouard 4550:
1.227 brouard 4551: int i, m, jk, j1, bool, z1,j, iv;
4552: int mi; /* Effective wave */
4553: int iage;
4554: double agebegin, ageend;
4555:
4556: double **prop;
4557: double posprop;
4558: double y2; /* in fractional years */
4559: int iagemin, iagemax;
4560: int first; /** to stop verbosity which is redirected to log file */
4561:
4562: iagemin= (int) agemin;
4563: iagemax= (int) agemax;
4564: /*pp=vector(1,nlstate);*/
4565: prop=matrix(1,nlstate,iagemin-AGEMARGE,iagemax+3+AGEMARGE);
4566: /* freq=ma3x(-1,nlstate+ndeath,-1,nlstate+ndeath,iagemin,iagemax+3);*/
4567: j1=0;
1.222 brouard 4568:
1.227 brouard 4569: /*j=cptcoveff;*/
4570: if (cptcovn<1) {j=1;ncodemax[1]=1;}
1.222 brouard 4571:
1.227 brouard 4572: first=1;
4573: for(j1=1; j1<= (int) pow(2,cptcoveff);j1++){ /* For each combination of covariate */
4574: for (i=1; i<=nlstate; i++)
4575: for(iage=iagemin-AGEMARGE; iage <= iagemax+3+AGEMARGE; iage++)
4576: prop[i][iage]=0.0;
4577: printf("Prevalence combination of varying and fixed dummies %d\n",j1);
4578: /* fprintf(ficlog," V%d=%d ",Tvaraff[j1],nbcode[Tvaraff[j1]][codtabm(k,j1)]); */
4579: fprintf(ficlog,"Prevalence combination of varying and fixed dummies %d\n",j1);
4580:
4581: for (i=1; i<=imx; i++) { /* Each individual */
4582: bool=1;
4583: /* for(m=firstpass; m<=lastpass; m++){/\* Other selection (we can limit to certain interviews*\/ */
4584: for(mi=1; mi<wav[i];mi++){ /* For this wave too look where individual can be counted V4=0 V3=0 */
4585: m=mw[mi][i];
4586: /* Tmodelind[z1]=k is the position of the varying covariate in the model, but which # within 1 to ntv? */
4587: /* Tvar[Tmodelind[z1]] is the n of Vn; n-ncovcol-nqv is the first time varying covariate or iv */
4588: for (z1=1; z1<=cptcoveff; z1++){
4589: if( Fixed[Tmodelind[z1]]==1){
4590: iv= Tvar[Tmodelind[z1]]-ncovcol-nqv;
4591: if (cotvar[m][iv][i]!= nbcode[Tvaraff[z1]][codtabm(j1,z1)]) /* iv=1 to ntv, right modality */
4592: bool=0;
4593: }else if( Fixed[Tmodelind[z1]]== 0) /* fixed */
4594: if (covar[Tvaraff[z1]][i]!= nbcode[Tvaraff[z1]][codtabm(j1,z1)]) {
4595: bool=0;
4596: }
4597: }
4598: if(bool==1){ /* Otherwise we skip that wave/person */
4599: agebegin=agev[m][i]; /* Age at beginning of wave before transition*/
4600: /* ageend=agev[m][i]+(dh[m][i])*stepm/YEARM; /\* Age at end of wave and transition *\/ */
4601: if(m >=firstpass && m <=lastpass){
4602: y2=anint[m][i]+(mint[m][i]/12.); /* Fractional date in year */
4603: if ((y2>=dateprev1) && (y2<=dateprev2)) { /* Here is the main selection (fractional years) */
4604: if(agev[m][i]==0) agev[m][i]=iagemax+1;
4605: if(agev[m][i]==1) agev[m][i]=iagemax+2;
4606: if((int)agev[m][i] <iagemin-AGEMARGE || (int)agev[m][i] >iagemax+3+AGEMARGE){
4607: 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);
4608: exit(1);
4609: }
4610: if (s[m][i]>0 && s[m][i]<=nlstate) {
4611: /*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]]);*/
4612: prop[s[m][i]][(int)agev[m][i]] += weight[i];/* At age of beginning of transition, where status is known */
4613: prop[s[m][i]][iagemax+3] += weight[i];
4614: } /* end valid statuses */
4615: } /* end selection of dates */
4616: } /* end selection of waves */
4617: } /* end bool */
4618: } /* end wave */
4619: } /* end individual */
4620: for(i=iagemin; i <= iagemax+3; i++){
4621: for(jk=1,posprop=0; jk <=nlstate ; jk++) {
4622: posprop += prop[jk][i];
4623: }
4624:
4625: for(jk=1; jk <=nlstate ; jk++){
4626: if( i <= iagemax){
4627: if(posprop>=1.e-5){
4628: probs[i][jk][j1]= prop[jk][i]/posprop;
4629: } else{
4630: if(first==1){
4631: first=0;
4632: 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]);
4633: }
4634: }
4635: }
4636: }/* end jk */
4637: }/* end i */
1.222 brouard 4638: /*} *//* end i1 */
1.227 brouard 4639: } /* end j1 */
1.222 brouard 4640:
1.227 brouard 4641: /* free_ma3x(freq,-1,nlstate+ndeath,-1,nlstate+ndeath, iagemin, iagemax+3);*/
4642: /*free_vector(pp,1,nlstate);*/
4643: free_matrix(prop,1,nlstate, iagemin-AGEMARGE,iagemax+3+AGEMARGE);
4644: } /* End of prevalence */
1.126 brouard 4645:
4646: /************* Waves Concatenation ***************/
4647:
4648: 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)
4649: {
4650: /* Concatenates waves: wav[i] is the number of effective (useful waves) of individual i.
4651: Death is a valid wave (if date is known).
4652: mw[mi][i] is the mi (mi=1 to wav[i]) effective wave of individual i
4653: dh[m][i] or dh[mw[mi][i]][i] is the delay between two effective waves m=mw[mi][i]
4654: and mw[mi+1][i]. dh depends on stepm.
1.227 brouard 4655: */
1.126 brouard 4656:
1.224 brouard 4657: int i=0, mi=0, m=0, mli=0;
1.126 brouard 4658: /* int j, k=0,jk, ju, jl,jmin=1e+5, jmax=-1;
4659: double sum=0., jmean=0.;*/
1.224 brouard 4660: int first=0, firstwo=0, firsthree=0, firstfour=0, firstfiv=0;
1.126 brouard 4661: int j, k=0,jk, ju, jl;
4662: double sum=0.;
4663: first=0;
1.214 brouard 4664: firstwo=0;
1.217 brouard 4665: firsthree=0;
1.218 brouard 4666: firstfour=0;
1.164 brouard 4667: jmin=100000;
1.126 brouard 4668: jmax=-1;
4669: jmean=0.;
1.224 brouard 4670:
4671: /* Treating live states */
1.214 brouard 4672: for(i=1; i<=imx; i++){ /* For simple cases and if state is death */
1.224 brouard 4673: mi=0; /* First valid wave */
1.227 brouard 4674: mli=0; /* Last valid wave */
1.126 brouard 4675: m=firstpass;
1.214 brouard 4676: while(s[m][i] <= nlstate){ /* a live state */
1.227 brouard 4677: 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 */
4678: mli=m-1;/* mw[++mi][i]=m-1; */
4679: }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 */
4680: mw[++mi][i]=m;
4681: mli=m;
1.224 brouard 4682: } /* else might be a useless wave -1 and mi is not incremented and mw[mi] not updated */
4683: if(m < lastpass){ /* m < lastpass, standard case */
1.227 brouard 4684: m++; /* mi gives the "effective" current wave, m the current wave, go to next wave by incrementing m */
1.216 brouard 4685: }
1.227 brouard 4686: else{ /* m >= lastpass, eventual special issue with warning */
1.224 brouard 4687: #ifdef UNKNOWNSTATUSNOTCONTRIBUTING
1.227 brouard 4688: break;
1.224 brouard 4689: #else
1.227 brouard 4690: if(s[m][i]==-1 && (int) andc[i] == 9999 && (int)anint[m][i] != 9999){
4691: if(firsthree == 0){
4692: 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 pi. .\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);
4693: firsthree=1;
4694: }
4695: 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 pi. .\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);
4696: mw[++mi][i]=m;
4697: mli=m;
4698: }
4699: if(s[m][i]==-2){ /* Vital status is really unknown */
4700: nbwarn++;
4701: if((int)anint[m][i] == 9999){ /* Has the vital status really been verified? */
4702: 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);
4703: 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);
4704: }
4705: break;
4706: }
4707: break;
1.224 brouard 4708: #endif
1.227 brouard 4709: }/* End m >= lastpass */
1.126 brouard 4710: }/* end while */
1.224 brouard 4711:
1.227 brouard 4712: /* 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 4713: /* After last pass */
1.224 brouard 4714: /* Treating death states */
1.214 brouard 4715: if (s[m][i] > nlstate){ /* In a death state */
1.227 brouard 4716: /* if( mint[m][i]==mdc[m][i] && anint[m][i]==andc[m][i]){ /\* same date of death and date of interview *\/ */
4717: /* } */
1.126 brouard 4718: mi++; /* Death is another wave */
4719: /* if(mi==0) never been interviewed correctly before death */
1.227 brouard 4720: /* Only death is a correct wave */
1.126 brouard 4721: mw[mi][i]=m;
1.224 brouard 4722: }
4723: #ifndef DISPATCHINGKNOWNDEATHAFTERLASTWAVE
1.227 brouard 4724: else if ((int) andc[i] != 9999) { /* Status is negative. A death occured after lastpass, we can't take it into account because of potential bias */
1.216 brouard 4725: /* m++; */
4726: /* mi++; */
4727: /* s[m][i]=nlstate+1; /\* We are setting the status to the last of non live state *\/ */
4728: /* mw[mi][i]=m; */
1.218 brouard 4729: if ((int)anint[m][i]!= 9999) { /* date of last interview is known */
1.227 brouard 4730: 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 */
4731: nbwarn++;
4732: if(firstfiv==0){
4733: 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 );
4734: firstfiv=1;
4735: }else{
4736: 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 );
4737: }
4738: }else{ /* Death occured afer last wave potential bias */
4739: nberr++;
4740: if(firstwo==0){
4741: 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.\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 );
4742: firstwo=1;
4743: }
4744: 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.\n",num[i],i,(int) moisdc[i], (int) andc[i], lastpass,(int)mint[m][i],(int)anint[m][i], i,m );
4745: }
1.218 brouard 4746: }else{ /* end date of interview is known */
1.227 brouard 4747: /* death is known but not confirmed by death status at any wave */
4748: if(firstfour==0){
4749: 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 );
4750: firstfour=1;
4751: }
4752: 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 4753: }
1.224 brouard 4754: } /* end if date of death is known */
4755: #endif
4756: wav[i]=mi; /* mi should be the last effective wave (or mli) */
4757: /* wav[i]=mw[mi][i]; */
1.126 brouard 4758: if(mi==0){
4759: nbwarn++;
4760: if(first==0){
1.227 brouard 4761: printf("Warning! No valid information for individual %ld line=%d (skipped) and may be others, see log file\n",num[i],i);
4762: first=1;
1.126 brouard 4763: }
4764: if(first==1){
1.227 brouard 4765: fprintf(ficlog,"Warning! No valid information for individual %ld line=%d (skipped)\n",num[i],i);
1.126 brouard 4766: }
4767: } /* end mi==0 */
4768: } /* End individuals */
1.214 brouard 4769: /* wav and mw are no more changed */
1.223 brouard 4770:
1.214 brouard 4771:
1.126 brouard 4772: for(i=1; i<=imx; i++){
4773: for(mi=1; mi<wav[i];mi++){
4774: if (stepm <=0)
1.227 brouard 4775: dh[mi][i]=1;
1.126 brouard 4776: else{
1.227 brouard 4777: if (s[mw[mi+1][i]][i] > nlstate) { /* A death */
4778: if (agedc[i] < 2*AGESUP) {
4779: j= rint(agedc[i]*12-agev[mw[mi][i]][i]*12);
4780: if(j==0) j=1; /* Survives at least one month after exam */
4781: else if(j<0){
4782: nberr++;
4783: 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]);
4784: j=1; /* Temporary Dangerous patch */
4785: 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);
4786: 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]);
4787: 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);
4788: }
4789: k=k+1;
4790: if (j >= jmax){
4791: jmax=j;
4792: ijmax=i;
4793: }
4794: if (j <= jmin){
4795: jmin=j;
4796: ijmin=i;
4797: }
4798: sum=sum+j;
4799: /*if (j<0) printf("j=%d num=%d \n",j,i);*/
4800: /* printf("%d %d %d %d\n", s[mw[mi][i]][i] ,s[mw[mi+1][i]][i],j,i);*/
4801: }
4802: }
4803: else{
4804: j= rint( (agev[mw[mi+1][i]][i]*12 - agev[mw[mi][i]][i]*12));
1.126 brouard 4805: /* 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 4806:
1.227 brouard 4807: k=k+1;
4808: if (j >= jmax) {
4809: jmax=j;
4810: ijmax=i;
4811: }
4812: else if (j <= jmin){
4813: jmin=j;
4814: ijmin=i;
4815: }
4816: /* if (j<10) printf("j=%d jmin=%d num=%d ",j,jmin,i); */
4817: /*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]);*/
4818: if(j<0){
4819: nberr++;
4820: 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]);
4821: 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]);
4822: }
4823: sum=sum+j;
4824: }
4825: jk= j/stepm;
4826: jl= j -jk*stepm;
4827: ju= j -(jk+1)*stepm;
4828: if(mle <=1){ /* only if we use a the linear-interpoloation pseudo-likelihood */
4829: if(jl==0){
4830: dh[mi][i]=jk;
4831: bh[mi][i]=0;
4832: }else{ /* We want a negative bias in order to only have interpolation ie
4833: * to avoid the price of an extra matrix product in likelihood */
4834: dh[mi][i]=jk+1;
4835: bh[mi][i]=ju;
4836: }
4837: }else{
4838: if(jl <= -ju){
4839: dh[mi][i]=jk;
4840: bh[mi][i]=jl; /* bias is positive if real duration
4841: * is higher than the multiple of stepm and negative otherwise.
4842: */
4843: }
4844: else{
4845: dh[mi][i]=jk+1;
4846: bh[mi][i]=ju;
4847: }
4848: if(dh[mi][i]==0){
4849: dh[mi][i]=1; /* At least one step */
4850: bh[mi][i]=ju; /* At least one step */
4851: /* 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);*/
4852: }
4853: } /* end if mle */
1.126 brouard 4854: }
4855: } /* end wave */
4856: }
4857: jmean=sum/k;
4858: 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 4859: 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 4860: }
1.126 brouard 4861:
4862: /*********** Tricode ****************************/
1.220 brouard 4863: void tricode(int *cptcov, int *Tvar, int **nbcode, int imx, int *Ndum)
1.242 brouard 4864: {
4865: /**< Uses cptcovn+2*cptcovprod as the number of covariates */
4866: /* Tvar[i]=atoi(stre); find 'n' in Vn and stores in Tvar. If model=V2+V1 Tvar[1]=2 and Tvar[2]=1
4867: * Boring subroutine which should only output nbcode[Tvar[j]][k]
4868: * Tvar[5] in V2+V1+V3*age+V2*V4 is 4 (V4) even it is a time varying or quantitative variable
4869: * nbcode[Tvar[5]][1]= nbcode[4][1]=0, nbcode[4][2]=1 (usually);
4870: */
1.130 brouard 4871:
1.242 brouard 4872: int ij=1, k=0, j=0, i=0, maxncov=NCOVMAX;
4873: int modmaxcovj=0; /* Modality max of covariates j */
4874: int cptcode=0; /* Modality max of covariates j */
4875: int modmincovj=0; /* Modality min of covariates j */
1.145 brouard 4876:
4877:
1.242 brouard 4878: /* cptcoveff=0; */
4879: /* *cptcov=0; */
1.126 brouard 4880:
1.242 brouard 4881: for (k=1; k <= maxncov; k++) ncodemax[k]=0; /* Horrible constant again replaced by NCOVMAX */
1.126 brouard 4882:
1.242 brouard 4883: /* Loop on covariates without age and products and no quantitative variable */
4884: /* for (j=1; j<=(cptcovs); j++) { /\* From model V1 + V2*age+ V3 + V3*V4 keeps V1 + V3 = 2 only *\/ */
4885: for (k=1; k<=cptcovt; k++) { /* From model V1 + V2*age + V3 + V3*V4 keeps V1 + V3 = 2 only */
4886: for (j=-1; (j < maxncov); j++) Ndum[j]=0;
4887: if(Dummy[k]==0 && Typevar[k] !=1){ /* Dummy covariate and not age product */
4888: switch(Fixed[k]) {
4889: case 0: /* Testing on fixed dummy covariate, simple or product of fixed */
4890: 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*/
4891: ij=(int)(covar[Tvar[k]][i]);
4892: /* ij=0 or 1 or -1. Value of the covariate Tvar[j] for individual i
4893: * If product of Vn*Vm, still boolean *:
4894: * If it was coded 1, 2, 3, 4 should be splitted into 3 boolean variables
4895: * 1 => 0 0 0, 2 => 0 0 1, 3 => 0 1 1, 4=1 0 0 */
4896: /* Finds for covariate j, n=Tvar[j] of Vn . ij is the
4897: modality of the nth covariate of individual i. */
4898: if (ij > modmaxcovj)
4899: modmaxcovj=ij;
4900: else if (ij < modmincovj)
4901: modmincovj=ij;
4902: if ((ij < -1) && (ij > NCOVMAX)){
4903: printf( "Error: minimal is less than -1 or maximal is bigger than %d. Exiting. \n", NCOVMAX );
4904: exit(1);
4905: }else
4906: Ndum[ij]++; /*counts and stores the occurence of this modality 0, 1, -1*/
4907: /* If coded 1, 2, 3 , counts the number of 1 Ndum[1], number of 2, Ndum[2], etc */
4908: /*printf("i=%d ij=%d Ndum[ij]=%d imx=%d",i,ij,Ndum[ij],imx);*/
4909: /* getting the maximum value of the modality of the covariate
4910: (should be 0 or 1 now) Tvar[j]. If V=sex and male is coded 0 and
4911: female ies 1, then modmaxcovj=1.
4912: */
4913: } /* end for loop on individuals i */
4914: printf(" Minimal and maximal values of %d th (fixed) covariate V%d: min=%d max=%d \n", k, Tvar[k], modmincovj, modmaxcovj);
4915: fprintf(ficlog," Minimal and maximal values of %d th (fixed) covariate V%d: min=%d max=%d \n", k, Tvar[k], modmincovj, modmaxcovj);
4916: cptcode=modmaxcovj;
4917: /* Ndum[0] = frequency of 0 for model-covariate j, Ndum[1] frequency of 1 etc. */
4918: /*for (i=0; i<=cptcode; i++) {*/
4919: for (j=modmincovj; j<=modmaxcovj; j++) { /* j=-1 ? 0 and 1*//* For each value j of the modality of model-cov k */
4920: printf("Frequencies of (fixed) covariate %d ie V%d with value %d: %d\n", k, Tvar[k], j, Ndum[j]);
4921: fprintf(ficlog, "Frequencies of (fixed) covariate %d ie V%d with value %d: %d\n", k, Tvar[k], j, Ndum[j]);
4922: if( Ndum[j] != 0 ){ /* Counts if nobody answered modality j ie empty modality, we skip it and reorder */
4923: if( j != -1){
4924: ncodemax[k]++; /* ncodemax[k]= Number of modalities of the k th
4925: covariate for which somebody answered excluding
4926: undefined. Usually 2: 0 and 1. */
4927: }
4928: ncodemaxwundef[k]++; /* ncodemax[j]= Number of modalities of the k th
4929: covariate for which somebody answered including
4930: undefined. Usually 3: -1, 0 and 1. */
4931: } /* In fact ncodemax[k]=2 (dichotom. variables only) but it could be more for
4932: * historical reasons: 3 if coded 1, 2, 3 and 4 and Ndum[2]=0 */
4933: } /* Ndum[-1] number of undefined modalities */
1.231 brouard 4934:
1.242 brouard 4935: /* j is a covariate, n=Tvar[j] of Vn; Fills nbcode */
4936: /* For covariate j, modalities could be 1, 2, 3, 4, 5, 6, 7. */
4937: /* If Ndum[1]=0, Ndum[2]=0, Ndum[3]= 635, Ndum[4]=0, Ndum[5]=0, Ndum[6]=27, Ndum[7]=125; */
4938: /* modmincovj=3; modmaxcovj = 7; */
4939: /* There are only 3 modalities non empty 3, 6, 7 (or 2 if 27 is too few) : ncodemax[j]=3; */
4940: /* which will be coded 0, 1, 2 which in binary on 2=3-1 digits are 0=00 1=01, 2=10; */
4941: /* defining two dummy variables: variables V1_1 and V1_2.*/
4942: /* nbcode[Tvar[j]][ij]=k; */
4943: /* nbcode[Tvar[j]][1]=0; */
4944: /* nbcode[Tvar[j]][2]=1; */
4945: /* nbcode[Tvar[j]][3]=2; */
4946: /* To be continued (not working yet). */
4947: ij=0; /* ij is similar to i but can jump over null modalities */
4948: 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*/
4949: if (Ndum[i] == 0) { /* If nobody responded to this modality k */
4950: break;
4951: }
4952: ij++;
4953: 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*/
4954: cptcode = ij; /* New max modality for covar j */
4955: } /* end of loop on modality i=-1 to 1 or more */
4956: break;
4957: case 1: /* Testing on varying covariate, could be simple and
4958: * should look at waves or product of fixed *
4959: * varying. No time to test -1, assuming 0 and 1 only */
4960: ij=0;
4961: for(i=0; i<=1;i++){
4962: nbcode[Tvar[k]][++ij]=i;
4963: }
4964: break;
4965: default:
4966: break;
4967: } /* end switch */
4968: } /* end dummy test */
4969:
4970: /* for (k=0; k<= cptcode; k++) { /\* k=-1 ? k=0 to 1 *\//\* Could be 1 to 4 *\//\* cptcode=modmaxcovj *\/ */
4971: /* /\*recode from 0 *\/ */
4972: /* k is a modality. If we have model=V1+V1*sex */
4973: /* then: nbcode[1][1]=0 ; nbcode[1][2]=1; nbcode[2][1]=0 ; nbcode[2][2]=1; */
4974: /* But if some modality were not used, it is recoded from 0 to a newer modmaxcovj=cptcode *\/ */
4975: /* } */
4976: /* /\* cptcode = ij; *\/ /\* New max modality for covar j *\/ */
4977: /* if (ij > ncodemax[j]) { */
4978: /* printf( " Error ij=%d > ncodemax[%d]=%d\n", ij, j, ncodemax[j]); */
4979: /* fprintf(ficlog, " Error ij=%d > ncodemax[%d]=%d\n", ij, j, ncodemax[j]); */
4980: /* break; */
4981: /* } */
4982: /* } /\* end of loop on modality k *\/ */
4983: } /* end of loop on model-covariate j. nbcode[Tvarj][1]=0 and nbcode[Tvarj][2]=1 sets the value of covariate j*/
4984:
4985: for (k=-1; k< maxncov; k++) Ndum[k]=0;
4986: /* Look at fixed dummy (single or product) covariates to check empty modalities */
4987: for (i=1; i<=ncovmodel-2-nagesqr; i++) { /* -2, cste and age and eventually age*age */
4988: /* Listing of all covariables in statement model to see if some covariates appear twice. For example, V1 appears twice in V1+V1*V2.*/
4989: 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 */
4990: 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 */
4991: /* V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1, {2, 1, 1, 1, 2, 1, 1, 0, 0} */
4992: } /* V4+V3+V5, Ndum[1]@5={0, 0, 1, 1, 1} */
4993:
4994: ij=0;
4995: /* for (i=0; i<= maxncov-1; i++) { /\* modmaxcovj is unknown here. Only Ndum[2(V2),3(age*V3), 5(V3*V2) 6(V1*V4) *\/ */
4996: for (k=1; k<= cptcovt; k++) { /* modmaxcovj is unknown here. Only Ndum[2(V2),3(age*V3), 5(V3*V2) 6(V1*V4) */
4997: /*printf("Ndum[%d]=%d\n",i, Ndum[i]);*/
4998: /* if((Ndum[i]!=0) && (i<=ncovcol)){ /\* Tvar[i] <= ncovmodel ? *\/ */
4999: if(Ndum[Tvar[k]]!=0 && Dummy[k] == 0 && Typevar[k]==0){ /* Only Dummy and non empty in the model */
5000: /* If product not in single variable we don't print results */
5001: /*printf("diff Ndum[%d]=%d\n",i, Ndum[i]);*/
5002: ++ij;/* V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1, */
5003: 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*/
5004: Tmodelind[ij]=k; /* Tmodelind: index in model of dummies Tmodelind[1]=2 V4: pos=2; V3: pos=3, V1=9 {2, 3, 9, ?, ?,} */
5005: 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 */
5006: if(Fixed[k]!=0)
5007: anyvaryingduminmodel=1;
5008: /* }else if((Ndum[i]!=0) && (i<=ncovcol+nqv)){ */
5009: /* Tvaraff[++ij]=-10; /\* Dont'n know how to treat quantitative variables yet *\/ */
5010: /* }else if((Ndum[i]!=0) && (i<=ncovcol+nqv+ntv)){ */
5011: /* Tvaraff[++ij]=i; /\*For printing (unclear) *\/ */
5012: /* }else if((Ndum[i]!=0) && (i<=ncovcol+nqv+ntv+nqtv)){ */
5013: /* Tvaraff[++ij]=-20; /\* Dont'n know how to treat quantitative variables yet *\/ */
5014: }
5015: } /* Tvaraff[1]@5 {3, 4, -20, 0, 0} Very strange */
5016: /* ij--; */
5017: /* cptcoveff=ij; /\*Number of total covariates*\/ */
5018: *cptcov=ij; /*Number of total real effective covariates: effective
5019: * because they can be excluded from the model and real
5020: * if in the model but excluded because missing values, but how to get k from ij?*/
5021: for(j=ij+1; j<= cptcovt; j++){
5022: Tvaraff[j]=0;
5023: Tmodelind[j]=0;
5024: }
5025: for(j=ntveff+1; j<= cptcovt; j++){
5026: TmodelInvind[j]=0;
5027: }
5028: /* To be sorted */
5029: ;
5030: }
1.126 brouard 5031:
1.145 brouard 5032:
1.126 brouard 5033: /*********** Health Expectancies ****************/
5034:
1.235 brouard 5035: 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 5036:
5037: {
5038: /* Health expectancies, no variances */
1.164 brouard 5039: int i, j, nhstepm, hstepm, h, nstepm;
1.126 brouard 5040: int nhstepma, nstepma; /* Decreasing with age */
5041: double age, agelim, hf;
5042: double ***p3mat;
5043: double eip;
5044:
1.238 brouard 5045: /* pstamp(ficreseij); */
1.126 brouard 5046: fprintf(ficreseij,"# (a) Life expectancies by health status at initial age and (b) health expectancies by health status at initial age\n");
5047: fprintf(ficreseij,"# Age");
5048: for(i=1; i<=nlstate;i++){
5049: for(j=1; j<=nlstate;j++){
5050: fprintf(ficreseij," e%1d%1d ",i,j);
5051: }
5052: fprintf(ficreseij," e%1d. ",i);
5053: }
5054: fprintf(ficreseij,"\n");
5055:
5056:
5057: if(estepm < stepm){
5058: printf ("Problem %d lower than %d\n",estepm, stepm);
5059: }
5060: else hstepm=estepm;
5061: /* We compute the life expectancy from trapezoids spaced every estepm months
5062: * This is mainly to measure the difference between two models: for example
5063: * if stepm=24 months pijx are given only every 2 years and by summing them
5064: * we are calculating an estimate of the Life Expectancy assuming a linear
5065: * progression in between and thus overestimating or underestimating according
5066: * to the curvature of the survival function. If, for the same date, we
5067: * estimate the model with stepm=1 month, we can keep estepm to 24 months
5068: * to compare the new estimate of Life expectancy with the same linear
5069: * hypothesis. A more precise result, taking into account a more precise
5070: * curvature will be obtained if estepm is as small as stepm. */
5071:
5072: /* For example we decided to compute the life expectancy with the smallest unit */
5073: /* hstepm beeing the number of stepms, if hstepm=1 the length of hstepm is stepm.
5074: nhstepm is the number of hstepm from age to agelim
5075: nstepm is the number of stepm from age to agelin.
5076: Look at hpijx to understand the reason of that which relies in memory size
5077: and note for a fixed period like estepm months */
5078: /* We decided (b) to get a life expectancy respecting the most precise curvature of the
5079: survival function given by stepm (the optimization length). Unfortunately it
5080: means that if the survival funtion is printed only each two years of age and if
5081: you sum them up and add 1 year (area under the trapezoids) you won't get the same
5082: results. So we changed our mind and took the option of the best precision.
5083: */
5084: hstepm=hstepm/stepm; /* Typically in stepm units, if stepm=6 & estepm=24 , = 24/6 months = 4 */
5085:
5086: agelim=AGESUP;
5087: /* If stepm=6 months */
5088: /* Computed by stepm unit matrices, product of hstepm matrices, stored
5089: in an array of nhstepm length: nhstepm=10, hstepm=4, stepm=6 months */
5090:
5091: /* nhstepm age range expressed in number of stepm */
5092: nstepm=(int) rint((agelim-bage)*YEARM/stepm); /* Biggest nstepm */
5093: /* Typically if 20 years nstepm = 20*12/6=40 stepm */
5094: /* if (stepm >= YEARM) hstepm=1;*/
5095: nhstepm = nstepm/hstepm;/* Expressed in hstepm, typically nhstepm=40/4=10 */
5096: p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5097:
5098: for (age=bage; age<=fage; age ++){
5099: nstepma=(int) rint((agelim-bage)*YEARM/stepm); /* Biggest nstepm */
5100: /* Typically if 20 years nstepm = 20*12/6=40 stepm */
5101: /* if (stepm >= YEARM) hstepm=1;*/
5102: nhstepma = nstepma/hstepm;/* Expressed in hstepm, typically nhstepma=40/4=10 */
5103:
5104: /* If stepm=6 months */
5105: /* Computed by stepm unit matrices, product of hstepma matrices, stored
5106: in an array of nhstepma length: nhstepma=10, hstepm=4, stepm=6 months */
5107:
1.235 brouard 5108: hpxij(p3mat,nhstepma,age,hstepm,x,nlstate,stepm,oldm, savm, cij, nres);
1.126 brouard 5109:
5110: hf=hstepm*stepm/YEARM; /* Duration of hstepm expressed in year unit. */
5111:
5112: printf("%d|",(int)age);fflush(stdout);
5113: fprintf(ficlog,"%d|",(int)age);fflush(ficlog);
5114:
5115: /* Computing expectancies */
5116: for(i=1; i<=nlstate;i++)
5117: for(j=1; j<=nlstate;j++)
5118: for (h=0, eij[i][j][(int)age]=0; h<=nhstepm-1; h++){
5119: eij[i][j][(int)age] += (p3mat[i][j][h]+p3mat[i][j][h+1])/2.0*hf;
5120:
5121: /* 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]);*/
5122:
5123: }
5124:
5125: fprintf(ficreseij,"%3.0f",age );
5126: for(i=1; i<=nlstate;i++){
5127: eip=0;
5128: for(j=1; j<=nlstate;j++){
5129: eip +=eij[i][j][(int)age];
5130: fprintf(ficreseij,"%9.4f", eij[i][j][(int)age] );
5131: }
5132: fprintf(ficreseij,"%9.4f", eip );
5133: }
5134: fprintf(ficreseij,"\n");
5135:
5136: }
5137: free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5138: printf("\n");
5139: fprintf(ficlog,"\n");
5140:
5141: }
5142:
1.235 brouard 5143: 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 5144:
5145: {
5146: /* Covariances of health expectancies eij and of total life expectancies according
1.222 brouard 5147: to initial status i, ei. .
1.126 brouard 5148: */
5149: int i, j, nhstepm, hstepm, h, nstepm, k, cptj, cptj2, i2, j2, ij, ji;
5150: int nhstepma, nstepma; /* Decreasing with age */
5151: double age, agelim, hf;
5152: double ***p3matp, ***p3matm, ***varhe;
5153: double **dnewm,**doldm;
5154: double *xp, *xm;
5155: double **gp, **gm;
5156: double ***gradg, ***trgradg;
5157: int theta;
5158:
5159: double eip, vip;
5160:
5161: varhe=ma3x(1,nlstate*nlstate,1,nlstate*nlstate,(int) bage, (int) fage);
5162: xp=vector(1,npar);
5163: xm=vector(1,npar);
5164: dnewm=matrix(1,nlstate*nlstate,1,npar);
5165: doldm=matrix(1,nlstate*nlstate,1,nlstate*nlstate);
5166:
5167: pstamp(ficresstdeij);
5168: fprintf(ficresstdeij,"# Health expectancies with standard errors\n");
5169: fprintf(ficresstdeij,"# Age");
5170: for(i=1; i<=nlstate;i++){
5171: for(j=1; j<=nlstate;j++)
5172: fprintf(ficresstdeij," e%1d%1d (SE)",i,j);
5173: fprintf(ficresstdeij," e%1d. ",i);
5174: }
5175: fprintf(ficresstdeij,"\n");
5176:
5177: pstamp(ficrescveij);
5178: fprintf(ficrescveij,"# Subdiagonal matrix of covariances of health expectancies by age: cov(eij,ekl)\n");
5179: fprintf(ficrescveij,"# Age");
5180: for(i=1; i<=nlstate;i++)
5181: for(j=1; j<=nlstate;j++){
5182: cptj= (j-1)*nlstate+i;
5183: for(i2=1; i2<=nlstate;i2++)
5184: for(j2=1; j2<=nlstate;j2++){
5185: cptj2= (j2-1)*nlstate+i2;
5186: if(cptj2 <= cptj)
5187: fprintf(ficrescveij," %1d%1d,%1d%1d",i,j,i2,j2);
5188: }
5189: }
5190: fprintf(ficrescveij,"\n");
5191:
5192: if(estepm < stepm){
5193: printf ("Problem %d lower than %d\n",estepm, stepm);
5194: }
5195: else hstepm=estepm;
5196: /* We compute the life expectancy from trapezoids spaced every estepm months
5197: * This is mainly to measure the difference between two models: for example
5198: * if stepm=24 months pijx are given only every 2 years and by summing them
5199: * we are calculating an estimate of the Life Expectancy assuming a linear
5200: * progression in between and thus overestimating or underestimating according
5201: * to the curvature of the survival function. If, for the same date, we
5202: * estimate the model with stepm=1 month, we can keep estepm to 24 months
5203: * to compare the new estimate of Life expectancy with the same linear
5204: * hypothesis. A more precise result, taking into account a more precise
5205: * curvature will be obtained if estepm is as small as stepm. */
5206:
5207: /* For example we decided to compute the life expectancy with the smallest unit */
5208: /* hstepm beeing the number of stepms, if hstepm=1 the length of hstepm is stepm.
5209: nhstepm is the number of hstepm from age to agelim
5210: nstepm is the number of stepm from age to agelin.
5211: Look at hpijx to understand the reason of that which relies in memory size
5212: and note for a fixed period like estepm months */
5213: /* We decided (b) to get a life expectancy respecting the most precise curvature of the
5214: survival function given by stepm (the optimization length). Unfortunately it
5215: means that if the survival funtion is printed only each two years of age and if
5216: you sum them up and add 1 year (area under the trapezoids) you won't get the same
5217: results. So we changed our mind and took the option of the best precision.
5218: */
5219: hstepm=hstepm/stepm; /* Typically in stepm units, if stepm=6 & estepm=24 , = 24/6 months = 4 */
5220:
5221: /* If stepm=6 months */
5222: /* nhstepm age range expressed in number of stepm */
5223: agelim=AGESUP;
5224: nstepm=(int) rint((agelim-bage)*YEARM/stepm);
5225: /* Typically if 20 years nstepm = 20*12/6=40 stepm */
5226: /* if (stepm >= YEARM) hstepm=1;*/
5227: nhstepm = nstepm/hstepm;/* Expressed in hstepm, typically nhstepm=40/4=10 */
5228:
5229: p3matp=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5230: p3matm=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5231: gradg=ma3x(0,nhstepm,1,npar,1,nlstate*nlstate);
5232: trgradg =ma3x(0,nhstepm,1,nlstate*nlstate,1,npar);
5233: gp=matrix(0,nhstepm,1,nlstate*nlstate);
5234: gm=matrix(0,nhstepm,1,nlstate*nlstate);
5235:
5236: for (age=bage; age<=fage; age ++){
5237: nstepma=(int) rint((agelim-bage)*YEARM/stepm); /* Biggest nstepm */
5238: /* Typically if 20 years nstepm = 20*12/6=40 stepm */
5239: /* if (stepm >= YEARM) hstepm=1;*/
5240: nhstepma = nstepma/hstepm;/* Expressed in hstepm, typically nhstepma=40/4=10 */
1.218 brouard 5241:
1.126 brouard 5242: /* If stepm=6 months */
5243: /* Computed by stepm unit matrices, product of hstepma matrices, stored
5244: in an array of nhstepma length: nhstepma=10, hstepm=4, stepm=6 months */
5245:
5246: hf=hstepm*stepm/YEARM; /* Duration of hstepm expressed in year unit. */
1.218 brouard 5247:
1.126 brouard 5248: /* Computing Variances of health expectancies */
5249: /* Gradient is computed with plus gp and minus gm. Code is duplicated in order to
5250: decrease memory allocation */
5251: for(theta=1; theta <=npar; theta++){
5252: for(i=1; i<=npar; i++){
1.222 brouard 5253: xp[i] = x[i] + (i==theta ?delti[theta]:0);
5254: xm[i] = x[i] - (i==theta ?delti[theta]:0);
1.126 brouard 5255: }
1.235 brouard 5256: hpxij(p3matp,nhstepm,age,hstepm,xp,nlstate,stepm,oldm,savm, cij, nres);
5257: hpxij(p3matm,nhstepm,age,hstepm,xm,nlstate,stepm,oldm,savm, cij, nres);
1.218 brouard 5258:
1.126 brouard 5259: for(j=1; j<= nlstate; j++){
1.222 brouard 5260: for(i=1; i<=nlstate; i++){
5261: for(h=0; h<=nhstepm-1; h++){
5262: gp[h][(j-1)*nlstate + i] = (p3matp[i][j][h]+p3matp[i][j][h+1])/2.;
5263: gm[h][(j-1)*nlstate + i] = (p3matm[i][j][h]+p3matm[i][j][h+1])/2.;
5264: }
5265: }
1.126 brouard 5266: }
1.218 brouard 5267:
1.126 brouard 5268: for(ij=1; ij<= nlstate*nlstate; ij++)
1.222 brouard 5269: for(h=0; h<=nhstepm-1; h++){
5270: gradg[h][theta][ij]= (gp[h][ij]-gm[h][ij])/2./delti[theta];
5271: }
1.126 brouard 5272: }/* End theta */
5273:
5274:
5275: for(h=0; h<=nhstepm-1; h++)
5276: for(j=1; j<=nlstate*nlstate;j++)
1.222 brouard 5277: for(theta=1; theta <=npar; theta++)
5278: trgradg[h][j][theta]=gradg[h][theta][j];
1.126 brouard 5279:
1.218 brouard 5280:
1.222 brouard 5281: for(ij=1;ij<=nlstate*nlstate;ij++)
1.126 brouard 5282: for(ji=1;ji<=nlstate*nlstate;ji++)
1.222 brouard 5283: varhe[ij][ji][(int)age] =0.;
1.218 brouard 5284:
1.222 brouard 5285: printf("%d|",(int)age);fflush(stdout);
5286: fprintf(ficlog,"%d|",(int)age);fflush(ficlog);
5287: for(h=0;h<=nhstepm-1;h++){
1.126 brouard 5288: for(k=0;k<=nhstepm-1;k++){
1.222 brouard 5289: matprod2(dnewm,trgradg[h],1,nlstate*nlstate,1,npar,1,npar,matcov);
5290: matprod2(doldm,dnewm,1,nlstate*nlstate,1,npar,1,nlstate*nlstate,gradg[k]);
5291: for(ij=1;ij<=nlstate*nlstate;ij++)
5292: for(ji=1;ji<=nlstate*nlstate;ji++)
5293: varhe[ij][ji][(int)age] += doldm[ij][ji]*hf*hf;
1.126 brouard 5294: }
5295: }
1.218 brouard 5296:
1.126 brouard 5297: /* Computing expectancies */
1.235 brouard 5298: hpxij(p3matm,nhstepm,age,hstepm,x,nlstate,stepm,oldm, savm, cij,nres);
1.126 brouard 5299: for(i=1; i<=nlstate;i++)
5300: for(j=1; j<=nlstate;j++)
1.222 brouard 5301: for (h=0, eij[i][j][(int)age]=0; h<=nhstepm-1; h++){
5302: eij[i][j][(int)age] += (p3matm[i][j][h]+p3matm[i][j][h+1])/2.0*hf;
1.218 brouard 5303:
1.222 brouard 5304: /* 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 5305:
1.222 brouard 5306: }
1.218 brouard 5307:
1.126 brouard 5308: fprintf(ficresstdeij,"%3.0f",age );
5309: for(i=1; i<=nlstate;i++){
5310: eip=0.;
5311: vip=0.;
5312: for(j=1; j<=nlstate;j++){
1.222 brouard 5313: eip += eij[i][j][(int)age];
5314: for(k=1; k<=nlstate;k++) /* Sum on j and k of cov(eij,eik) */
5315: vip += varhe[(j-1)*nlstate+i][(k-1)*nlstate+i][(int)age];
5316: 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 5317: }
5318: fprintf(ficresstdeij," %9.4f (%.4f)", eip, sqrt(vip));
5319: }
5320: fprintf(ficresstdeij,"\n");
1.218 brouard 5321:
1.126 brouard 5322: fprintf(ficrescveij,"%3.0f",age );
5323: for(i=1; i<=nlstate;i++)
5324: for(j=1; j<=nlstate;j++){
1.222 brouard 5325: cptj= (j-1)*nlstate+i;
5326: for(i2=1; i2<=nlstate;i2++)
5327: for(j2=1; j2<=nlstate;j2++){
5328: cptj2= (j2-1)*nlstate+i2;
5329: if(cptj2 <= cptj)
5330: fprintf(ficrescveij," %.4f", varhe[cptj][cptj2][(int)age]);
5331: }
1.126 brouard 5332: }
5333: fprintf(ficrescveij,"\n");
1.218 brouard 5334:
1.126 brouard 5335: }
5336: free_matrix(gm,0,nhstepm,1,nlstate*nlstate);
5337: free_matrix(gp,0,nhstepm,1,nlstate*nlstate);
5338: free_ma3x(gradg,0,nhstepm,1,npar,1,nlstate*nlstate);
5339: free_ma3x(trgradg,0,nhstepm,1,nlstate*nlstate,1,npar);
5340: free_ma3x(p3matm,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5341: free_ma3x(p3matp,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5342: printf("\n");
5343: fprintf(ficlog,"\n");
1.218 brouard 5344:
1.126 brouard 5345: free_vector(xm,1,npar);
5346: free_vector(xp,1,npar);
5347: free_matrix(dnewm,1,nlstate*nlstate,1,npar);
5348: free_matrix(doldm,1,nlstate*nlstate,1,nlstate*nlstate);
5349: free_ma3x(varhe,1,nlstate*nlstate,1,nlstate*nlstate,(int) bage, (int)fage);
5350: }
1.218 brouard 5351:
1.126 brouard 5352: /************ Variance ******************/
1.235 brouard 5353: 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 5354: {
5355: /* Variance of health expectancies */
5356: /* double **prevalim(double **prlim, int nlstate, double *xp, double age, double **oldm, double ** savm,double ftolpl);*/
5357: /* double **newm;*/
5358: /* int movingaverage(double ***probs, double bage,double fage, double ***mobaverage, int mobilav)*/
5359:
5360: /* int movingaverage(); */
5361: double **dnewm,**doldm;
5362: double **dnewmp,**doldmp;
5363: int i, j, nhstepm, hstepm, h, nstepm ;
5364: int k;
5365: double *xp;
5366: double **gp, **gm; /* for var eij */
5367: double ***gradg, ***trgradg; /*for var eij */
5368: double **gradgp, **trgradgp; /* for var p point j */
5369: double *gpp, *gmp; /* for var p point j */
5370: double **varppt; /* for var p point j nlstate to nlstate+ndeath */
5371: double ***p3mat;
5372: double age,agelim, hf;
5373: /* double ***mobaverage; */
5374: int theta;
5375: char digit[4];
5376: char digitp[25];
5377:
5378: char fileresprobmorprev[FILENAMELENGTH];
5379:
5380: if(popbased==1){
5381: if(mobilav!=0)
5382: strcpy(digitp,"-POPULBASED-MOBILAV_");
5383: else strcpy(digitp,"-POPULBASED-NOMOBIL_");
5384: }
5385: else
5386: strcpy(digitp,"-STABLBASED_");
1.126 brouard 5387:
1.218 brouard 5388: /* if (mobilav!=0) { */
5389: /* mobaverage= ma3x(1, AGESUP,1,NCOVMAX, 1,NCOVMAX); */
5390: /* if (movingaverage(probs, bage, fage, mobaverage,mobilav)!=0){ */
5391: /* fprintf(ficlog," Error in movingaverage mobilav=%d\n",mobilav); */
5392: /* printf(" Error in movingaverage mobilav=%d\n",mobilav); */
5393: /* } */
5394: /* } */
5395:
5396: strcpy(fileresprobmorprev,"PRMORPREV-");
5397: sprintf(digit,"%-d",ij);
5398: /*printf("DIGIT=%s, ij=%d ijr=%-d|\n",digit, ij,ij);*/
5399: strcat(fileresprobmorprev,digit); /* Tvar to be done */
5400: strcat(fileresprobmorprev,digitp); /* Popbased or not, mobilav or not */
5401: strcat(fileresprobmorprev,fileresu);
5402: if((ficresprobmorprev=fopen(fileresprobmorprev,"w"))==NULL) {
5403: printf("Problem with resultfile: %s\n", fileresprobmorprev);
5404: fprintf(ficlog,"Problem with resultfile: %s\n", fileresprobmorprev);
5405: }
5406: printf("Computing total mortality p.j=w1*p1j+w2*p2j+..: result on file '%s' \n",fileresprobmorprev);
5407: fprintf(ficlog,"Computing total mortality p.j=w1*p1j+w2*p2j+..: result on file '%s' \n",fileresprobmorprev);
5408: pstamp(ficresprobmorprev);
5409: 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 5410: fprintf(ficresprobmorprev,"# Selected quantitative variables and dummies");
5411: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
5412: fprintf(ficresprobmorprev," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
5413: }
5414: for(j=1;j<=cptcoveff;j++)
5415: fprintf(ficresprobmorprev,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(ij,j)]);
5416: fprintf(ficresprobmorprev,"\n");
5417:
1.218 brouard 5418: fprintf(ficresprobmorprev,"# Age cov=%-d",ij);
5419: for(j=nlstate+1; j<=(nlstate+ndeath);j++){
5420: fprintf(ficresprobmorprev," p.%-d SE",j);
5421: for(i=1; i<=nlstate;i++)
5422: fprintf(ficresprobmorprev," w%1d p%-d%-d",i,i,j);
5423: }
5424: fprintf(ficresprobmorprev,"\n");
5425:
5426: fprintf(ficgp,"\n# Routine varevsij");
5427: fprintf(ficgp,"\nunset title \n");
5428: /* fprintf(fichtm, "#Local time at start: %s", strstart);*/
5429: 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");
5430: fprintf(fichtm,"\n<br>%s <br>\n",digitp);
5431: /* } */
5432: varppt = matrix(nlstate+1,nlstate+ndeath,nlstate+1,nlstate+ndeath);
5433: pstamp(ficresvij);
5434: fprintf(ficresvij,"# Variance and covariance of health expectancies e.j \n# (weighted average of eij where weights are ");
5435: if(popbased==1)
5436: 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);
5437: else
5438: fprintf(ficresvij,"the age specific period (stable) prevalences in each health state \n");
5439: fprintf(ficresvij,"# Age");
5440: for(i=1; i<=nlstate;i++)
5441: for(j=1; j<=nlstate;j++)
5442: fprintf(ficresvij," Cov(e.%1d, e.%1d)",i,j);
5443: fprintf(ficresvij,"\n");
5444:
5445: xp=vector(1,npar);
5446: dnewm=matrix(1,nlstate,1,npar);
5447: doldm=matrix(1,nlstate,1,nlstate);
5448: dnewmp= matrix(nlstate+1,nlstate+ndeath,1,npar);
5449: doldmp= matrix(nlstate+1,nlstate+ndeath,nlstate+1,nlstate+ndeath);
5450:
5451: gradgp=matrix(1,npar,nlstate+1,nlstate+ndeath);
5452: gpp=vector(nlstate+1,nlstate+ndeath);
5453: gmp=vector(nlstate+1,nlstate+ndeath);
5454: trgradgp =matrix(nlstate+1,nlstate+ndeath,1,npar); /* mu or p point j*/
1.126 brouard 5455:
1.218 brouard 5456: if(estepm < stepm){
5457: printf ("Problem %d lower than %d\n",estepm, stepm);
5458: }
5459: else hstepm=estepm;
5460: /* For example we decided to compute the life expectancy with the smallest unit */
5461: /* hstepm beeing the number of stepms, if hstepm=1 the length of hstepm is stepm.
5462: nhstepm is the number of hstepm from age to agelim
5463: nstepm is the number of stepm from age to agelim.
5464: Look at function hpijx to understand why because of memory size limitations,
5465: we decided (b) to get a life expectancy respecting the most precise curvature of the
5466: survival function given by stepm (the optimization length). Unfortunately it
5467: means that if the survival funtion is printed every two years of age and if
5468: you sum them up and add 1 year (area under the trapezoids) you won't get the same
5469: results. So we changed our mind and took the option of the best precision.
5470: */
5471: hstepm=hstepm/stepm; /* Typically in stepm units, if stepm=6 & estepm=24 , = 24/6 months = 4 */
5472: agelim = AGESUP;
5473: for (age=bage; age<=fage; age ++){ /* If stepm=6 months */
5474: nstepm=(int) rint((agelim-age)*YEARM/stepm); /* Typically 20 years = 20*12/6=40 */
5475: nhstepm = nstepm/hstepm;/* Expressed in hstepm, typically nhstepm=40/4=10 */
5476: p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5477: gradg=ma3x(0,nhstepm,1,npar,1,nlstate);
5478: gp=matrix(0,nhstepm,1,nlstate);
5479: gm=matrix(0,nhstepm,1,nlstate);
5480:
5481:
5482: for(theta=1; theta <=npar; theta++){
5483: for(i=1; i<=npar; i++){ /* Computes gradient x + delta*/
5484: xp[i] = x[i] + (i==theta ?delti[theta]:0);
5485: }
5486:
1.242 brouard 5487: prevalim(prlim,nlstate,xp,age,oldm,savm,ftolpl,ncvyearp,ij, nres);
1.218 brouard 5488:
5489: if (popbased==1) {
5490: if(mobilav ==0){
5491: for(i=1; i<=nlstate;i++)
5492: prlim[i][i]=probs[(int)age][i][ij];
5493: }else{ /* mobilav */
5494: for(i=1; i<=nlstate;i++)
5495: prlim[i][i]=mobaverage[(int)age][i][ij];
5496: }
5497: }
5498:
1.235 brouard 5499: 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 5500: for(j=1; j<= nlstate; j++){
5501: for(h=0; h<=nhstepm; h++){
5502: for(i=1, gp[h][j]=0.;i<=nlstate;i++)
5503: gp[h][j] += prlim[i][i]*p3mat[i][j][h];
5504: }
5505: }
5506: /* Next for computing probability of death (h=1 means
5507: computed over hstepm matrices product = hstepm*stepm months)
5508: as a weighted average of prlim.
5509: */
5510: for(j=nlstate+1;j<=nlstate+ndeath;j++){
5511: for(i=1,gpp[j]=0.; i<= nlstate; i++)
5512: gpp[j] += prlim[i][i]*p3mat[i][j][1];
5513: }
5514: /* end probability of death */
5515:
5516: for(i=1; i<=npar; i++) /* Computes gradient x - delta */
5517: xp[i] = x[i] - (i==theta ?delti[theta]:0);
5518:
1.242 brouard 5519: prevalim(prlim,nlstate,xp,age,oldm,savm,ftolpl,ncvyearp, ij, nres);
1.218 brouard 5520:
5521: if (popbased==1) {
5522: if(mobilav ==0){
5523: for(i=1; i<=nlstate;i++)
5524: prlim[i][i]=probs[(int)age][i][ij];
5525: }else{ /* mobilav */
5526: for(i=1; i<=nlstate;i++)
5527: prlim[i][i]=mobaverage[(int)age][i][ij];
5528: }
5529: }
5530:
1.235 brouard 5531: hpxij(p3mat,nhstepm,age,hstepm,xp,nlstate,stepm,oldm,savm, ij,nres);
1.218 brouard 5532:
5533: for(j=1; j<= nlstate; j++){ /* Sum of wi * eij = e.j */
5534: for(h=0; h<=nhstepm; h++){
5535: for(i=1, gm[h][j]=0.;i<=nlstate;i++)
5536: gm[h][j] += prlim[i][i]*p3mat[i][j][h];
5537: }
5538: }
5539: /* This for computing probability of death (h=1 means
5540: computed over hstepm matrices product = hstepm*stepm months)
5541: as a weighted average of prlim.
5542: */
5543: for(j=nlstate+1;j<=nlstate+ndeath;j++){
5544: for(i=1,gmp[j]=0.; i<= nlstate; i++)
5545: gmp[j] += prlim[i][i]*p3mat[i][j][1];
5546: }
5547: /* end probability of death */
5548:
5549: for(j=1; j<= nlstate; j++) /* vareij */
5550: for(h=0; h<=nhstepm; h++){
5551: gradg[h][theta][j]= (gp[h][j]-gm[h][j])/2./delti[theta];
5552: }
5553:
5554: for(j=nlstate+1; j<= nlstate+ndeath; j++){ /* var mu */
5555: gradgp[theta][j]= (gpp[j]-gmp[j])/2./delti[theta];
5556: }
5557:
5558: } /* End theta */
5559:
5560: trgradg =ma3x(0,nhstepm,1,nlstate,1,npar); /* veij */
5561:
5562: for(h=0; h<=nhstepm; h++) /* veij */
5563: for(j=1; j<=nlstate;j++)
5564: for(theta=1; theta <=npar; theta++)
5565: trgradg[h][j][theta]=gradg[h][theta][j];
5566:
5567: for(j=nlstate+1; j<=nlstate+ndeath;j++) /* mu */
5568: for(theta=1; theta <=npar; theta++)
5569: trgradgp[j][theta]=gradgp[theta][j];
5570:
5571:
5572: hf=hstepm*stepm/YEARM; /* Duration of hstepm expressed in year unit. */
5573: for(i=1;i<=nlstate;i++)
5574: for(j=1;j<=nlstate;j++)
5575: vareij[i][j][(int)age] =0.;
5576:
5577: for(h=0;h<=nhstepm;h++){
5578: for(k=0;k<=nhstepm;k++){
5579: matprod2(dnewm,trgradg[h],1,nlstate,1,npar,1,npar,matcov);
5580: matprod2(doldm,dnewm,1,nlstate,1,npar,1,nlstate,gradg[k]);
5581: for(i=1;i<=nlstate;i++)
5582: for(j=1;j<=nlstate;j++)
5583: vareij[i][j][(int)age] += doldm[i][j]*hf*hf;
5584: }
5585: }
5586:
5587: /* pptj */
5588: matprod2(dnewmp,trgradgp,nlstate+1,nlstate+ndeath,1,npar,1,npar,matcov);
5589: matprod2(doldmp,dnewmp,nlstate+1,nlstate+ndeath,1,npar,nlstate+1,nlstate+ndeath,gradgp);
5590: for(j=nlstate+1;j<=nlstate+ndeath;j++)
5591: for(i=nlstate+1;i<=nlstate+ndeath;i++)
5592: varppt[j][i]=doldmp[j][i];
5593: /* end ppptj */
5594: /* x centered again */
5595:
1.242 brouard 5596: prevalim(prlim,nlstate,x,age,oldm,savm,ftolpl,ncvyearp,ij, nres);
1.218 brouard 5597:
5598: if (popbased==1) {
5599: if(mobilav ==0){
5600: for(i=1; i<=nlstate;i++)
5601: prlim[i][i]=probs[(int)age][i][ij];
5602: }else{ /* mobilav */
5603: for(i=1; i<=nlstate;i++)
5604: prlim[i][i]=mobaverage[(int)age][i][ij];
5605: }
5606: }
5607:
5608: /* This for computing probability of death (h=1 means
5609: computed over hstepm (estepm) matrices product = hstepm*stepm months)
5610: as a weighted average of prlim.
5611: */
1.235 brouard 5612: hpxij(p3mat,nhstepm,age,hstepm,x,nlstate,stepm,oldm,savm, ij, nres);
1.218 brouard 5613: for(j=nlstate+1;j<=nlstate+ndeath;j++){
5614: for(i=1,gmp[j]=0.;i<= nlstate; i++)
5615: gmp[j] += prlim[i][i]*p3mat[i][j][1];
5616: }
5617: /* end probability of death */
5618:
5619: fprintf(ficresprobmorprev,"%3d %d ",(int) age, ij);
5620: for(j=nlstate+1; j<=(nlstate+ndeath);j++){
5621: fprintf(ficresprobmorprev," %11.3e %11.3e",gmp[j], sqrt(varppt[j][j]));
5622: for(i=1; i<=nlstate;i++){
5623: fprintf(ficresprobmorprev," %11.3e %11.3e ",prlim[i][i],p3mat[i][j][1]);
5624: }
5625: }
5626: fprintf(ficresprobmorprev,"\n");
5627:
5628: fprintf(ficresvij,"%.0f ",age );
5629: for(i=1; i<=nlstate;i++)
5630: for(j=1; j<=nlstate;j++){
5631: fprintf(ficresvij," %.4f", vareij[i][j][(int)age]);
5632: }
5633: fprintf(ficresvij,"\n");
5634: free_matrix(gp,0,nhstepm,1,nlstate);
5635: free_matrix(gm,0,nhstepm,1,nlstate);
5636: free_ma3x(gradg,0,nhstepm,1,npar,1,nlstate);
5637: free_ma3x(trgradg,0,nhstepm,1,nlstate,1,npar);
5638: free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
5639: } /* End age */
5640: free_vector(gpp,nlstate+1,nlstate+ndeath);
5641: free_vector(gmp,nlstate+1,nlstate+ndeath);
5642: free_matrix(gradgp,1,npar,nlstate+1,nlstate+ndeath);
5643: free_matrix(trgradgp,nlstate+1,nlstate+ndeath,1,npar); /* mu or p point j*/
5644: /* fprintf(ficgp,"\nunset parametric;unset label; set ter png small size 320, 240"); */
5645: fprintf(ficgp,"\nunset parametric;unset label; set ter svg size 640, 480");
5646: /* for(j=nlstate+1; j<= nlstate+ndeath; j++){ *//* Only the first actually */
5647: fprintf(ficgp,"\n set log y; unset log x;set xlabel \"Age\"; set ylabel \"Force of mortality (year-1)\";");
5648: fprintf(ficgp,"\nset out \"%s%s.svg\";",subdirf3(optionfilefiname,"VARMUPTJGR-",digitp),digit);
5649: /* fprintf(ficgp,"\n plot \"%s\" u 1:($3*%6.3f) not w l 1 ",fileresprobmorprev,YEARM/estepm); */
5650: /* fprintf(ficgp,"\n replot \"%s\" u 1:(($3+1.96*$4)*%6.3f) t \"95\%% interval\" w l 2 ",fileresprobmorprev,YEARM/estepm); */
5651: /* fprintf(ficgp,"\n replot \"%s\" u 1:(($3-1.96*$4)*%6.3f) not w l 2 ",fileresprobmorprev,YEARM/estepm); */
5652: fprintf(ficgp,"\n plot \"%s\" u 1:($3) not w l lt 1 ",subdirf(fileresprobmorprev));
5653: fprintf(ficgp,"\n replot \"%s\" u 1:(($3+1.96*$4)) t \"95%% interval\" w l lt 2 ",subdirf(fileresprobmorprev));
5654: fprintf(ficgp,"\n replot \"%s\" u 1:(($3-1.96*$4)) not w l lt 2 ",subdirf(fileresprobmorprev));
5655: fprintf(fichtm,"\n<br> File (multiple files are possible if covariates are present): <A href=\"%s\">%s</a>\n",subdirf(fileresprobmorprev),subdirf(fileresprobmorprev));
5656: 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);
5657: /* 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 5658: */
1.218 brouard 5659: /* fprintf(ficgp,"\nset out \"varmuptjgr%s%s%s.svg\";replot;",digitp,optionfilefiname,digit); */
5660: fprintf(ficgp,"\nset out;\nset out \"%s%s.svg\";replot;set out;\n",subdirf3(optionfilefiname,"VARMUPTJGR-",digitp),digit);
1.126 brouard 5661:
1.218 brouard 5662: free_vector(xp,1,npar);
5663: free_matrix(doldm,1,nlstate,1,nlstate);
5664: free_matrix(dnewm,1,nlstate,1,npar);
5665: free_matrix(doldmp,nlstate+1,nlstate+ndeath,nlstate+1,nlstate+ndeath);
5666: free_matrix(dnewmp,nlstate+1,nlstate+ndeath,1,npar);
5667: free_matrix(varppt,nlstate+1,nlstate+ndeath,nlstate+1,nlstate+ndeath);
5668: /* if (mobilav!=0) free_ma3x(mobaverage,1, AGESUP,1,NCOVMAX, 1,NCOVMAX); */
5669: fclose(ficresprobmorprev);
5670: fflush(ficgp);
5671: fflush(fichtm);
5672: } /* end varevsij */
1.126 brouard 5673:
5674: /************ Variance of prevlim ******************/
1.235 brouard 5675: 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 5676: {
1.205 brouard 5677: /* Variance of prevalence limit for each state ij using current parameters x[] and estimates of neighbourhood give by delti*/
1.126 brouard 5678: /* double **prevalim(double **prlim, int nlstate, double *xp, double age, double **oldm, double **savm,double ftolpl);*/
1.164 brouard 5679:
1.126 brouard 5680: double **dnewm,**doldm;
5681: int i, j, nhstepm, hstepm;
5682: double *xp;
5683: double *gp, *gm;
5684: double **gradg, **trgradg;
1.208 brouard 5685: double **mgm, **mgp;
1.126 brouard 5686: double age,agelim;
5687: int theta;
5688:
5689: pstamp(ficresvpl);
5690: fprintf(ficresvpl,"# Standard deviation of period (stable) prevalences \n");
1.241 brouard 5691: fprintf(ficresvpl,"# Age ");
5692: if(nresult >=1)
5693: fprintf(ficresvpl," Result# ");
1.126 brouard 5694: for(i=1; i<=nlstate;i++)
5695: fprintf(ficresvpl," %1d-%1d",i,i);
5696: fprintf(ficresvpl,"\n");
5697:
5698: xp=vector(1,npar);
5699: dnewm=matrix(1,nlstate,1,npar);
5700: doldm=matrix(1,nlstate,1,nlstate);
5701:
5702: hstepm=1*YEARM; /* Every year of age */
5703: hstepm=hstepm/stepm; /* Typically in stepm units, if j= 2 years, = 2/6 months = 4 */
5704: agelim = AGESUP;
5705: for (age=bage; age<=fage; age ++){ /* If stepm=6 months */
5706: nhstepm=(int) rint((agelim-age)*YEARM/stepm); /* Typically 20 years = 20*12/6=40 */
5707: if (stepm >= YEARM) hstepm=1;
5708: nhstepm = nhstepm/hstepm; /* Typically 40/4=10 */
5709: gradg=matrix(1,npar,1,nlstate);
1.208 brouard 5710: mgp=matrix(1,npar,1,nlstate);
5711: mgm=matrix(1,npar,1,nlstate);
1.126 brouard 5712: gp=vector(1,nlstate);
5713: gm=vector(1,nlstate);
5714:
5715: for(theta=1; theta <=npar; theta++){
5716: for(i=1; i<=npar; i++){ /* Computes gradient */
5717: xp[i] = x[i] + (i==theta ?delti[theta]:0);
5718: }
1.209 brouard 5719: if((int)age==79 ||(int)age== 80 ||(int)age== 81 )
1.235 brouard 5720: prevalim(prlim,nlstate,xp,age,oldm,savm,ftolpl,ncvyearp,ij,nres);
1.209 brouard 5721: else
1.235 brouard 5722: prevalim(prlim,nlstate,xp,age,oldm,savm,ftolpl,ncvyearp,ij,nres);
1.208 brouard 5723: for(i=1;i<=nlstate;i++){
1.126 brouard 5724: gp[i] = prlim[i][i];
1.208 brouard 5725: mgp[theta][i] = prlim[i][i];
5726: }
1.126 brouard 5727: for(i=1; i<=npar; i++) /* Computes gradient */
5728: xp[i] = x[i] - (i==theta ?delti[theta]:0);
1.209 brouard 5729: if((int)age==79 ||(int)age== 80 ||(int)age== 81 )
1.235 brouard 5730: prevalim(prlim,nlstate,xp,age,oldm,savm,ftolpl,ncvyearp,ij,nres);
1.209 brouard 5731: else
1.235 brouard 5732: prevalim(prlim,nlstate,xp,age,oldm,savm,ftolpl,ncvyearp,ij,nres);
1.208 brouard 5733: for(i=1;i<=nlstate;i++){
1.126 brouard 5734: gm[i] = prlim[i][i];
1.208 brouard 5735: mgm[theta][i] = prlim[i][i];
5736: }
1.126 brouard 5737: for(i=1;i<=nlstate;i++)
5738: gradg[theta][i]= (gp[i]-gm[i])/2./delti[theta];
1.209 brouard 5739: /* gradg[theta][2]= -gradg[theta][1]; */ /* For testing if nlstate=2 */
1.126 brouard 5740: } /* End theta */
5741:
5742: trgradg =matrix(1,nlstate,1,npar);
5743:
5744: for(j=1; j<=nlstate;j++)
5745: for(theta=1; theta <=npar; theta++)
5746: trgradg[j][theta]=gradg[theta][j];
1.209 brouard 5747: /* if((int)age==79 ||(int)age== 80 ||(int)age== 81 ){ */
5748: /* printf("\nmgm mgp %d ",(int)age); */
5749: /* for(j=1; j<=nlstate;j++){ */
5750: /* printf(" %d ",j); */
5751: /* for(theta=1; theta <=npar; theta++) */
5752: /* printf(" %d %lf %lf",theta,mgm[theta][j],mgp[theta][j]); */
5753: /* printf("\n "); */
5754: /* } */
5755: /* } */
5756: /* if((int)age==79 ||(int)age== 80 ||(int)age== 81 ){ */
5757: /* printf("\n gradg %d ",(int)age); */
5758: /* for(j=1; j<=nlstate;j++){ */
5759: /* printf("%d ",j); */
5760: /* for(theta=1; theta <=npar; theta++) */
5761: /* printf("%d %lf ",theta,gradg[theta][j]); */
5762: /* printf("\n "); */
5763: /* } */
5764: /* } */
1.126 brouard 5765:
5766: for(i=1;i<=nlstate;i++)
5767: varpl[i][(int)age] =0.;
1.209 brouard 5768: if((int)age==79 ||(int)age== 80 ||(int)age== 81){
1.205 brouard 5769: matprod2(dnewm,trgradg,1,nlstate,1,npar,1,npar,matcov);
5770: matprod2(doldm,dnewm,1,nlstate,1,npar,1,nlstate,gradg);
5771: }else{
1.126 brouard 5772: matprod2(dnewm,trgradg,1,nlstate,1,npar,1,npar,matcov);
5773: matprod2(doldm,dnewm,1,nlstate,1,npar,1,nlstate,gradg);
1.205 brouard 5774: }
1.126 brouard 5775: for(i=1;i<=nlstate;i++)
5776: varpl[i][(int)age] = doldm[i][i]; /* Covariances are useless */
5777:
5778: fprintf(ficresvpl,"%.0f ",age );
1.241 brouard 5779: if(nresult >=1)
5780: fprintf(ficresvpl,"%d ",nres );
1.126 brouard 5781: for(i=1; i<=nlstate;i++)
5782: fprintf(ficresvpl," %.5f (%.5f)",prlim[i][i],sqrt(varpl[i][(int)age]));
5783: fprintf(ficresvpl,"\n");
5784: free_vector(gp,1,nlstate);
5785: free_vector(gm,1,nlstate);
1.208 brouard 5786: free_matrix(mgm,1,npar,1,nlstate);
5787: free_matrix(mgp,1,npar,1,nlstate);
1.126 brouard 5788: free_matrix(gradg,1,npar,1,nlstate);
5789: free_matrix(trgradg,1,nlstate,1,npar);
5790: } /* End age */
5791:
5792: free_vector(xp,1,npar);
5793: free_matrix(doldm,1,nlstate,1,npar);
5794: free_matrix(dnewm,1,nlstate,1,nlstate);
5795:
5796: }
5797:
5798: /************ Variance of one-step probabilities ******************/
5799: 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 5800: {
5801: int i, j=0, k1, l1, tj;
5802: int k2, l2, j1, z1;
5803: int k=0, l;
5804: int first=1, first1, first2;
5805: double cv12, mu1, mu2, lc1, lc2, v12, v21, v11, v22,v1,v2, c12, tnalp;
5806: double **dnewm,**doldm;
5807: double *xp;
5808: double *gp, *gm;
5809: double **gradg, **trgradg;
5810: double **mu;
5811: double age, cov[NCOVMAX+1];
5812: double std=2.0; /* Number of standard deviation wide of confidence ellipsoids */
5813: int theta;
5814: char fileresprob[FILENAMELENGTH];
5815: char fileresprobcov[FILENAMELENGTH];
5816: char fileresprobcor[FILENAMELENGTH];
5817: double ***varpij;
5818:
5819: strcpy(fileresprob,"PROB_");
5820: strcat(fileresprob,fileres);
5821: if((ficresprob=fopen(fileresprob,"w"))==NULL) {
5822: printf("Problem with resultfile: %s\n", fileresprob);
5823: fprintf(ficlog,"Problem with resultfile: %s\n", fileresprob);
5824: }
5825: strcpy(fileresprobcov,"PROBCOV_");
5826: strcat(fileresprobcov,fileresu);
5827: if((ficresprobcov=fopen(fileresprobcov,"w"))==NULL) {
5828: printf("Problem with resultfile: %s\n", fileresprobcov);
5829: fprintf(ficlog,"Problem with resultfile: %s\n", fileresprobcov);
5830: }
5831: strcpy(fileresprobcor,"PROBCOR_");
5832: strcat(fileresprobcor,fileresu);
5833: if((ficresprobcor=fopen(fileresprobcor,"w"))==NULL) {
5834: printf("Problem with resultfile: %s\n", fileresprobcor);
5835: fprintf(ficlog,"Problem with resultfile: %s\n", fileresprobcor);
5836: }
5837: printf("Computing standard deviation of one-step probabilities: result on file '%s' \n",fileresprob);
5838: fprintf(ficlog,"Computing standard deviation of one-step probabilities: result on file '%s' \n",fileresprob);
5839: printf("Computing matrix of variance covariance of one-step probabilities: result on file '%s' \n",fileresprobcov);
5840: fprintf(ficlog,"Computing matrix of variance covariance of one-step probabilities: result on file '%s' \n",fileresprobcov);
5841: printf("and correlation matrix of one-step probabilities: result on file '%s' \n",fileresprobcor);
5842: fprintf(ficlog,"and correlation matrix of one-step probabilities: result on file '%s' \n",fileresprobcor);
5843: pstamp(ficresprob);
5844: fprintf(ficresprob,"#One-step probabilities and stand. devi in ()\n");
5845: fprintf(ficresprob,"# Age");
5846: pstamp(ficresprobcov);
5847: fprintf(ficresprobcov,"#One-step probabilities and covariance matrix\n");
5848: fprintf(ficresprobcov,"# Age");
5849: pstamp(ficresprobcor);
5850: fprintf(ficresprobcor,"#One-step probabilities and correlation matrix\n");
5851: fprintf(ficresprobcor,"# Age");
1.126 brouard 5852:
5853:
1.222 brouard 5854: for(i=1; i<=nlstate;i++)
5855: for(j=1; j<=(nlstate+ndeath);j++){
5856: fprintf(ficresprob," p%1d-%1d (SE)",i,j);
5857: fprintf(ficresprobcov," p%1d-%1d ",i,j);
5858: fprintf(ficresprobcor," p%1d-%1d ",i,j);
5859: }
5860: /* fprintf(ficresprob,"\n");
5861: fprintf(ficresprobcov,"\n");
5862: fprintf(ficresprobcor,"\n");
5863: */
5864: xp=vector(1,npar);
5865: dnewm=matrix(1,(nlstate)*(nlstate+ndeath),1,npar);
5866: doldm=matrix(1,(nlstate)*(nlstate+ndeath),1,(nlstate)*(nlstate+ndeath));
5867: mu=matrix(1,(nlstate)*(nlstate+ndeath), (int) bage, (int)fage);
5868: varpij=ma3x(1,nlstate*(nlstate+ndeath),1,nlstate*(nlstate+ndeath),(int) bage, (int) fage);
5869: first=1;
5870: fprintf(ficgp,"\n# Routine varprob");
5871: fprintf(fichtm,"\n<li><h4> Computing and drawing one step probabilities with their confidence intervals</h4></li>\n");
5872: fprintf(fichtm,"\n");
5873:
5874: 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);
5875: 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);
5876: fprintf(fichtmcov,"\nEllipsoids of confidence centered on point (p<inf>ij</inf>, p<inf>kl</inf>) are estimated \
1.126 brouard 5877: and drawn. It helps understanding how is the covariance between two incidences.\
5878: They are expressed in year<sup>-1</sup> in order to be less dependent of stepm.<br>\n");
1.222 brouard 5879: 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 5880: It can be understood this way: if pij and pkl where uncorrelated the (2x2) matrix of covariance \
5881: would have been (1/(var pij), 0 , 0, 1/(var pkl)), and the confidence interval would be 2 \
5882: standard deviations wide on each axis. <br>\
5883: Now, if both incidences are correlated (usual case) we diagonalised the inverse of the covariance matrix\
5884: and made the appropriate rotation to look at the uncorrelated principal directions.<br>\
5885: To be simple, these graphs help to understand the significativity of each parameter in relation to a second other one.<br> \n");
5886:
1.222 brouard 5887: cov[1]=1;
5888: /* tj=cptcoveff; */
1.225 brouard 5889: tj = (int) pow(2,cptcoveff);
1.222 brouard 5890: if (cptcovn<1) {tj=1;ncodemax[1]=1;}
5891: j1=0;
1.224 brouard 5892: for(j1=1; j1<=tj;j1++){ /* For each valid combination of covariates or only once*/
1.222 brouard 5893: if (cptcovn>0) {
5894: fprintf(ficresprob, "\n#********** Variable ");
1.225 brouard 5895: for (z1=1; z1<=cptcoveff; z1++) fprintf(ficresprob, "V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
1.222 brouard 5896: fprintf(ficresprob, "**********\n#\n");
5897: fprintf(ficresprobcov, "\n#********** Variable ");
1.225 brouard 5898: for (z1=1; z1<=cptcoveff; z1++) fprintf(ficresprobcov, "V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
1.222 brouard 5899: fprintf(ficresprobcov, "**********\n#\n");
1.220 brouard 5900:
1.222 brouard 5901: fprintf(ficgp, "\n#********** Variable ");
1.225 brouard 5902: for (z1=1; z1<=cptcoveff; z1++) fprintf(ficgp, " V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
1.222 brouard 5903: fprintf(ficgp, "**********\n#\n");
1.220 brouard 5904:
5905:
1.222 brouard 5906: fprintf(fichtmcov, "\n<hr size=\"2\" color=\"#EC5E5E\">********** Variable ");
1.225 brouard 5907: for (z1=1; z1<=cptcoveff; z1++) fprintf(fichtm, "V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
1.222 brouard 5908: fprintf(fichtmcov, "**********\n<hr size=\"2\" color=\"#EC5E5E\">");
1.220 brouard 5909:
1.222 brouard 5910: fprintf(ficresprobcor, "\n#********** Variable ");
1.225 brouard 5911: for (z1=1; z1<=cptcoveff; z1++) fprintf(ficresprobcor, "V%d=%d ",Tvaraff[z1],nbcode[Tvaraff[z1]][codtabm(j1,z1)]);
1.222 brouard 5912: fprintf(ficresprobcor, "**********\n#");
5913: if(invalidvarcomb[j1]){
5914: fprintf(ficgp,"\n#Combination (%d) ignored because no cases \n",j1);
5915: fprintf(fichtmcov,"\n<h3>Combination (%d) ignored because no cases </h3>\n",j1);
5916: continue;
5917: }
5918: }
5919: gradg=matrix(1,npar,1,(nlstate)*(nlstate+ndeath));
5920: trgradg=matrix(1,(nlstate)*(nlstate+ndeath),1,npar);
5921: gp=vector(1,(nlstate)*(nlstate+ndeath));
5922: gm=vector(1,(nlstate)*(nlstate+ndeath));
5923: for (age=bage; age<=fage; age ++){
5924: cov[2]=age;
5925: if(nagesqr==1)
5926: cov[3]= age*age;
5927: for (k=1; k<=cptcovn;k++) {
5928: cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(j1,k)];
5929: /*cov[2+nagesqr+k]=nbcode[Tvar[k]][codtabm(j1,Tvar[k])];*//* j1 1 2 3 4
5930: * 1 1 1 1 1
5931: * 2 2 1 1 1
5932: * 3 1 2 1 1
5933: */
5934: /* nbcode[1][1]=0 nbcode[1][2]=1;*/
5935: }
5936: /* for (k=1; k<=cptcovage;k++) cov[2+Tage[k]]=cov[2+Tage[k]]*cov[2]; */
5937: for (k=1; k<=cptcovage;k++) cov[2+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,k)]*cov[2];
5938: for (k=1; k<=cptcovprod;k++)
5939: cov[2+nagesqr+Tprod[k]]=nbcode[Tvard[k][1]][codtabm(ij,k)]*nbcode[Tvard[k][2]][codtabm(ij,k)];
1.220 brouard 5940:
5941:
1.222 brouard 5942: for(theta=1; theta <=npar; theta++){
5943: for(i=1; i<=npar; i++)
5944: xp[i] = x[i] + (i==theta ?delti[theta]:(double)0);
1.220 brouard 5945:
1.222 brouard 5946: pmij(pmmij,cov,ncovmodel,xp,nlstate);
1.220 brouard 5947:
1.222 brouard 5948: k=0;
5949: for(i=1; i<= (nlstate); i++){
5950: for(j=1; j<=(nlstate+ndeath);j++){
5951: k=k+1;
5952: gp[k]=pmmij[i][j];
5953: }
5954: }
1.220 brouard 5955:
1.222 brouard 5956: for(i=1; i<=npar; i++)
5957: xp[i] = x[i] - (i==theta ?delti[theta]:(double)0);
1.220 brouard 5958:
1.222 brouard 5959: pmij(pmmij,cov,ncovmodel,xp,nlstate);
5960: k=0;
5961: for(i=1; i<=(nlstate); i++){
5962: for(j=1; j<=(nlstate+ndeath);j++){
5963: k=k+1;
5964: gm[k]=pmmij[i][j];
5965: }
5966: }
1.220 brouard 5967:
1.222 brouard 5968: for(i=1; i<= (nlstate)*(nlstate+ndeath); i++)
5969: gradg[theta][i]=(gp[i]-gm[i])/(double)2./delti[theta];
5970: }
1.126 brouard 5971:
1.222 brouard 5972: for(j=1; j<=(nlstate)*(nlstate+ndeath);j++)
5973: for(theta=1; theta <=npar; theta++)
5974: trgradg[j][theta]=gradg[theta][j];
1.220 brouard 5975:
1.222 brouard 5976: matprod2(dnewm,trgradg,1,(nlstate)*(nlstate+ndeath),1,npar,1,npar,matcov);
5977: matprod2(doldm,dnewm,1,(nlstate)*(nlstate+ndeath),1,npar,1,(nlstate)*(nlstate+ndeath),gradg);
1.220 brouard 5978:
1.222 brouard 5979: pmij(pmmij,cov,ncovmodel,x,nlstate);
1.220 brouard 5980:
1.222 brouard 5981: k=0;
5982: for(i=1; i<=(nlstate); i++){
5983: for(j=1; j<=(nlstate+ndeath);j++){
5984: k=k+1;
5985: mu[k][(int) age]=pmmij[i][j];
5986: }
5987: }
5988: for(i=1;i<=(nlstate)*(nlstate+ndeath);i++)
5989: for(j=1;j<=(nlstate)*(nlstate+ndeath);j++)
5990: varpij[i][j][(int)age] = doldm[i][j];
1.220 brouard 5991:
1.222 brouard 5992: /*printf("\n%d ",(int)age);
5993: for (i=1; i<=(nlstate)*(nlstate+ndeath);i++){
5994: printf("%e [%e ;%e] ",gm[i],gm[i]-2*sqrt(doldm[i][i]),gm[i]+2*sqrt(doldm[i][i]));
5995: fprintf(ficlog,"%e [%e ;%e] ",gm[i],gm[i]-2*sqrt(doldm[i][i]),gm[i]+2*sqrt(doldm[i][i]));
5996: }*/
1.220 brouard 5997:
1.222 brouard 5998: fprintf(ficresprob,"\n%d ",(int)age);
5999: fprintf(ficresprobcov,"\n%d ",(int)age);
6000: fprintf(ficresprobcor,"\n%d ",(int)age);
1.220 brouard 6001:
1.222 brouard 6002: for (i=1; i<=(nlstate)*(nlstate+ndeath);i++)
6003: fprintf(ficresprob,"%11.3e (%11.3e) ",mu[i][(int) age],sqrt(varpij[i][i][(int)age]));
6004: for (i=1; i<=(nlstate)*(nlstate+ndeath);i++){
6005: fprintf(ficresprobcov,"%11.3e ",mu[i][(int) age]);
6006: fprintf(ficresprobcor,"%11.3e ",mu[i][(int) age]);
6007: }
6008: i=0;
6009: for (k=1; k<=(nlstate);k++){
6010: for (l=1; l<=(nlstate+ndeath);l++){
6011: i++;
6012: fprintf(ficresprobcov,"\n%d %d-%d",(int)age,k,l);
6013: fprintf(ficresprobcor,"\n%d %d-%d",(int)age,k,l);
6014: for (j=1; j<=i;j++){
6015: /* printf(" k=%d l=%d i=%d j=%d\n",k,l,i,j);fflush(stdout); */
6016: fprintf(ficresprobcov," %11.3e",varpij[i][j][(int)age]);
6017: fprintf(ficresprobcor," %11.3e",varpij[i][j][(int) age]/sqrt(varpij[i][i][(int) age])/sqrt(varpij[j][j][(int)age]));
6018: }
6019: }
6020: }/* end of loop for state */
6021: } /* end of loop for age */
6022: free_vector(gp,1,(nlstate+ndeath)*(nlstate+ndeath));
6023: free_vector(gm,1,(nlstate+ndeath)*(nlstate+ndeath));
6024: free_matrix(trgradg,1,(nlstate+ndeath)*(nlstate+ndeath),1,npar);
6025: free_matrix(gradg,1,(nlstate+ndeath)*(nlstate+ndeath),1,npar);
6026:
6027: /* Confidence intervalle of pij */
6028: /*
6029: fprintf(ficgp,"\nunset parametric;unset label");
6030: fprintf(ficgp,"\nset log y;unset log x; set xlabel \"Age\";set ylabel \"probability (year-1)\"");
6031: fprintf(ficgp,"\nset ter png small\nset size 0.65,0.65");
6032: 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);
6033: fprintf(fichtm,"\n<br><img src=\"pijgr%s.png\"> ",optionfilefiname);
6034: fprintf(ficgp,"\nset out \"pijgr%s.png\"",optionfilefiname);
6035: fprintf(ficgp,"\nplot \"%s\" every :::%d::%d u 1:2 \"\%%lf",k1,k2,xfilevarprob);
6036: */
6037:
6038: /* Drawing ellipsoids of confidence of two variables p(k1-l1,k2-l2)*/
6039: first1=1;first2=2;
6040: for (k2=1; k2<=(nlstate);k2++){
6041: for (l2=1; l2<=(nlstate+ndeath);l2++){
6042: if(l2==k2) continue;
6043: j=(k2-1)*(nlstate+ndeath)+l2;
6044: for (k1=1; k1<=(nlstate);k1++){
6045: for (l1=1; l1<=(nlstate+ndeath);l1++){
6046: if(l1==k1) continue;
6047: i=(k1-1)*(nlstate+ndeath)+l1;
6048: if(i<=j) continue;
6049: for (age=bage; age<=fage; age ++){
6050: if ((int)age %5==0){
6051: v1=varpij[i][i][(int)age]/stepm*YEARM/stepm*YEARM;
6052: v2=varpij[j][j][(int)age]/stepm*YEARM/stepm*YEARM;
6053: cv12=varpij[i][j][(int)age]/stepm*YEARM/stepm*YEARM;
6054: mu1=mu[i][(int) age]/stepm*YEARM ;
6055: mu2=mu[j][(int) age]/stepm*YEARM;
6056: c12=cv12/sqrt(v1*v2);
6057: /* Computing eigen value of matrix of covariance */
6058: lc1=((v1+v2)+sqrt((v1+v2)*(v1+v2) - 4*(v1*v2-cv12*cv12)))/2.;
6059: lc2=((v1+v2)-sqrt((v1+v2)*(v1+v2) - 4*(v1*v2-cv12*cv12)))/2.;
6060: if ((lc2 <0) || (lc1 <0) ){
6061: if(first2==1){
6062: first1=0;
6063: 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);
6064: }
6065: 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);
6066: /* lc1=fabs(lc1); */ /* If we want to have them positive */
6067: /* lc2=fabs(lc2); */
6068: }
1.220 brouard 6069:
1.222 brouard 6070: /* Eigen vectors */
6071: v11=(1./sqrt(1+(v1-lc1)*(v1-lc1)/cv12/cv12));
6072: /*v21=sqrt(1.-v11*v11); *//* error */
6073: v21=(lc1-v1)/cv12*v11;
6074: v12=-v21;
6075: v22=v11;
6076: tnalp=v21/v11;
6077: if(first1==1){
6078: first1=0;
6079: 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);
6080: }
6081: 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);
6082: /*printf(fignu*/
6083: /* mu1+ v11*lc1*cost + v12*lc2*sin(t) */
6084: /* mu2+ v21*lc1*cost + v22*lc2*sin(t) */
6085: if(first==1){
6086: first=0;
6087: fprintf(ficgp,"\n# Ellipsoids of confidence\n#\n");
6088: fprintf(ficgp,"\nset parametric;unset label");
6089: 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);
6090: fprintf(ficgp,"\nset ter svg size 640, 480");
6091: fprintf(fichtmcov,"\n<br>Ellipsoids of confidence cov(p%1d%1d,p%1d%1d) expressed in year<sup>-1</sup>\
1.220 brouard 6092: :<a href=\"%s_%d%1d%1d-%1d%1d.svg\"> \
1.201 brouard 6093: %s_%d%1d%1d-%1d%1d.svg</A>, ",k1,l1,k2,l2,\
1.222 brouard 6094: subdirf2(optionfilefiname,"VARPIJGR_"), j1,k1,l1,k2,l2, \
6095: subdirf2(optionfilefiname,"VARPIJGR_"), j1,k1,l1,k2,l2);
6096: fprintf(fichtmcov,"\n<br><img src=\"%s_%d%1d%1d-%1d%1d.svg\"> ",subdirf2(optionfilefiname,"VARPIJGR_"), j1,k1,l1,k2,l2);
6097: fprintf(fichtmcov,"\n<br> Correlation at age %d (%.3f),",(int) age, c12);
6098: fprintf(ficgp,"\nset out \"%s_%d%1d%1d-%1d%1d.svg\"",subdirf2(optionfilefiname,"VARPIJGR_"), j1,k1,l1,k2,l2);
6099: fprintf(ficgp,"\nset label \"%d\" at %11.3e,%11.3e center",(int) age, mu1,mu2);
6100: fprintf(ficgp,"\n# Age %d, p%1d%1d - p%1d%1d",(int) age, k1,l1,k2,l2);
6101: 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", \
6102: mu1,std,v11,sqrt(lc1),v12,sqrt(lc2), \
6103: mu2,std,v21,sqrt(lc1),v22,sqrt(lc2));
6104: }else{
6105: first=0;
6106: fprintf(fichtmcov," %d (%.3f),",(int) age, c12);
6107: fprintf(ficgp,"\n# Age %d, p%1d%1d - p%1d%1d",(int) age, k1,l1,k2,l2);
6108: fprintf(ficgp,"\nset label \"%d\" at %11.3e,%11.3e center",(int) age, mu1,mu2);
6109: 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", \
6110: mu1,std,v11,sqrt(lc1),v12,sqrt(lc2), \
6111: mu2,std,v21,sqrt(lc1),v22,sqrt(lc2));
6112: }/* if first */
6113: } /* age mod 5 */
6114: } /* end loop age */
6115: fprintf(ficgp,"\nset out;\nset out \"%s_%d%1d%1d-%1d%1d.svg\";replot;set out;",subdirf2(optionfilefiname,"VARPIJGR_"), j1,k1,l1,k2,l2);
6116: first=1;
6117: } /*l12 */
6118: } /* k12 */
6119: } /*l1 */
6120: }/* k1 */
6121: } /* loop on combination of covariates j1 */
6122: free_ma3x(varpij,1,nlstate,1,nlstate+ndeath,(int) bage, (int)fage);
6123: free_matrix(mu,1,(nlstate+ndeath)*(nlstate+ndeath),(int) bage, (int)fage);
6124: free_matrix(doldm,1,(nlstate)*(nlstate+ndeath),1,(nlstate)*(nlstate+ndeath));
6125: free_matrix(dnewm,1,(nlstate)*(nlstate+ndeath),1,npar);
6126: free_vector(xp,1,npar);
6127: fclose(ficresprob);
6128: fclose(ficresprobcov);
6129: fclose(ficresprobcor);
6130: fflush(ficgp);
6131: fflush(fichtmcov);
6132: }
1.126 brouard 6133:
6134:
6135: /******************* Printing html file ***********/
1.201 brouard 6136: void printinghtml(char fileresu[], char title[], char datafile[], int firstpass, \
1.126 brouard 6137: int lastpass, int stepm, int weightopt, char model[],\
6138: int imx,int jmin, int jmax, double jmeanint,char rfileres[],\
1.217 brouard 6139: int popforecast, int prevfcast, int backcast, int estepm , \
1.213 brouard 6140: double jprev1, double mprev1,double anprev1, double dateprev1, \
6141: double jprev2, double mprev2,double anprev2, double dateprev2){
1.237 brouard 6142: int jj1, k1, i1, cpt, k4, nres;
1.126 brouard 6143:
6144: fprintf(fichtm,"<ul><li><a href='#firstorder'>Result files (first order: no variance)</a>\n \
6145: <li><a href='#secondorder'>Result files (second order (variance)</a>\n \
6146: </ul>");
1.237 brouard 6147: fprintf(fichtm,"<ul><li> model=1+age+%s\n \
6148: </ul>", model);
1.214 brouard 6149: fprintf(fichtm,"<ul><li><h4><a name='firstorder'>Result files (first order: no variance)</a></h4>\n");
6150: 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",
6151: jprev1, mprev1,anprev1,jprev2, mprev2,anprev2,subdirfext3(optionfilefiname,"PHTMFR_",".htm"),subdirfext3(optionfilefiname,"PHTMFR_",".htm"));
6152: 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 6153: jprev1, mprev1,anprev1,jprev2, mprev2,anprev2,subdirfext3(optionfilefiname,"PHTM_",".htm"),subdirfext3(optionfilefiname,"PHTM_",".htm"));
6154: fprintf(fichtm,", <a href=\"%s\">%s</a> (text file) <br>\n",subdirf2(fileresu,"P_"),subdirf2(fileresu,"P_"));
1.126 brouard 6155: fprintf(fichtm,"\
6156: - Estimated transition probabilities over %d (stepm) months: <a href=\"%s\">%s</a><br>\n ",
1.201 brouard 6157: stepm,subdirf2(fileresu,"PIJ_"),subdirf2(fileresu,"PIJ_"));
1.126 brouard 6158: fprintf(fichtm,"\
1.217 brouard 6159: - Estimated back transition probabilities over %d (stepm) months: <a href=\"%s\">%s</a><br>\n ",
6160: stepm,subdirf2(fileresu,"PIJB_"),subdirf2(fileresu,"PIJB_"));
6161: fprintf(fichtm,"\
1.126 brouard 6162: - Period (stable) prevalence in each health state: <a href=\"%s\">%s</a> <br>\n",
1.201 brouard 6163: subdirf2(fileresu,"PL_"),subdirf2(fileresu,"PL_"));
1.126 brouard 6164: fprintf(fichtm,"\
1.217 brouard 6165: - Period (stable) back prevalence in each health state: <a href=\"%s\">%s</a> <br>\n",
6166: subdirf2(fileresu,"PLB_"),subdirf2(fileresu,"PLB_"));
6167: fprintf(fichtm,"\
1.211 brouard 6168: - (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 6169: <a href=\"%s\">%s</a> <br>\n",
1.201 brouard 6170: estepm,subdirf2(fileresu,"E_"),subdirf2(fileresu,"E_"));
1.211 brouard 6171: if(prevfcast==1){
6172: fprintf(fichtm,"\
6173: - Prevalence projections by age and states: \
1.201 brouard 6174: <a href=\"%s\">%s</a> <br>\n</li>", subdirf2(fileresu,"F_"),subdirf2(fileresu,"F_"));
1.211 brouard 6175: }
1.126 brouard 6176:
1.222 brouard 6177: fprintf(fichtm," \n<ul><li><b>Graphs</b></li><p>");
1.126 brouard 6178:
1.225 brouard 6179: m=pow(2,cptcoveff);
1.222 brouard 6180: if (cptcovn < 1) {m=1;ncodemax[1]=1;}
1.126 brouard 6181:
1.222 brouard 6182: jj1=0;
1.237 brouard 6183:
6184: for(nres=1; nres <= nresult; nres++) /* For each resultline */
1.241 brouard 6185: for(k1=1; k1<=m;k1++){ /* For each combination of covariate */
1.237 brouard 6186: if(TKresult[nres]!= k1)
6187: continue;
1.220 brouard 6188:
1.222 brouard 6189: /* for(i1=1; i1<=ncodemax[k1];i1++){ */
6190: jj1++;
6191: if (cptcovn > 0) {
6192: fprintf(fichtm,"<hr size=\"2\" color=\"#EC5E5E\">************ Results for covariates");
1.225 brouard 6193: for (cpt=1; cpt<=cptcoveff;cpt++){
1.237 brouard 6194: fprintf(fichtm," V%d=%d ",Tvresult[nres][cpt],(int)Tresult[nres][cpt]);
6195: printf(" V%d=%d ",Tvresult[nres][cpt],Tresult[nres][cpt]);fflush(stdout);
6196: /* fprintf(fichtm," V%d=%d ",Tvaraff[cpt],nbcode[Tvaraff[cpt]][codtabm(jj1,cpt)]); */
6197: /* printf(" V%d=%d ",Tvaraff[cpt],nbcode[Tvaraff[cpt]][codtabm(jj1,cpt)]);fflush(stdout); */
1.222 brouard 6198: }
1.237 brouard 6199: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
6200: fprintf(fichtm," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
6201: printf(" V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);fflush(stdout);
6202: }
6203:
1.230 brouard 6204: /* if(nqfveff+nqtveff 0) */ /* Test to be done */
1.222 brouard 6205: fprintf(fichtm," ************\n<hr size=\"2\" color=\"#EC5E5E\">");
6206: if(invalidvarcomb[k1]){
6207: fprintf(fichtm,"\n<h3>Combination (%d) ignored because no cases </h3>\n",k1);
6208: printf("\nCombination (%d) ignored because no cases \n",k1);
6209: continue;
6210: }
6211: }
6212: /* aij, bij */
1.241 brouard 6213: fprintf(fichtm,"<br>- Logit model (yours is: 1+age+%s), for example: logit(pij)=log(pij/pii)= aij+ bij age + V1 age + etc. as a function of age: <a href=\"%s_%d-1-%d.svg\">%s_%d-1-%d.svg</a><br> \
6214: <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 6215: /* Pij */
1.241 brouard 6216: 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> \
6217: <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 6218: /* Quasi-incidences */
6219: 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 6220: before but expressed in per year i.e. quasi incidences if stepm is small and probabilities too, \
1.211 brouard 6221: 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 6222: 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> \
6223: <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 6224: /* Survival functions (period) in state j */
6225: for(cpt=1; cpt<=nlstate;cpt++){
1.241 brouard 6226: 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> \
6227: <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 6228: }
6229: /* State specific survival functions (period) */
6230: for(cpt=1; cpt<=nlstate;cpt++){
6231: fprintf(fichtm,"<br>\n- Survival functions from state %d in each live state and total.\
1.220 brouard 6232: Or probability to survive in various states (1 to %d) being in state %d at different ages. \
1.241 brouard 6233: <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 6234: }
6235: /* Period (stable) prevalence in each health state */
6236: for(cpt=1; cpt<=nlstate;cpt++){
1.241 brouard 6237: fprintf(fichtm,"<br>\n- Convergence 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> \
6238: <img src=\"%s_%d-%d-%d.svg\">", cpt, cpt, nlstate, subdirf2(optionfilefiname,"P_"),cpt,k1,nres,subdirf2(optionfilefiname,"P_"),cpt,k1,nres,subdirf2(optionfilefiname,"P_"),cpt,k1,nres);
1.222 brouard 6239: }
6240: if(backcast==1){
6241: /* Period (stable) back prevalence in each health state */
6242: for(cpt=1; cpt<=nlstate;cpt++){
1.241 brouard 6243: fprintf(fichtm,"<br>\n- Convergence to period (stable) back 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> \
6244: <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 6245: }
1.217 brouard 6246: }
1.222 brouard 6247: if(prevfcast==1){
6248: /* Projection of prevalence up to period (stable) prevalence in each health state */
6249: for(cpt=1; cpt<=nlstate;cpt++){
1.241 brouard 6250: fprintf(fichtm,"<br>\n- Projection of cross-sectional prevalence (estimated with cases observed from %.1f to %.1f) 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> \
6251: <img src=\"%s_%d-%d-%d.svg\">", dateprev1, dateprev2, cpt, cpt, nlstate, subdirf2(optionfilefiname,"PROJ_"),cpt,k1,nres,subdirf2(optionfilefiname,"PROJ_"),cpt,k1,nres,subdirf2(optionfilefiname,"PROJ_"),cpt,k1,nres);
1.222 brouard 6252: }
6253: }
1.220 brouard 6254:
1.222 brouard 6255: for(cpt=1; cpt<=nlstate;cpt++) {
1.241 brouard 6256: 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> \
6257: <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 6258: }
6259: /* } /\* end i1 *\/ */
6260: }/* End k1 */
6261: fprintf(fichtm,"</ul>");
1.126 brouard 6262:
1.222 brouard 6263: fprintf(fichtm,"\
1.126 brouard 6264: \n<br><li><h4> <a name='secondorder'>Result files (second order: variances)</a></h4>\n\
1.193 brouard 6265: - Parameter file with estimated parameters and covariance matrix: <a href=\"%s\">%s</a> <br> \
1.203 brouard 6266: - 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 6267: But because parameters are usually highly correlated (a higher incidence of disability \
6268: and a higher incidence of recovery can give very close observed transition) it might \
6269: be very useful to look not only at linear confidence intervals estimated from the \
6270: variances but at the covariance matrix. And instead of looking at the estimated coefficients \
6271: (parameters) of the logistic regression, it might be more meaningful to visualize the \
6272: covariance matrix of the one-step probabilities. \
6273: See page 'Matrix of variance-covariance of one-step probabilities' below. \n", rfileres,rfileres);
1.126 brouard 6274:
1.222 brouard 6275: fprintf(fichtm," - Standard deviation of one-step probabilities: <a href=\"%s\">%s</a> <br>\n",
6276: subdirf2(fileresu,"PROB_"),subdirf2(fileresu,"PROB_"));
6277: fprintf(fichtm,"\
1.126 brouard 6278: - Variance-covariance of one-step probabilities: <a href=\"%s\">%s</a> <br>\n",
1.222 brouard 6279: subdirf2(fileresu,"PROBCOV_"),subdirf2(fileresu,"PROBCOV_"));
1.126 brouard 6280:
1.222 brouard 6281: fprintf(fichtm,"\
1.126 brouard 6282: - Correlation matrix of one-step probabilities: <a href=\"%s\">%s</a> <br>\n",
1.222 brouard 6283: subdirf2(fileresu,"PROBCOR_"),subdirf2(fileresu,"PROBCOR_"));
6284: fprintf(fichtm,"\
1.126 brouard 6285: - 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): \
6286: <a href=\"%s\">%s</a> <br>\n</li>",
1.201 brouard 6287: estepm,subdirf2(fileresu,"CVE_"),subdirf2(fileresu,"CVE_"));
1.222 brouard 6288: fprintf(fichtm,"\
1.126 brouard 6289: - (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): \
6290: <a href=\"%s\">%s</a> <br>\n</li>",
1.201 brouard 6291: estepm,subdirf2(fileresu,"STDE_"),subdirf2(fileresu,"STDE_"));
1.222 brouard 6292: fprintf(fichtm,"\
1.128 brouard 6293: - 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 6294: estepm, subdirf2(fileresu,"V_"),subdirf2(fileresu,"V_"));
6295: fprintf(fichtm,"\
1.128 brouard 6296: - 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 6297: estepm, subdirf2(fileresu,"T_"),subdirf2(fileresu,"T_"));
6298: fprintf(fichtm,"\
1.126 brouard 6299: - Standard deviation of period (stable) prevalences: <a href=\"%s\">%s</a> <br>\n",\
1.222 brouard 6300: subdirf2(fileresu,"VPL_"),subdirf2(fileresu,"VPL_"));
1.126 brouard 6301:
6302: /* if(popforecast==1) fprintf(fichtm,"\n */
6303: /* - Prevalences forecasting: <a href=\"f%s\">f%s</a> <br>\n */
6304: /* - Population forecasting (if popforecast=1): <a href=\"pop%s\">pop%s</a> <br>\n */
6305: /* <br>",fileres,fileres,fileres,fileres); */
6306: /* else */
6307: /* 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 6308: fflush(fichtm);
6309: fprintf(fichtm," <ul><li><b>Graphs</b></li><p>");
1.126 brouard 6310:
1.225 brouard 6311: m=pow(2,cptcoveff);
1.222 brouard 6312: if (cptcovn < 1) {m=1;ncodemax[1]=1;}
1.126 brouard 6313:
1.222 brouard 6314: jj1=0;
1.237 brouard 6315:
1.241 brouard 6316: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
1.222 brouard 6317: for(k1=1; k1<=m;k1++){
1.237 brouard 6318: if(TKresult[nres]!= k1)
6319: continue;
1.222 brouard 6320: /* for(i1=1; i1<=ncodemax[k1];i1++){ */
6321: jj1++;
1.126 brouard 6322: if (cptcovn > 0) {
6323: fprintf(fichtm,"<hr size=\"2\" color=\"#EC5E5E\">************ Results for covariates");
1.225 brouard 6324: for (cpt=1; cpt<=cptcoveff;cpt++) /**< cptcoveff number of variables */
1.237 brouard 6325: fprintf(fichtm," V%d=%d ",Tvresult[nres][cpt],Tresult[nres][cpt]);
6326: /* fprintf(fichtm," V%d=%d ",Tvaraff[cpt],nbcode[Tvaraff[cpt]][codtabm(jj1,cpt)]); */
6327: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
6328: fprintf(fichtm," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
6329: }
6330:
1.126 brouard 6331: fprintf(fichtm," ************\n<hr size=\"2\" color=\"#EC5E5E\">");
1.220 brouard 6332:
1.222 brouard 6333: if(invalidvarcomb[k1]){
6334: fprintf(fichtm,"\n<h4>Combination (%d) ignored because no cases </h4>\n",k1);
6335: continue;
6336: }
1.126 brouard 6337: }
6338: for(cpt=1; cpt<=nlstate;cpt++) {
1.218 brouard 6339: fprintf(fichtm,"\n<br>- Observed (cross-sectional) and period (incidence based) \
1.241 brouard 6340: prevalence (with 95%% confidence interval) in state (%d): <a href=\"%s_%d-%d-%d.svg\"> %s_%d-%d-%d.svg</a>\n <br>\
6341: <img src=\"%s_%d-%d-%d.svg\">",cpt,subdirf2(optionfilefiname,"V_"),cpt,k1,nres,subdirf2(optionfilefiname,"V_"),cpt,k1,nres,subdirf2(optionfilefiname,"V_"),cpt,k1,nres);
1.126 brouard 6342: }
6343: fprintf(fichtm,"\n<br>- Total life expectancy by age and \
1.128 brouard 6344: health expectancies in states (1) and (2). If popbased=1 the smooth (due to the model) \
6345: true period expectancies (those weighted with period prevalences are also\
6346: drawn in addition to the population based expectancies computed using\
1.241 brouard 6347: observed and cahotic prevalences: <a href=\"%s_%d-%d.svg\">%s_%d-%d.svg</a>\n<br>\
6348: <img src=\"%s_%d-%d.svg\">",subdirf2(optionfilefiname,"E_"),k1,nres,subdirf2(optionfilefiname,"E_"),k1,nres,subdirf2(optionfilefiname,"E_"),k1,nres);
1.222 brouard 6349: /* } /\* end i1 *\/ */
6350: }/* End k1 */
1.241 brouard 6351: }/* End nres */
1.222 brouard 6352: fprintf(fichtm,"</ul>");
6353: fflush(fichtm);
1.126 brouard 6354: }
6355:
6356: /******************* Gnuplot file **************/
1.223 brouard 6357: void printinggnuplot(char fileresu[], char optionfilefiname[], double ageminpar, double agemaxpar, double fage , int prevfcast, int backcast, char pathc[], double p[]){
1.126 brouard 6358:
6359: char dirfileres[132],optfileres[132];
1.223 brouard 6360: char gplotcondition[132];
1.237 brouard 6361: 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 6362: int lv=0, vlv=0, kl=0;
1.130 brouard 6363: int ng=0;
1.201 brouard 6364: int vpopbased;
1.223 brouard 6365: int ioffset; /* variable offset for columns */
1.235 brouard 6366: int nres=0; /* Index of resultline */
1.219 brouard 6367:
1.126 brouard 6368: /* if((ficgp=fopen(optionfilegnuplot,"a"))==NULL) { */
6369: /* printf("Problem with file %s",optionfilegnuplot); */
6370: /* fprintf(ficlog,"Problem with file %s",optionfilegnuplot); */
6371: /* } */
6372:
6373: /*#ifdef windows */
6374: fprintf(ficgp,"cd \"%s\" \n",pathc);
1.223 brouard 6375: /*#endif */
1.225 brouard 6376: m=pow(2,cptcoveff);
1.126 brouard 6377:
1.202 brouard 6378: /* Contribution to likelihood */
6379: /* Plot the probability implied in the likelihood */
1.223 brouard 6380: fprintf(ficgp,"\n# Contributions to the Likelihood, mle >=1. For mle=4 no interpolation, pure matrix products.\n#\n");
6381: fprintf(ficgp,"\n set log y; unset log x;set xlabel \"Age\"; set ylabel \"Likelihood (-2Log(L))\";");
6382: /* fprintf(ficgp,"\nset ter svg size 640, 480"); */ /* Too big for svg */
6383: fprintf(ficgp,"\nset ter pngcairo size 640, 480");
1.204 brouard 6384: /* nice for mle=4 plot by number of matrix products.
1.202 brouard 6385: replot "rrtest1/toto.txt" u 2:($4 == 1 && $5==2 ? $9 : 1/0):5 t "p12" with point lc 1 */
6386: /* replot exp(p1+p2*x)/(1+exp(p1+p2*x)+exp(p3+p4*x)+exp(p5+p6*x)) t "p12(x)" */
1.223 brouard 6387: /* fprintf(ficgp,"\nset out \"%s.svg\";",subdirf2(optionfilefiname,"ILK_")); */
6388: fprintf(ficgp,"\nset out \"%s-dest.png\";",subdirf2(optionfilefiname,"ILK_"));
6389: 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));
6390: fprintf(ficgp,"\nset out \"%s-ori.png\";",subdirf2(optionfilefiname,"ILK_"));
6391: 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));
6392: for (i=1; i<= nlstate ; i ++) {
6393: fprintf(ficgp,"\nset out \"%s-p%dj.png\";set ylabel \"Probability for each individual/wave\";",subdirf2(optionfilefiname,"ILK_"),i);
6394: fprintf(ficgp,"unset log;\n# plot weighted, mean weight should have point size of 0.5\n plot \"%s\"",subdirf(fileresilk));
6395: 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);
6396: for (j=2; j<= nlstate+ndeath ; j ++) {
6397: 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);
6398: }
6399: fprintf(ficgp,";\nset out; unset ylabel;\n");
6400: }
6401: /* 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 */
6402: /* fprintf(ficgp,"\nset log y;plot \"%s\" u 2:(-$11):3 t \"All sample, all transitions\" with dots lc variable",subdirf(fileresilk)); */
6403: /* fprintf(ficgp,"\nreplot \"%s\" u 2:($3 <= 3 ? -$11 : 1/0):3 t \"First 3 individuals\" with line lc variable", subdirf(fileresilk)); */
6404: fprintf(ficgp,"\nset out;unset log\n");
6405: /* fprintf(ficgp,"\nset out \"%s.svg\"; replot; set out; # bug gnuplot",subdirf2(optionfilefiname,"ILK_")); */
1.202 brouard 6406:
1.126 brouard 6407: strcpy(dirfileres,optionfilefiname);
6408: strcpy(optfileres,"vpl");
1.223 brouard 6409: /* 1eme*/
1.238 brouard 6410: for (cpt=1; cpt<= nlstate ; cpt ++){ /* For each live state */
6411: for (k1=1; k1<= m ; k1 ++){ /* For each valid combination of covariate */
1.236 brouard 6412: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
1.238 brouard 6413: /* plot [100000000000000000000:-100000000000000000000] "mysbiaspar/vplrmysbiaspar.txt to check */
6414: if(TKresult[nres]!= k1)
6415: continue;
6416: /* We are interested in selected combination by the resultline */
1.246 brouard 6417: /* printf("\n# 1st: Period (stable) prevalence with CI: 'VPL_' files and live state =%d ", cpt); */
1.238 brouard 6418: fprintf(ficgp,"\n# 1st: Period (stable) prevalence with CI: 'VPL_' files and live state =%d ", cpt);
6419: for (k=1; k<=cptcoveff; k++){ /* For each covariate k get corresponding value lv for combination k1 */
6420: lv= decodtabm(k1,k,cptcoveff); /* Should be the value of the covariate corresponding to k1 combination */
6421: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6422: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6423: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6424: vlv= nbcode[Tvaraff[k]][lv]; /* vlv is the value of the covariate lv, 0 or 1 */
6425: /* For each combination of covariate k1 (V1=1, V3=0), we printed the current covariate k and its value vlv */
1.246 brouard 6426: /* printf(" V%d=%d ",Tvaraff[k],vlv); */
1.238 brouard 6427: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
6428: }
6429: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
1.246 brouard 6430: /* printf(" V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]); */
1.238 brouard 6431: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
6432: }
1.246 brouard 6433: /* printf("\n#\n"); */
1.238 brouard 6434: fprintf(ficgp,"\n#\n");
6435: if(invalidvarcomb[k1]){
6436: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6437: continue;
6438: }
1.235 brouard 6439:
1.241 brouard 6440: fprintf(ficgp,"\nset out \"%s_%d-%d-%d.svg\" \n",subdirf2(optionfilefiname,"V_"),cpt,k1,nres);
6441: fprintf(ficgp,"\n#set out \"V_%s_%d-%d-%d.svg\" \n",optionfilefiname,cpt,k1,nres);
6442: 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);
1.235 brouard 6443:
1.238 brouard 6444: for (i=1; i<= nlstate ; i ++) {
6445: if (i==cpt) fprintf(ficgp," %%lf (%%lf)");
6446: else fprintf(ficgp," %%*lf (%%*lf)");
6447: }
1.242 brouard 6448: 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_"),k1-1,k1-1,nres);
1.238 brouard 6449: for (i=1; i<= nlstate ; i ++) {
6450: if (i==cpt) fprintf(ficgp," %%lf (%%lf)");
6451: else fprintf(ficgp," %%*lf (%%*lf)");
6452: }
1.242 brouard 6453: fprintf(ficgp,"\" t\"95%% CI\" w l lt 1,\"%s\" every :::%d::%d u 1:($2==%d ? $3-1.96*$4 : 1/0) \"%%lf %%lf",subdirf2(fileresu,"VPL_"),k1-1,k1-1,nres);
1.238 brouard 6454: for (i=1; i<= nlstate ; i ++) {
6455: if (i==cpt) fprintf(ficgp," %%lf (%%lf)");
6456: else fprintf(ficgp," %%*lf (%%*lf)");
6457: }
6458: fprintf(ficgp,"\" t\"\" w l lt 1,\"%s\" every :::%d::%d u 1:($%d) t\"Observed prevalence\" w l lt 2",subdirf2(fileresu,"P_"),k1-1,k1-1,2+4*(cpt-1));
6459: if(backcast==1){ /* We need to get the corresponding values of the covariates involved in this combination k1 */
6460: /* fprintf(ficgp,",\"%s\" every :::%d::%d u 1:($%d) t\"Backward stable prevalence\" w l lt 3",subdirf2(fileresu,"PLB_"),k1-1,k1-1,1+cpt); */
1.242 brouard 6461: fprintf(ficgp,",\"%s\" u 1:((",subdirf2(fileresu,"PLB_")); /* Age is in 1, nres in 2 to be fixed */
1.238 brouard 6462: if(cptcoveff ==0){
1.245 brouard 6463: fprintf(ficgp,"$%d)) t 'Backward prevalence in state %d' with line lt 3", 2+(cpt-1), cpt );
1.238 brouard 6464: }else{
6465: kl=0;
6466: for (k=1; k<=cptcoveff; k++){ /* For each combination of covariate */
6467: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate value corresponding to k1 combination and kth covariate */
6468: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6469: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6470: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6471: vlv= nbcode[Tvaraff[k]][lv];
1.223 brouard 6472: kl++;
1.238 brouard 6473: /* 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 *\/ */
6474: /*6+(cpt-1)*(nlstate+1)+1+(i-1)+(nlstate+1)*nlstate; 6+(1-1)*(2+1)+1+(1-1) +(2+1)*2=13 */
6475: /*6+1+(i-1)+(nlstate+1)*nlstate; 6+1+(1-1) +(2+1)*2=13 */
6476: /* '' u 6:(($1==1 && $2==0 && $3==2 && $4==0)? $9/(1.-$15) : 1/0):($5==2000? 3:2) t 'p.1' with line lc variable*/
6477: if(k==cptcoveff){
1.245 brouard 6478: fprintf(ficgp,"$%d==%d && $%d==%d)? $%d : 1/0) t 'Backward prevalence in state %d' w l lt 3",kl+1, Tvaraff[k],kl+1+1,nbcode[Tvaraff[k]][lv], \
1.242 brouard 6479: 2+cptcoveff*2+(cpt-1), cpt ); /* 4 or 6 ?*/
1.238 brouard 6480: }else{
6481: fprintf(ficgp,"$%d==%d && $%d==%d && ",kl+1, Tvaraff[k],kl+1+1,nbcode[Tvaraff[k]][lv]);
6482: kl++;
6483: }
6484: } /* end covariate */
6485: } /* end if no covariate */
6486: } /* end if backcast */
6487: fprintf(ficgp,"\nset out \n");
6488: } /* nres */
1.201 brouard 6489: } /* k1 */
6490: } /* cpt */
1.235 brouard 6491:
6492:
1.126 brouard 6493: /*2 eme*/
1.238 brouard 6494: for (k1=1; k1<= m ; k1 ++){
6495: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
6496: if(TKresult[nres]!= k1)
6497: continue;
6498: fprintf(ficgp,"\n# 2nd: Total life expectancy with CI: 't' files ");
6499: for (k=1; k<=cptcoveff; k++){ /* For each covariate and each value */
1.225 brouard 6500: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate number corresponding to k1 combination */
1.223 brouard 6501: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6502: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6503: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6504: vlv= nbcode[Tvaraff[k]][lv];
6505: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
1.211 brouard 6506: }
1.237 brouard 6507: /* for(k=1; k <= ncovds; k++){ */
1.236 brouard 6508: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
1.238 brouard 6509: printf(" V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
1.236 brouard 6510: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
1.238 brouard 6511: }
1.211 brouard 6512: fprintf(ficgp,"\n#\n");
1.223 brouard 6513: if(invalidvarcomb[k1]){
6514: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6515: continue;
6516: }
1.219 brouard 6517:
1.241 brouard 6518: fprintf(ficgp,"\nset out \"%s_%d-%d.svg\" \n",subdirf2(optionfilefiname,"E_"),k1,nres);
1.238 brouard 6519: for(vpopbased=0; vpopbased <= popbased; vpopbased++){ /* Done for vpopbased=0 and vpopbased=1 if popbased==1*/
6520: if(vpopbased==0)
6521: fprintf(ficgp,"set ylabel \"Years\" \nset ter svg size 640, 480\nplot [%.f:%.f] ",ageminpar,fage);
6522: else
6523: fprintf(ficgp,"\nreplot ");
6524: for (i=1; i<= nlstate+1 ; i ++) {
6525: k=2*i;
6526: fprintf(ficgp,"\"%s\" every :::%d::%d u 1:($2==%d && $4!=0 ?$4 : 1/0) \"%%lf %%lf %%lf",subdirf2(fileresu,"T_"),k1-1,k1-1, vpopbased);
6527: for (j=1; j<= nlstate+1 ; j ++) {
6528: if (j==i) fprintf(ficgp," %%lf (%%lf)");
6529: else fprintf(ficgp," %%*lf (%%*lf)");
6530: }
6531: if (i== 1) fprintf(ficgp,"\" t\"TLE\" w l lt %d, \\\n",i);
6532: else fprintf(ficgp,"\" t\"LE in state (%d)\" w l lt %d, \\\n",i-1,i+1);
6533: fprintf(ficgp,"\"%s\" every :::%d::%d u 1:($2==%d && $4!=0 ? $4-$5*2 : 1/0) \"%%lf %%lf %%lf",subdirf2(fileresu,"T_"),k1-1,k1-1,vpopbased);
6534: for (j=1; j<= nlstate+1 ; j ++) {
6535: if (j==i) fprintf(ficgp," %%lf (%%lf)");
6536: else fprintf(ficgp," %%*lf (%%*lf)");
6537: }
6538: fprintf(ficgp,"\" t\"\" w l lt 0,");
6539: fprintf(ficgp,"\"%s\" every :::%d::%d u 1:($2==%d && $4!=0 ? $4+$5*2 : 1/0) \"%%lf %%lf %%lf",subdirf2(fileresu,"T_"),k1-1,k1-1,vpopbased);
6540: for (j=1; j<= nlstate+1 ; j ++) {
6541: if (j==i) fprintf(ficgp," %%lf (%%lf)");
6542: else fprintf(ficgp," %%*lf (%%*lf)");
6543: }
6544: if (i== (nlstate+1)) fprintf(ficgp,"\" t\"\" w l lt 0");
6545: else fprintf(ficgp,"\" t\"\" w l lt 0,\\\n");
6546: } /* state */
6547: } /* vpopbased */
1.244 brouard 6548: fprintf(ficgp,"\nset out;set out \"%s_%d-%d.svg\"; replot; set out; \n",subdirf2(optionfilefiname,"E_"),k1,nres); /* Buggy gnuplot */
1.238 brouard 6549: } /* end nres */
6550: } /* k1 end 2 eme*/
6551:
6552:
6553: /*3eme*/
6554: for (k1=1; k1<= m ; k1 ++){
6555: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
1.240 brouard 6556: if(TKresult[nres]!= k1)
1.238 brouard 6557: continue;
6558:
6559: for (cpt=1; cpt<= nlstate ; cpt ++) {
6560: fprintf(ficgp,"\n# 3d: Life expectancy with EXP_ files: combination=%d state=%d",k1, cpt);
6561: for (k=1; k<=cptcoveff; k++){ /* For each covariate and each value */
6562: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate number corresponding to k1 combination */
6563: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6564: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6565: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6566: vlv= nbcode[Tvaraff[k]][lv];
6567: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
6568: }
6569: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
6570: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
6571: }
6572: fprintf(ficgp,"\n#\n");
6573: if(invalidvarcomb[k1]){
6574: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6575: continue;
6576: }
6577:
6578: /* k=2+nlstate*(2*cpt-2); */
6579: k=2+(nlstate+1)*(cpt-1);
1.241 brouard 6580: fprintf(ficgp,"\nset out \"%s_%d-%d-%d.svg\" \n",subdirf2(optionfilefiname,"EXP_"),cpt,k1,nres);
1.238 brouard 6581: fprintf(ficgp,"set ter svg size 640, 480\n\
1.201 brouard 6582: plot [%.f:%.f] \"%s\" every :::%d::%d u 1:%d t \"e%d1\" w l",ageminpar,fage,subdirf2(fileresu,"E_"),k1-1,k1-1,k,cpt);
1.238 brouard 6583: /*fprintf(ficgp,",\"e%s\" every :::%d::%d u 1:($%d-2*$%d) \"\%%lf ",fileres,k1-1,k1-1,k,k+1);
6584: for (i=1; i<= nlstate*2 ; i ++) fprintf(ficgp,"\%%lf (\%%lf) ");
6585: fprintf(ficgp,"\" t \"e%d1\" w l",cpt);
6586: fprintf(ficgp,",\"e%s\" every :::%d::%d u 1:($%d+2*$%d) \"\%%lf ",fileres,k1-1,k1-1,k,k+1);
6587: for (i=1; i<= nlstate*2 ; i ++) fprintf(ficgp,"\%%lf (\%%lf) ");
6588: fprintf(ficgp,"\" t \"e%d1\" w l",cpt);
1.219 brouard 6589:
1.238 brouard 6590: */
6591: for (i=1; i< nlstate ; i ++) {
6592: fprintf(ficgp," ,\"%s\" every :::%d::%d u 1:%d t \"e%d%d\" w l",subdirf2(fileresu,"E_"),k1-1,k1-1,k+i,cpt,i+1);
6593: /* 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 6594:
1.238 brouard 6595: }
6596: fprintf(ficgp," ,\"%s\" every :::%d::%d u 1:%d t \"e%d.\" w l",subdirf2(fileresu,"E_"),k1-1,k1-1,k+nlstate,cpt);
6597: }
6598: } /* end nres */
6599: } /* end kl 3eme */
1.126 brouard 6600:
1.223 brouard 6601: /* 4eme */
1.201 brouard 6602: /* Survival functions (period) from state i in state j by initial state i */
1.238 brouard 6603: for (k1=1; k1<=m; k1++){ /* For each covariate and each value */
6604: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
6605: if(TKresult[nres]!= k1)
1.223 brouard 6606: continue;
1.238 brouard 6607: for (cpt=1; cpt<=nlstate ; cpt ++) { /* For each life state cpt*/
6608: fprintf(ficgp,"\n#\n#\n# Survival functions in state j : 'LIJ_' files, cov=%d state=%d",k1, cpt);
6609: for (k=1; k<=cptcoveff; k++){ /* For each covariate and each value */
6610: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate number corresponding to k1 combination */
6611: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6612: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6613: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6614: vlv= nbcode[Tvaraff[k]][lv];
6615: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
6616: }
6617: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
6618: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
6619: }
6620: fprintf(ficgp,"\n#\n");
6621: if(invalidvarcomb[k1]){
6622: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6623: continue;
1.223 brouard 6624: }
1.238 brouard 6625:
1.241 brouard 6626: fprintf(ficgp,"\nset out \"%s_%d-%d-%d.svg\" \n",subdirf2(optionfilefiname,"LIJ_"),cpt,k1,nres);
1.238 brouard 6627: fprintf(ficgp,"set xlabel \"Age\" \nset ylabel \"Probability to be alive\" \n\
6628: set ter svg size 640, 480\nunset log y\nplot [%.f:%.f] ", ageminpar, agemaxpar);
6629: k=3;
6630: for (i=1; i<= nlstate ; i ++){
6631: if(i==1){
6632: fprintf(ficgp,"\"%s\"",subdirf2(fileresu,"PIJ_"));
6633: }else{
6634: fprintf(ficgp,", '' ");
6635: }
6636: l=(nlstate+ndeath)*(i-1)+1;
6637: fprintf(ficgp," u ($1==%d ? ($3):1/0):($%d/($%d",k1,k+l+(cpt-1),k+l);
6638: for (j=2; j<= nlstate+ndeath ; j ++)
6639: fprintf(ficgp,"+$%d",k+l+j-1);
6640: fprintf(ficgp,")) t \"l(%d,%d)\" w l",i,cpt);
6641: } /* nlstate */
6642: fprintf(ficgp,"\nset out\n");
6643: } /* end cpt state*/
6644: } /* end nres */
6645: } /* end covariate k1 */
6646:
1.220 brouard 6647: /* 5eme */
1.201 brouard 6648: /* Survival functions (period) from state i in state j by final state j */
1.238 brouard 6649: for (k1=1; k1<= m ; k1++){ /* For each covariate combination if any */
6650: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
6651: if(TKresult[nres]!= k1)
1.227 brouard 6652: continue;
1.238 brouard 6653: for (cpt=1; cpt<=nlstate ; cpt ++) { /* For each inital state */
6654: 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);
6655: for (k=1; k<=cptcoveff; k++){ /* For each covariate and each value */
6656: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate number corresponding to k1 combination */
6657: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6658: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6659: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6660: vlv= nbcode[Tvaraff[k]][lv];
6661: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
6662: }
6663: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
6664: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
6665: }
6666: fprintf(ficgp,"\n#\n");
6667: if(invalidvarcomb[k1]){
6668: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6669: continue;
6670: }
1.227 brouard 6671:
1.241 brouard 6672: fprintf(ficgp,"\nset out \"%s_%d-%d-%d.svg\" \n",subdirf2(optionfilefiname,"LIJT_"),cpt,k1,nres);
1.238 brouard 6673: fprintf(ficgp,"set xlabel \"Age\" \nset ylabel \"Probability to be alive\" \n\
6674: set ter svg size 640, 480\nunset log y\nplot [%.f:%.f] ", ageminpar, agemaxpar);
6675: k=3;
6676: for (j=1; j<= nlstate ; j ++){ /* Lived in state j */
6677: if(j==1)
6678: fprintf(ficgp,"\"%s\"",subdirf2(fileresu,"PIJ_"));
6679: else
6680: fprintf(ficgp,", '' ");
6681: l=(nlstate+ndeath)*(cpt-1) +j;
6682: fprintf(ficgp," u (($1==%d && (floor($2)%%5 == 0)) ? ($3):1/0):($%d",k1,k+l);
6683: /* for (i=2; i<= nlstate+ndeath ; i ++) */
6684: /* fprintf(ficgp,"+$%d",k+l+i-1); */
6685: fprintf(ficgp,") t \"l(%d,%d)\" w l",cpt,j);
6686: } /* nlstate */
6687: fprintf(ficgp,", '' ");
6688: fprintf(ficgp," u (($1==%d && (floor($2)%%5 == 0)) ? ($3):1/0):(",k1);
6689: for (j=1; j<= nlstate ; j ++){ /* Lived in state j */
6690: l=(nlstate+ndeath)*(cpt-1) +j;
6691: if(j < nlstate)
6692: fprintf(ficgp,"$%d +",k+l);
6693: else
6694: fprintf(ficgp,"$%d) t\"l(%d,.)\" w l",k+l,cpt);
6695: }
6696: fprintf(ficgp,"\nset out\n");
6697: } /* end cpt state*/
6698: } /* end covariate */
6699: } /* end nres */
1.227 brouard 6700:
1.220 brouard 6701: /* 6eme */
1.202 brouard 6702: /* CV preval stable (period) for each covariate */
1.237 brouard 6703: for (k1=1; k1<= m ; k1 ++) /* For each covariate combination if any */
6704: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
6705: if(TKresult[nres]!= k1)
6706: continue;
1.153 brouard 6707: for (cpt=1; cpt<=nlstate ; cpt ++) { /* For each life state */
1.227 brouard 6708:
1.211 brouard 6709: fprintf(ficgp,"\n#\n#\n#CV preval stable (period): 'pij' files, covariatecombination#=%d state=%d",k1, cpt);
1.225 brouard 6710: for (k=1; k<=cptcoveff; k++){ /* For each covariate and each value */
1.227 brouard 6711: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate number corresponding to k1 combination */
6712: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6713: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6714: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6715: vlv= nbcode[Tvaraff[k]][lv];
6716: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
1.211 brouard 6717: }
1.237 brouard 6718: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
6719: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
6720: }
1.211 brouard 6721: fprintf(ficgp,"\n#\n");
1.223 brouard 6722: if(invalidvarcomb[k1]){
1.227 brouard 6723: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6724: continue;
1.223 brouard 6725: }
1.227 brouard 6726:
1.241 brouard 6727: fprintf(ficgp,"\nset out \"%s_%d-%d-%d.svg\" \n",subdirf2(optionfilefiname,"P_"),cpt,k1,nres);
1.126 brouard 6728: fprintf(ficgp,"set xlabel \"Age\" \nset ylabel \"Probability\" \n\
1.238 brouard 6729: set ter svg size 640, 480\nunset log y\nplot [%.f:%.f] ", ageminpar, agemaxpar);
1.211 brouard 6730: k=3; /* Offset */
1.153 brouard 6731: for (i=1; i<= nlstate ; i ++){
1.227 brouard 6732: if(i==1)
6733: fprintf(ficgp,"\"%s\"",subdirf2(fileresu,"PIJ_"));
6734: else
6735: fprintf(ficgp,", '' ");
6736: l=(nlstate+ndeath)*(i-1)+1;
6737: fprintf(ficgp," u ($1==%d ? ($3):1/0):($%d/($%d",k1,k+l+(cpt-1),k+l);
6738: for (j=2; j<= nlstate ; j ++)
6739: fprintf(ficgp,"+$%d",k+l+j-1);
6740: fprintf(ficgp,")) t \"prev(%d,%d)\" w l",i,cpt);
1.153 brouard 6741: } /* nlstate */
1.201 brouard 6742: fprintf(ficgp,"\nset out\n");
1.153 brouard 6743: } /* end cpt state*/
6744: } /* end covariate */
1.227 brouard 6745:
6746:
1.220 brouard 6747: /* 7eme */
1.218 brouard 6748: if(backcast == 1){
1.217 brouard 6749: /* CV back preval stable (period) for each covariate */
1.237 brouard 6750: for (k1=1; k1<= m ; k1 ++) /* For each covariate combination if any */
6751: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
6752: if(TKresult[nres]!= k1)
6753: continue;
1.218 brouard 6754: for (cpt=1; cpt<=nlstate ; cpt ++) { /* For each life state */
1.227 brouard 6755: fprintf(ficgp,"\n#\n#\n#CV Back preval stable (period): 'pij' files, covariatecombination#=%d state=%d",k1, cpt);
6756: for (k=1; k<=cptcoveff; k++){ /* For each covariate and each value */
6757: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate number corresponding to k1 combination */
6758: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6759: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
1.223 brouard 6760: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
1.227 brouard 6761: vlv= nbcode[Tvaraff[k]][lv];
6762: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
6763: }
1.237 brouard 6764: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
6765: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
6766: }
1.227 brouard 6767: fprintf(ficgp,"\n#\n");
6768: if(invalidvarcomb[k1]){
6769: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6770: continue;
6771: }
6772:
1.241 brouard 6773: fprintf(ficgp,"\nset out \"%s_%d-%d-%d.svg\" \n",subdirf2(optionfilefiname,"PB_"),cpt,k1,nres);
1.227 brouard 6774: fprintf(ficgp,"set xlabel \"Age\" \nset ylabel \"Probability\" \n\
1.238 brouard 6775: set ter svg size 640, 480\nunset log y\nplot [%.f:%.f] ", ageminpar, agemaxpar);
1.227 brouard 6776: k=3; /* Offset */
6777: for (i=1; i<= nlstate ; i ++){
6778: if(i==1)
6779: fprintf(ficgp,"\"%s\"",subdirf2(fileresu,"PIJB_"));
6780: else
6781: fprintf(ficgp,", '' ");
6782: /* l=(nlstate+ndeath)*(i-1)+1; */
6783: l=(nlstate+ndeath)*(cpt-1)+1;
6784: /* fprintf(ficgp," u ($1==%d ? ($3):1/0):($%d/($%d",k1,k+l+(cpt-1),k+l); /\* a vérifier *\/ */
6785: /* fprintf(ficgp," u ($1==%d ? ($3):1/0):($%d/($%d",k1,k+l+(cpt-1),k+l+(cpt-1)+i-1); /\* a vérifier *\/ */
6786: fprintf(ficgp," u ($1==%d ? ($3):1/0):($%d",k1,k+l+(cpt-1)+i-1); /* a vérifier */
6787: /* for (j=2; j<= nlstate ; j ++) */
6788: /* fprintf(ficgp,"+$%d",k+l+j-1); */
6789: /* /\* fprintf(ficgp,"+$%d",k+l+j-1); *\/ */
6790: fprintf(ficgp,") t \"bprev(%d,%d)\" w l",i,cpt);
6791: } /* nlstate */
6792: fprintf(ficgp,"\nset out\n");
1.218 brouard 6793: } /* end cpt state*/
6794: } /* end covariate */
6795: } /* End if backcast */
6796:
1.223 brouard 6797: /* 8eme */
1.218 brouard 6798: if(prevfcast==1){
6799: /* Projection from cross-sectional to stable (period) for each covariate */
6800:
1.237 brouard 6801: for (k1=1; k1<= m ; k1 ++) /* For each covariate combination if any */
6802: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
6803: if(TKresult[nres]!= k1)
6804: continue;
1.211 brouard 6805: for (cpt=1; cpt<=nlstate ; cpt ++) { /* For each life state */
1.227 brouard 6806: fprintf(ficgp,"\n#\n#\n#Projection of prevalence to stable (period): 'PROJ_' files, covariatecombination#=%d state=%d",k1, cpt);
6807: for (k=1; k<=cptcoveff; k++){ /* For each correspondig covariate value */
6808: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate value corresponding to k1 combination and kth covariate */
6809: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6810: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6811: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6812: vlv= nbcode[Tvaraff[k]][lv];
6813: fprintf(ficgp," V%d=%d ",Tvaraff[k],vlv);
6814: }
1.237 brouard 6815: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
6816: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
6817: }
1.227 brouard 6818: fprintf(ficgp,"\n#\n");
6819: if(invalidvarcomb[k1]){
6820: fprintf(ficgp,"#Combination (%d) ignored because no cases \n",k1);
6821: continue;
6822: }
6823:
6824: fprintf(ficgp,"# hpijx=probability over h years, hp.jx is weighted by observed prev\n ");
1.241 brouard 6825: fprintf(ficgp,"\nset out \"%s_%d-%d-%d.svg\" \n",subdirf2(optionfilefiname,"PROJ_"),cpt,k1,nres);
1.227 brouard 6826: fprintf(ficgp,"set xlabel \"Age\" \nset ylabel \"Prevalence\" \n\
1.238 brouard 6827: set ter svg size 640, 480\nunset log y\nplot [%.f:%.f] ", ageminpar, agemaxpar);
1.227 brouard 6828: for (i=1; i<= nlstate+1 ; i ++){ /* nlstate +1 p11 p21 p.1 */
6829: /*# V1 = 1 V2 = 0 yearproj age p11 p21 p.1 p12 p22 p.2 p13 p23 p.3*/
6830: /*# 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 */
6831: /*# yearproj age p11 p21 p.1 p12 p22 p.2 p13 p23 p.3*/
6832: /*# 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 */
6833: if(i==1){
6834: fprintf(ficgp,"\"%s\"",subdirf2(fileresu,"F_"));
6835: }else{
6836: fprintf(ficgp,",\\\n '' ");
6837: }
6838: if(cptcoveff ==0){ /* No covariate */
6839: ioffset=2; /* Age is in 2 */
6840: /*# yearproj age p11 p21 p31 p.1 p12 p22 p32 p.2 p13 p23 p33 p.3 p14 p24 p34 p.4*/
6841: /*# 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 */
6842: /*# V1 = 1 yearproj age p11 p21 p31 p.1 p12 p22 p32 p.2 p13 p23 p33 p.3 p14 p24 p34 p.4*/
6843: /*# 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 */
6844: fprintf(ficgp," u %d:(", ioffset);
6845: if(i==nlstate+1)
6846: fprintf(ficgp," $%d/(1.-$%d)) t 'pw.%d' with line ", \
6847: ioffset+(cpt-1)*(nlstate+1)+1+(i-1), ioffset+1+(i-1)+(nlstate+1)*nlstate,cpt );
6848: else
6849: fprintf(ficgp," $%d/(1.-$%d)) t 'p%d%d' with line ", \
6850: ioffset+(cpt-1)*(nlstate+1)+1+(i-1), ioffset+1+(i-1)+(nlstate+1)*nlstate,i,cpt );
6851: }else{ /* more than 2 covariates */
6852: if(cptcoveff ==1){
6853: ioffset=4; /* Age is in 4 */
6854: }else{
6855: ioffset=6; /* Age is in 6 */
6856: /*# V1 = 1 V2 = 0 yearproj age p11 p21 p.1 p12 p22 p.2 p13 p23 p.3*/
6857: /*# 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 */
6858: }
6859: fprintf(ficgp," u %d:(",ioffset);
6860: kl=0;
6861: strcpy(gplotcondition,"(");
6862: for (k=1; k<=cptcoveff; k++){ /* For each covariate writing the chain of conditions */
6863: lv= decodtabm(k1,k,cptcoveff); /* Should be the covariate value corresponding to combination k1 and covariate k */
6864: /* decodtabm(1,1,4) = 1 because h=1 k= (1) 1 1 1 */
6865: /* decodtabm(1,2,4) = 1 because h=1 k= 1 (1) 1 1 */
6866: /* decodtabm(13,3,4)= 2 because h=13 k= 1 1 (2) 2 */
6867: vlv= nbcode[Tvaraff[k]][lv]; /* Value of the modality of Tvaraff[k] */
6868: kl++;
6869: sprintf(gplotcondition+strlen(gplotcondition),"$%d==%d && $%d==%d " ,kl,Tvaraff[k], kl+1, nbcode[Tvaraff[k]][lv]);
6870: kl++;
6871: if(k <cptcoveff && cptcoveff>1)
6872: sprintf(gplotcondition+strlen(gplotcondition)," && ");
6873: }
6874: strcpy(gplotcondition+strlen(gplotcondition),")");
6875: /* 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 *\/ */
6876: /*6+(cpt-1)*(nlstate+1)+1+(i-1)+(nlstate+1)*nlstate; 6+(1-1)*(2+1)+1+(1-1) +(2+1)*2=13 */
6877: /*6+1+(i-1)+(nlstate+1)*nlstate; 6+1+(1-1) +(2+1)*2=13 */
6878: /* '' 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*/
6879: if(i==nlstate+1){
6880: fprintf(ficgp,"%s ? $%d/(1.-$%d) : 1/0) t 'p.%d' with line ", gplotcondition, \
6881: ioffset+(cpt-1)*(nlstate+1)+1+(i-1), ioffset+1+(i-1)+(nlstate+1)*nlstate,cpt );
6882: }else{
6883: fprintf(ficgp,"%s ? $%d/(1.-$%d) : 1/0) t 'p%d%d' with line ", gplotcondition, \
6884: ioffset+(cpt-1)*(nlstate+1)+1+(i-1), ioffset +1+(i-1)+(nlstate+1)*nlstate,i,cpt );
6885: }
6886: } /* end if covariate */
6887: } /* nlstate */
6888: fprintf(ficgp,"\nset out\n");
1.223 brouard 6889: } /* end cpt state*/
6890: } /* end covariate */
6891: } /* End if prevfcast */
1.227 brouard 6892:
6893:
1.238 brouard 6894: /* 9eme writing MLE parameters */
6895: fprintf(ficgp,"\n##############\n#9eme MLE estimated parameters\n#############\n");
1.126 brouard 6896: for(i=1,jk=1; i <=nlstate; i++){
1.187 brouard 6897: fprintf(ficgp,"# initial state %d\n",i);
1.126 brouard 6898: for(k=1; k <=(nlstate+ndeath); k++){
6899: if (k != i) {
1.227 brouard 6900: fprintf(ficgp,"# current state %d\n",k);
6901: for(j=1; j <=ncovmodel; j++){
6902: fprintf(ficgp,"p%d=%f; ",jk,p[jk]);
6903: jk++;
6904: }
6905: fprintf(ficgp,"\n");
1.126 brouard 6906: }
6907: }
1.223 brouard 6908: }
1.187 brouard 6909: fprintf(ficgp,"##############\n#\n");
1.227 brouard 6910:
1.145 brouard 6911: /*goto avoid;*/
1.238 brouard 6912: /* 10eme Graphics of probabilities or incidences using written MLE parameters */
6913: fprintf(ficgp,"\n##############\n#10eme Graphics of probabilities or incidences\n#############\n");
1.187 brouard 6914: fprintf(ficgp,"# logi(p12/p11)=a12+b12*age+c12age*age+d12*V1+e12*V1*age\n");
6915: fprintf(ficgp,"# logi(p12/p11)=p1 +p2*age +p3*age*age+ p4*V1+ p5*V1*age\n");
6916: fprintf(ficgp,"# logi(p13/p11)=a13+b13*age+c13age*age+d13*V1+e13*V1*age\n");
6917: fprintf(ficgp,"# logi(p13/p11)=p6 +p7*age +p8*age*age+ p9*V1+ p10*V1*age\n");
6918: fprintf(ficgp,"# p12+p13+p14+p11=1=p11(1+exp(a12+b12*age+c12age*age+d12*V1+e12*V1*age)\n");
6919: fprintf(ficgp,"# +exp(a13+b13*age+c13age*age+d13*V1+e13*V1*age)+...)\n");
6920: fprintf(ficgp,"# p11=1/(1+exp(a12+b12*age+c12age*age+d12*V1+e12*V1*age)\n");
6921: fprintf(ficgp,"# +exp(a13+b13*age+c13age*age+d13*V1+e13*V1*age)+...)\n");
6922: fprintf(ficgp,"# p12=exp(a12+b12*age+c12age*age+d12*V1+e12*V1*age)/\n");
6923: fprintf(ficgp,"# (1+exp(a12+b12*age+c12age*age+d12*V1+e12*V1*age)\n");
6924: fprintf(ficgp,"# +exp(a13+b13*age+c13age*age+d13*V1+e13*V1*age))\n");
6925: fprintf(ficgp,"# +exp(a14+b14*age+c14age*age+d14*V1+e14*V1*age)+...)\n");
6926: fprintf(ficgp,"#\n");
1.223 brouard 6927: for(ng=1; ng<=3;ng++){ /* Number of graphics: first is logit, 2nd is probabilities, third is incidences per year*/
1.238 brouard 6928: fprintf(ficgp,"#Number of graphics: first is logit, 2nd is probabilities, third is incidences per year\n");
1.237 brouard 6929: fprintf(ficgp,"#model=%s \n",model);
1.238 brouard 6930: fprintf(ficgp,"# Type of graphic ng=%d\n",ng);
1.237 brouard 6931: fprintf(ficgp,"# jk=1 to 2^%d=%d\n",cptcoveff,m);/* to be checked */
6932: for(jk=1; jk <=m; jk++) /* For each combination of covariate */
6933: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
6934: if(TKresult[nres]!= jk)
6935: continue;
6936: fprintf(ficgp,"# Combination of dummy jk=%d and ",jk);
6937: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
6938: fprintf(ficgp," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
6939: }
6940: fprintf(ficgp,"\n#\n");
1.241 brouard 6941: fprintf(ficgp,"\nset out \"%s_%d-%d-%d.svg\" ",subdirf2(optionfilefiname,"PE_"),jk,ng,nres);
1.223 brouard 6942: fprintf(ficgp,"\nset ter svg size 640, 480 ");
6943: if (ng==1){
6944: fprintf(ficgp,"\nset ylabel \"Value of the logit of the model\"\n"); /* exp(a12+b12*x) could be nice */
6945: fprintf(ficgp,"\nunset log y");
6946: }else if (ng==2){
6947: fprintf(ficgp,"\nset ylabel \"Probability\"\n");
6948: fprintf(ficgp,"\nset log y");
6949: }else if (ng==3){
6950: fprintf(ficgp,"\nset ylabel \"Quasi-incidence per year\"\n");
6951: fprintf(ficgp,"\nset log y");
6952: }else
6953: fprintf(ficgp,"\nunset title ");
6954: fprintf(ficgp,"\nplot [%.f:%.f] ",ageminpar,agemaxpar);
6955: i=1;
6956: for(k2=1; k2<=nlstate; k2++) {
6957: k3=i;
6958: for(k=1; k<=(nlstate+ndeath); k++) {
6959: if (k != k2){
6960: switch( ng) {
6961: case 1:
6962: if(nagesqr==0)
6963: fprintf(ficgp," p%d+p%d*x",i,i+1);
6964: else /* nagesqr =1 */
6965: fprintf(ficgp," p%d+p%d*x+p%d*x*x",i,i+1,i+1+nagesqr);
6966: break;
6967: case 2: /* ng=2 */
6968: if(nagesqr==0)
6969: fprintf(ficgp," exp(p%d+p%d*x",i,i+1);
6970: else /* nagesqr =1 */
6971: fprintf(ficgp," exp(p%d+p%d*x+p%d*x*x",i,i+1,i+1+nagesqr);
6972: break;
6973: case 3:
6974: if(nagesqr==0)
6975: fprintf(ficgp," %f*exp(p%d+p%d*x",YEARM/stepm,i,i+1);
6976: else /* nagesqr =1 */
6977: fprintf(ficgp," %f*exp(p%d+p%d*x+p%d*x*x",YEARM/stepm,i,i+1,i+1+nagesqr);
6978: break;
6979: }
6980: ij=1;/* To be checked else nbcode[0][0] wrong */
1.237 brouard 6981: ijp=1; /* product no age */
6982: /* for(j=3; j <=ncovmodel-nagesqr; j++) { */
6983: for(j=1; j <=cptcovt; j++) { /* For each covariate of the simplified model */
1.223 brouard 6984: /* printf("Tage[%d]=%d, j=%d\n", ij, Tage[ij], j); */
1.237 brouard 6985: if(j==Tage[ij]) { /* Product by age */
6986: if(ij <=cptcovage) { /* V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1, 2 V5 and V1 */
1.238 brouard 6987: if(DummyV[j]==0){
1.237 brouard 6988: fprintf(ficgp,"+p%d*%d*x",i+j+2+nagesqr-1,Tinvresult[nres][Tvar[j]]);;
6989: }else{ /* quantitative */
6990: fprintf(ficgp,"+p%d*%f*x",i+j+2+nagesqr-1,Tqinvresult[nres][Tvar[j]]); /* Tqinvresult in decoderesult */
6991: /* fprintf(ficgp,"+p%d*%d*x",i+j+nagesqr-1,nbcode[Tvar[j-2]][codtabm(jk,Tvar[j-2])]); */
6992: }
6993: ij++;
6994: }
6995: }else if(j==Tprod[ijp]) { /* */
6996: /* printf("Tprod[%d]=%d, j=%d\n", ij, Tprod[ijp], j); */
6997: if(ijp <=cptcovprod) { /* Product */
1.238 brouard 6998: if(DummyV[Tvard[ijp][1]]==0){/* Vn is dummy */
6999: if(DummyV[Tvard[ijp][2]]==0){/* Vn and Vm are dummy */
1.237 brouard 7000: /* fprintf(ficgp,"+p%d*%d*%d",i+j+2+nagesqr-1,nbcode[Tvard[ijp][1]][codtabm(jk,j)],nbcode[Tvard[ijp][2]][codtabm(jk,j)]); */
7001: fprintf(ficgp,"+p%d*%d*%d",i+j+2+nagesqr-1,Tinvresult[nres][Tvard[ijp][1]],Tinvresult[nres][Tvard[ijp][2]]);
7002: }else{ /* Vn is dummy and Vm is quanti */
7003: /* fprintf(ficgp,"+p%d*%d*%f",i+j+2+nagesqr-1,nbcode[Tvard[ijp][1]][codtabm(jk,j)],Tqinvresult[nres][Tvard[ijp][2]]); */
7004: fprintf(ficgp,"+p%d*%d*%f",i+j+2+nagesqr-1,Tinvresult[nres][Tvard[ijp][1]],Tqinvresult[nres][Tvard[ijp][2]]);
7005: }
7006: }else{ /* Vn*Vm Vn is quanti */
1.238 brouard 7007: if(DummyV[Tvard[ijp][2]]==0){
1.237 brouard 7008: fprintf(ficgp,"+p%d*%d*%f",i+j+2+nagesqr-1,Tinvresult[nres][Tvard[ijp][2]],Tqinvresult[nres][Tvard[ijp][1]]);
7009: }else{ /* Both quanti */
7010: fprintf(ficgp,"+p%d*%f*%f",i+j+2+nagesqr-1,Tqinvresult[nres][Tvard[ijp][1]],Tqinvresult[nres][Tvard[ijp][2]]);
7011: }
7012: }
1.238 brouard 7013: ijp++;
1.237 brouard 7014: }
7015: } else{ /* simple covariate */
7016: /* fprintf(ficgp,"+p%d*%d",i+j+2+nagesqr-1,nbcode[Tvar[j]][codtabm(jk,j)]); /\* Valgrind bug nbcode *\/ */
7017: if(Dummy[j]==0){
7018: fprintf(ficgp,"+p%d*%d",i+j+2+nagesqr-1,Tinvresult[nres][Tvar[j]]); /* */
7019: }else{ /* quantitative */
7020: fprintf(ficgp,"+p%d*%f",i+j+2+nagesqr-1,Tqinvresult[nres][Tvar[j]]); /* */
1.223 brouard 7021: /* fprintf(ficgp,"+p%d*%d*x",i+j+nagesqr-1,nbcode[Tvar[j-2]][codtabm(jk,Tvar[j-2])]); */
7022: }
1.237 brouard 7023: } /* end simple */
7024: } /* end j */
1.223 brouard 7025: }else{
7026: i=i-ncovmodel;
7027: if(ng !=1 ) /* For logit formula of log p11 is more difficult to get */
7028: fprintf(ficgp," (1.");
7029: }
1.227 brouard 7030:
1.223 brouard 7031: if(ng != 1){
7032: fprintf(ficgp,")/(1");
1.227 brouard 7033:
1.223 brouard 7034: for(k1=1; k1 <=nlstate; k1++){
7035: if(nagesqr==0)
7036: fprintf(ficgp,"+exp(p%d+p%d*x",k3+(k1-1)*ncovmodel,k3+(k1-1)*ncovmodel+1);
7037: else /* nagesqr =1 */
7038: fprintf(ficgp,"+exp(p%d+p%d*x+p%d*x*x",k3+(k1-1)*ncovmodel,k3+(k1-1)*ncovmodel+1,k3+(k1-1)*ncovmodel+1+nagesqr);
1.217 brouard 7039:
1.223 brouard 7040: ij=1;
7041: for(j=3; j <=ncovmodel-nagesqr; j++){
1.237 brouard 7042: if((j-2)==Tage[ij]) { /* Bug valgrind */
7043: if(ij <=cptcovage) { /* Bug valgrind */
1.223 brouard 7044: fprintf(ficgp,"+p%d*%d*x",k3+(k1-1)*ncovmodel+1+j-2+nagesqr,nbcode[Tvar[j-2]][codtabm(jk,j-2)]);
7045: /* fprintf(ficgp,"+p%d*%d*x",k3+(k1-1)*ncovmodel+1+j-2+nagesqr,nbcode[Tvar[j-2]][codtabm(jk,Tvar[j-2])]); */
7046: ij++;
7047: }
7048: }
7049: else
1.225 brouard 7050: fprintf(ficgp,"+p%d*%d",k3+(k1-1)*ncovmodel+1+j-2+nagesqr,nbcode[Tvar[j-2]][codtabm(jk,j-2)]);/* Valgrind bug nbcode */
1.223 brouard 7051: }
7052: fprintf(ficgp,")");
7053: }
7054: fprintf(ficgp,")");
7055: if(ng ==2)
7056: fprintf(ficgp," t \"p%d%d\" ", k2,k);
7057: else /* ng= 3 */
7058: fprintf(ficgp," t \"i%d%d\" ", k2,k);
7059: }else{ /* end ng <> 1 */
7060: if( k !=k2) /* logit p11 is hard to draw */
7061: fprintf(ficgp," t \"logit(p%d%d)\" ", k2,k);
7062: }
7063: if ((k+k2)!= (nlstate*2+ndeath) && ng != 1)
7064: fprintf(ficgp,",");
7065: if (ng == 1 && k!=k2 && (k+k2)!= (nlstate*2+ndeath))
7066: fprintf(ficgp,",");
7067: i=i+ncovmodel;
7068: } /* end k */
7069: } /* end k2 */
7070: fprintf(ficgp,"\n set out\n");
7071: } /* end jk */
7072: } /* end ng */
7073: /* avoid: */
7074: fflush(ficgp);
1.126 brouard 7075: } /* end gnuplot */
7076:
7077:
7078: /*************** Moving average **************/
1.219 brouard 7079: /* int movingaverage(double ***probs, double bage, double fage, double ***mobaverage, int mobilav, double bageout, double fageout){ */
1.222 brouard 7080: int movingaverage(double ***probs, double bage, double fage, double ***mobaverage, int mobilav){
1.218 brouard 7081:
1.222 brouard 7082: int i, cpt, cptcod;
7083: int modcovmax =1;
7084: int mobilavrange, mob;
7085: int iage=0;
7086:
7087: double sum=0.;
7088: double age;
7089: double *sumnewp, *sumnewm;
7090: double *agemingood, *agemaxgood; /* Currently identical for all covariates */
7091:
7092:
1.225 brouard 7093: /* modcovmax=2*cptcoveff;/\* Max number of modalities. We suppose */
1.222 brouard 7094: /* a covariate has 2 modalities, should be equal to ncovcombmax *\/ */
7095:
7096: sumnewp = vector(1,ncovcombmax);
7097: sumnewm = vector(1,ncovcombmax);
7098: agemingood = vector(1,ncovcombmax);
7099: agemaxgood = vector(1,ncovcombmax);
7100:
7101: for (cptcod=1;cptcod<=ncovcombmax;cptcod++){
7102: sumnewm[cptcod]=0.;
7103: sumnewp[cptcod]=0.;
7104: agemingood[cptcod]=0;
7105: agemaxgood[cptcod]=0;
7106: }
7107: if (cptcovn<1) ncovcombmax=1; /* At least 1 pass */
7108:
7109: if(mobilav==1||mobilav ==3 ||mobilav==5 ||mobilav== 7){
7110: if(mobilav==1) mobilavrange=5; /* default */
7111: else mobilavrange=mobilav;
7112: for (age=bage; age<=fage; age++)
7113: for (i=1; i<=nlstate;i++)
7114: for (cptcod=1;cptcod<=ncovcombmax;cptcod++)
7115: mobaverage[(int)age][i][cptcod]=probs[(int)age][i][cptcod];
7116: /* We keep the original values on the extreme ages bage, fage and for
7117: fage+1 and bage-1 we use a 3 terms moving average; for fage+2 bage+2
7118: we use a 5 terms etc. until the borders are no more concerned.
7119: */
7120: for (mob=3;mob <=mobilavrange;mob=mob+2){
7121: for (age=bage+(mob-1)/2; age<=fage-(mob-1)/2; age++){
7122: for (i=1; i<=nlstate;i++){
7123: for (cptcod=1;cptcod<=ncovcombmax;cptcod++){
7124: mobaverage[(int)age][i][cptcod] =probs[(int)age][i][cptcod];
7125: for (cpt=1;cpt<=(mob-1)/2;cpt++){
7126: mobaverage[(int)age][i][cptcod] +=probs[(int)age-cpt][i][cptcod];
7127: mobaverage[(int)age][i][cptcod] +=probs[(int)age+cpt][i][cptcod];
7128: }
7129: mobaverage[(int)age][i][cptcod]=mobaverage[(int)age][i][cptcod]/mob;
7130: }
7131: }
7132: }/* end age */
7133: }/* end mob */
7134: }else
7135: return -1;
7136: for (cptcod=1;cptcod<=ncovcombmax;cptcod++){
7137: /* for (age=bage+(mob-1)/2; age<=fage-(mob-1)/2; age++){ */
7138: if(invalidvarcomb[cptcod]){
7139: printf("\nCombination (%d) ignored because no cases \n",cptcod);
7140: continue;
7141: }
1.219 brouard 7142:
1.222 brouard 7143: agemingood[cptcod]=fage-(mob-1)/2;
7144: for (age=fage-(mob-1)/2; age>=bage; age--){/* From oldest to youngest, finding the youngest wrong */
7145: sumnewm[cptcod]=0.;
7146: for (i=1; i<=nlstate;i++){
7147: sumnewm[cptcod]+=mobaverage[(int)age][i][cptcod];
7148: }
7149: if(fabs(sumnewm[cptcod] - 1.) <= 1.e-3) { /* good */
7150: agemingood[cptcod]=age;
7151: }else{ /* bad */
7152: for (i=1; i<=nlstate;i++){
7153: mobaverage[(int)age][i][cptcod]=mobaverage[(int)agemingood[cptcod]][i][cptcod];
7154: } /* i */
7155: } /* end bad */
7156: }/* age */
7157: sum=0.;
7158: for (i=1; i<=nlstate;i++){
7159: sum+=mobaverage[(int)agemingood[cptcod]][i][cptcod];
7160: }
7161: if(fabs(sum - 1.) > 1.e-3) { /* bad */
7162: 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);
7163: /* for (i=1; i<=nlstate;i++){ */
7164: /* mobaverage[(int)age][i][cptcod]=mobaverage[(int)agemingood[cptcod]][i][cptcod]; */
7165: /* } /\* i *\/ */
7166: } /* end bad */
7167: /* else{ /\* We found some ages summing to one, we will smooth the oldest *\/ */
7168: /* From youngest, finding the oldest wrong */
7169: agemaxgood[cptcod]=bage+(mob-1)/2;
7170: for (age=bage+(mob-1)/2; age<=fage; age++){
7171: sumnewm[cptcod]=0.;
7172: for (i=1; i<=nlstate;i++){
7173: sumnewm[cptcod]+=mobaverage[(int)age][i][cptcod];
7174: }
7175: if(fabs(sumnewm[cptcod] - 1.) <= 1.e-3) { /* good */
7176: agemaxgood[cptcod]=age;
7177: }else{ /* bad */
7178: for (i=1; i<=nlstate;i++){
7179: mobaverage[(int)age][i][cptcod]=mobaverage[(int)agemaxgood[cptcod]][i][cptcod];
7180: } /* i */
7181: } /* end bad */
7182: }/* age */
7183: sum=0.;
7184: for (i=1; i<=nlstate;i++){
7185: sum+=mobaverage[(int)agemaxgood[cptcod]][i][cptcod];
7186: }
7187: if(fabs(sum - 1.) > 1.e-3) { /* bad */
7188: 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);
7189: /* for (i=1; i<=nlstate;i++){ */
7190: /* mobaverage[(int)age][i][cptcod]=mobaverage[(int)agemingood[cptcod]][i][cptcod]; */
7191: /* } /\* i *\/ */
7192: } /* end bad */
7193:
7194: for (age=bage; age<=fage; age++){
1.235 brouard 7195: /* printf("%d %d ", cptcod, (int)age); */
1.222 brouard 7196: sumnewp[cptcod]=0.;
7197: sumnewm[cptcod]=0.;
7198: for (i=1; i<=nlstate;i++){
7199: sumnewp[cptcod]+=probs[(int)age][i][cptcod];
7200: sumnewm[cptcod]+=mobaverage[(int)age][i][cptcod];
7201: /* printf("%.4f %.4f ",probs[(int)age][i][cptcod], mobaverage[(int)age][i][cptcod]); */
7202: }
7203: /* printf("%.4f %.4f \n",sumnewp[cptcod], sumnewm[cptcod]); */
7204: }
7205: /* printf("\n"); */
7206: /* } */
7207: /* brutal averaging */
7208: for (i=1; i<=nlstate;i++){
7209: for (age=1; age<=bage; age++){
7210: mobaverage[(int)age][i][cptcod]=mobaverage[(int)agemingood[cptcod]][i][cptcod];
7211: /* printf("age=%d i=%d cptcod=%d mobaverage=%.4f \n",(int)age,i, cptcod, mobaverage[(int)age][i][cptcod]); */
7212: }
7213: for (age=fage; age<=AGESUP; age++){
7214: mobaverage[(int)age][i][cptcod]=mobaverage[(int)agemaxgood[cptcod]][i][cptcod];
7215: /* printf("age=%d i=%d cptcod=%d mobaverage=%.4f \n",(int)age,i, cptcod, mobaverage[(int)age][i][cptcod]); */
7216: }
7217: } /* end i status */
7218: for (i=nlstate+1; i<=nlstate+ndeath;i++){
7219: for (age=1; age<=AGESUP; age++){
7220: /*printf("i=%d, age=%d, cptcod=%d\n",i, (int)age, cptcod);*/
7221: mobaverage[(int)age][i][cptcod]=0.;
7222: }
7223: }
7224: }/* end cptcod */
7225: free_vector(sumnewm,1, ncovcombmax);
7226: free_vector(sumnewp,1, ncovcombmax);
7227: free_vector(agemaxgood,1, ncovcombmax);
7228: free_vector(agemingood,1, ncovcombmax);
7229: return 0;
7230: }/* End movingaverage */
1.218 brouard 7231:
1.126 brouard 7232:
7233: /************** Forecasting ******************/
1.235 brouard 7234: 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 7235: /* proj1, year, month, day of starting projection
7236: agemin, agemax range of age
7237: dateprev1 dateprev2 range of dates during which prevalence is computed
7238: anproj2 year of en of projection (same day and month as proj1).
7239: */
1.235 brouard 7240: int yearp, stepsize, hstepm, nhstepm, j, k, cptcod, i, h, i1, k4, nres=0;
1.126 brouard 7241: double agec; /* generic age */
7242: double agelim, ppij, yp,yp1,yp2,jprojmean,mprojmean,anprojmean;
7243: double *popeffectif,*popcount;
7244: double ***p3mat;
1.218 brouard 7245: /* double ***mobaverage; */
1.126 brouard 7246: char fileresf[FILENAMELENGTH];
7247:
7248: agelim=AGESUP;
1.211 brouard 7249: /* Compute observed prevalence between dateprev1 and dateprev2 by counting the number of people
7250: in each health status at the date of interview (if between dateprev1 and dateprev2).
7251: We still use firstpass and lastpass as another selection.
7252: */
1.214 brouard 7253: /* freqsummary(fileres, agemin, agemax, s, agev, nlstate, imx,Tvaraff,nbcode, ncodemax,mint,anint,strstart,\ */
7254: /* firstpass, lastpass, stepm, weightopt, model); */
1.126 brouard 7255:
1.201 brouard 7256: strcpy(fileresf,"F_");
7257: strcat(fileresf,fileresu);
1.126 brouard 7258: if((ficresf=fopen(fileresf,"w"))==NULL) {
7259: printf("Problem with forecast resultfile: %s\n", fileresf);
7260: fprintf(ficlog,"Problem with forecast resultfile: %s\n", fileresf);
7261: }
1.235 brouard 7262: printf("\nComputing forecasting: result on file '%s', please wait... \n", fileresf);
7263: fprintf(ficlog,"\nComputing forecasting: result on file '%s', please wait... \n", fileresf);
1.126 brouard 7264:
1.225 brouard 7265: if (cptcoveff==0) ncodemax[cptcoveff]=1;
1.126 brouard 7266:
7267:
7268: stepsize=(int) (stepm+YEARM-1)/YEARM;
7269: if (stepm<=12) stepsize=1;
7270: if(estepm < stepm){
7271: printf ("Problem %d lower than %d\n",estepm, stepm);
7272: }
7273: else hstepm=estepm;
7274:
7275: hstepm=hstepm/stepm;
7276: yp1=modf(dateintmean,&yp);/* extracts integral of datemean in yp and
7277: fractional in yp1 */
7278: anprojmean=yp;
7279: yp2=modf((yp1*12),&yp);
7280: mprojmean=yp;
7281: yp1=modf((yp2*30.5),&yp);
7282: jprojmean=yp;
7283: if(jprojmean==0) jprojmean=1;
7284: if(mprojmean==0) jprojmean=1;
7285:
1.227 brouard 7286: i1=pow(2,cptcoveff);
1.126 brouard 7287: if (cptcovn < 1){i1=1;}
7288:
7289: fprintf(ficresf,"# Mean day of interviews %.lf/%.lf/%.lf (%.2f) between %.2f and %.2f \n",jprojmean,mprojmean,anprojmean,dateintmean,dateprev1,dateprev2);
7290:
7291: fprintf(ficresf,"#****** Routine prevforecast **\n");
1.227 brouard 7292:
1.126 brouard 7293: /* if (h==(int)(YEARM*yearp)){ */
1.235 brouard 7294: for(nres=1; nres <= nresult; nres++) /* For each resultline */
7295: for(k=1; k<=i1;k++){
7296: if(TKresult[nres]!= k)
7297: continue;
1.227 brouard 7298: if(invalidvarcomb[k]){
7299: printf("\nCombination (%d) projection ignored because no cases \n",k);
7300: continue;
7301: }
7302: fprintf(ficresf,"\n#****** hpijx=probability over h years, hp.jx is weighted by observed prev \n#");
7303: for(j=1;j<=cptcoveff;j++) {
7304: fprintf(ficresf," V%d (=) %d",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
7305: }
1.235 brouard 7306: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
1.238 brouard 7307: fprintf(ficresf," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
1.235 brouard 7308: }
1.227 brouard 7309: fprintf(ficresf," yearproj age");
7310: for(j=1; j<=nlstate+ndeath;j++){
7311: for(i=1; i<=nlstate;i++)
7312: fprintf(ficresf," p%d%d",i,j);
7313: fprintf(ficresf," wp.%d",j);
7314: }
7315: for (yearp=0; yearp<=(anproj2-anproj1);yearp +=stepsize) {
7316: fprintf(ficresf,"\n");
7317: fprintf(ficresf,"\n# Forecasting at date %.lf/%.lf/%.lf ",jproj1,mproj1,anproj1+yearp);
7318: for (agec=fage; agec>=(ageminpar-1); agec--){
7319: nhstepm=(int) rint((agelim-agec)*YEARM/stepm);
7320: nhstepm = nhstepm/hstepm;
7321: p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
7322: oldm=oldms;savm=savms;
1.235 brouard 7323: hpxij(p3mat,nhstepm,agec,hstepm,p,nlstate,stepm,oldm,savm, k,nres);
1.227 brouard 7324:
7325: for (h=0; h<=nhstepm; h++){
7326: if (h*hstepm/YEARM*stepm ==yearp) {
7327: fprintf(ficresf,"\n");
7328: for(j=1;j<=cptcoveff;j++)
7329: fprintf(ficresf,"%d %d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
7330: fprintf(ficresf,"%.f %.f ",anproj1+yearp,agec+h*hstepm/YEARM*stepm);
7331: }
7332: for(j=1; j<=nlstate+ndeath;j++) {
7333: ppij=0.;
7334: for(i=1; i<=nlstate;i++) {
7335: if (mobilav==1)
7336: ppij=ppij+p3mat[i][j][h]*mobaverage[(int)agec][i][k];
7337: else {
7338: ppij=ppij+p3mat[i][j][h]*probs[(int)(agec)][i][k];
7339: }
7340: if (h*hstepm/YEARM*stepm== yearp) {
7341: fprintf(ficresf," %.3f", p3mat[i][j][h]);
7342: }
7343: } /* end i */
7344: if (h*hstepm/YEARM*stepm==yearp) {
7345: fprintf(ficresf," %.3f", ppij);
7346: }
7347: }/* end j */
7348: } /* end h */
7349: free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
7350: } /* end agec */
7351: } /* end yearp */
7352: } /* end k */
1.219 brouard 7353:
1.126 brouard 7354: fclose(ficresf);
1.215 brouard 7355: printf("End of Computing forecasting \n");
7356: fprintf(ficlog,"End of Computing forecasting\n");
7357:
1.126 brouard 7358: }
7359:
1.218 brouard 7360: /* /\************** Back Forecasting ******************\/ */
1.225 brouard 7361: /* 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 7362: /* /\* back1, year, month, day of starting backection */
7363: /* agemin, agemax range of age */
7364: /* dateprev1 dateprev2 range of dates during which prevalence is computed */
7365: /* anback2 year of en of backection (same day and month as back1). */
7366: /* *\/ */
7367: /* int yearp, stepsize, hstepm, nhstepm, j, k, cptcod, i, h, i1; */
7368: /* double agec; /\* generic age *\/ */
7369: /* double agelim, ppij, yp,yp1,yp2,jprojmean,mprojmean,anprojmean; */
7370: /* double *popeffectif,*popcount; */
7371: /* double ***p3mat; */
7372: /* /\* double ***mobaverage; *\/ */
7373: /* char fileresfb[FILENAMELENGTH]; */
7374:
7375: /* agelim=AGESUP; */
7376: /* /\* Compute observed prevalence between dateprev1 and dateprev2 by counting the number of people */
7377: /* in each health status at the date of interview (if between dateprev1 and dateprev2). */
7378: /* We still use firstpass and lastpass as another selection. */
7379: /* *\/ */
7380: /* /\* freqsummary(fileres, agemin, agemax, s, agev, nlstate, imx,Tvaraff,nbcode, ncodemax,mint,anint,strstart,\ *\/ */
7381: /* /\* firstpass, lastpass, stepm, weightopt, model); *\/ */
7382: /* prevalence(probs, ageminpar, agemax, s, agev, nlstate, imx, Tvar, nbcode, ncodemax, mint, anint, dateprev1, dateprev2, firstpass, lastpass); */
7383:
7384: /* strcpy(fileresfb,"FB_"); */
7385: /* strcat(fileresfb,fileresu); */
7386: /* if((ficresfb=fopen(fileresfb,"w"))==NULL) { */
7387: /* printf("Problem with back forecast resultfile: %s\n", fileresfb); */
7388: /* fprintf(ficlog,"Problem with back forecast resultfile: %s\n", fileresfb); */
7389: /* } */
7390: /* printf("Computing back forecasting: result on file '%s', please wait... \n", fileresfb); */
7391: /* fprintf(ficlog,"Computing back forecasting: result on file '%s', please wait... \n", fileresfb); */
7392:
1.225 brouard 7393: /* if (cptcoveff==0) ncodemax[cptcoveff]=1; */
1.218 brouard 7394:
7395: /* /\* if (mobilav!=0) { *\/ */
7396: /* /\* mobaverage= ma3x(1, AGESUP,1,NCOVMAX, 1,NCOVMAX); *\/ */
7397: /* /\* if (movingaverage(probs, ageminpar, fage, mobaverage,mobilav)!=0){ *\/ */
7398: /* /\* fprintf(ficlog," Error in movingaverage mobilav=%d\n",mobilav); *\/ */
7399: /* /\* printf(" Error in movingaverage mobilav=%d\n",mobilav); *\/ */
7400: /* /\* } *\/ */
7401: /* /\* } *\/ */
7402:
7403: /* stepsize=(int) (stepm+YEARM-1)/YEARM; */
7404: /* if (stepm<=12) stepsize=1; */
7405: /* if(estepm < stepm){ */
7406: /* printf ("Problem %d lower than %d\n",estepm, stepm); */
7407: /* } */
7408: /* else hstepm=estepm; */
7409:
7410: /* hstepm=hstepm/stepm; */
7411: /* yp1=modf(dateintmean,&yp);/\* extracts integral of datemean in yp and */
7412: /* fractional in yp1 *\/ */
7413: /* anprojmean=yp; */
7414: /* yp2=modf((yp1*12),&yp); */
7415: /* mprojmean=yp; */
7416: /* yp1=modf((yp2*30.5),&yp); */
7417: /* jprojmean=yp; */
7418: /* if(jprojmean==0) jprojmean=1; */
7419: /* if(mprojmean==0) jprojmean=1; */
7420:
1.225 brouard 7421: /* i1=cptcoveff; */
1.218 brouard 7422: /* if (cptcovn < 1){i1=1;} */
1.217 brouard 7423:
1.218 brouard 7424: /* fprintf(ficresfb,"# Mean day of interviews %.lf/%.lf/%.lf (%.2f) between %.2f and %.2f \n",jprojmean,mprojmean,anprojmean,dateintmean,dateprev1,dateprev2); */
1.217 brouard 7425:
1.218 brouard 7426: /* fprintf(ficresfb,"#****** Routine prevbackforecast **\n"); */
7427:
7428: /* /\* if (h==(int)(YEARM*yearp)){ *\/ */
7429: /* for(cptcov=1, k=0;cptcov<=i1;cptcov++){ */
1.225 brouard 7430: /* for(cptcod=1;cptcod<=ncodemax[cptcoveff];cptcod++){ */
1.218 brouard 7431: /* k=k+1; */
7432: /* fprintf(ficresfb,"\n#****** hbijx=probability over h years, hp.jx is weighted by observed prev \n#"); */
1.225 brouard 7433: /* for(j=1;j<=cptcoveff;j++) { */
1.218 brouard 7434: /* fprintf(ficresfb," V%d (=) %d",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]); */
7435: /* } */
7436: /* fprintf(ficresfb," yearbproj age"); */
7437: /* for(j=1; j<=nlstate+ndeath;j++){ */
7438: /* for(i=1; i<=nlstate;i++) */
7439: /* fprintf(ficresfb," p%d%d",i,j); */
7440: /* fprintf(ficresfb," p.%d",j); */
7441: /* } */
7442: /* for (yearp=0; yearp>=(anback2-anback1);yearp -=stepsize) { */
7443: /* /\* for (yearp=0; yearp<=(anproj2-anproj1);yearp +=stepsize) { *\/ */
7444: /* fprintf(ficresfb,"\n"); */
7445: /* fprintf(ficresfb,"\n# Back Forecasting at date %.lf/%.lf/%.lf ",jback1,mback1,anback1+yearp); */
7446: /* for (agec=fage; agec>=(ageminpar-1); agec--){ */
7447: /* nhstepm=(int) rint((agelim-agec)*YEARM/stepm); */
7448: /* nhstepm = nhstepm/hstepm; */
7449: /* p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm); */
7450: /* oldm=oldms;savm=savms; */
7451: /* hbxij(p3mat,nhstepm,agec,hstepm,p,prevacurrent,nlstate,stepm,oldm,savm,oldm,savm, dnewm, doldm, dsavm, k); */
7452: /* for (h=0; h<=nhstepm; h++){ */
7453: /* if (h*hstepm/YEARM*stepm ==yearp) { */
7454: /* fprintf(ficresfb,"\n"); */
1.225 brouard 7455: /* for(j=1;j<=cptcoveff;j++) */
1.218 brouard 7456: /* fprintf(ficresfb,"%d %d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]); */
7457: /* fprintf(ficresfb,"%.f %.f ",anback1+yearp,agec+h*hstepm/YEARM*stepm); */
7458: /* } */
7459: /* for(j=1; j<=nlstate+ndeath;j++) { */
7460: /* ppij=0.; */
7461: /* for(i=1; i<=nlstate;i++) { */
7462: /* if (mobilav==1) */
7463: /* ppij=ppij+p3mat[i][j][h]*mobaverage[(int)agec][i][cptcod]; */
7464: /* else { */
7465: /* ppij=ppij+p3mat[i][j][h]*probs[(int)(agec)][i][cptcod]; */
7466: /* } */
7467: /* if (h*hstepm/YEARM*stepm== yearp) { */
7468: /* fprintf(ficresfb," %.3f", p3mat[i][j][h]); */
7469: /* } */
7470: /* } /\* end i *\/ */
7471: /* if (h*hstepm/YEARM*stepm==yearp) { */
7472: /* fprintf(ficresfb," %.3f", ppij); */
7473: /* } */
7474: /* }/\* end j *\/ */
7475: /* } /\* end h *\/ */
7476: /* free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm); */
7477: /* } /\* end agec *\/ */
7478: /* } /\* end yearp *\/ */
7479: /* } /\* end cptcod *\/ */
7480: /* } /\* end cptcov *\/ */
7481:
7482: /* /\* if (mobilav!=0) free_ma3x(mobaverage,1, AGESUP,1,NCOVMAX, 1,NCOVMAX); *\/ */
7483:
7484: /* fclose(ficresfb); */
7485: /* printf("End of Computing Back forecasting \n"); */
7486: /* fprintf(ficlog,"End of Computing Back forecasting\n"); */
1.217 brouard 7487:
1.218 brouard 7488: /* } */
1.217 brouard 7489:
1.126 brouard 7490: /************** Forecasting *****not tested NB*************/
1.227 brouard 7491: /* 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 7492:
1.227 brouard 7493: /* int cpt, stepsize, hstepm, nhstepm, j,k,c, cptcod, i,h; */
7494: /* int *popage; */
7495: /* double calagedatem, agelim, kk1, kk2; */
7496: /* double *popeffectif,*popcount; */
7497: /* double ***p3mat,***tabpop,***tabpopprev; */
7498: /* /\* double ***mobaverage; *\/ */
7499: /* char filerespop[FILENAMELENGTH]; */
1.126 brouard 7500:
1.227 brouard 7501: /* tabpop= ma3x(1, AGESUP,1,NCOVMAX, 1,NCOVMAX); */
7502: /* tabpopprev= ma3x(1, AGESUP,1,NCOVMAX, 1,NCOVMAX); */
7503: /* agelim=AGESUP; */
7504: /* calagedatem=(anpyram+mpyram/12.+jpyram/365.-dateintmean)*YEARM; */
1.126 brouard 7505:
1.227 brouard 7506: /* prevalence(probs, ageminpar, agemax, s, agev, nlstate, imx, Tvar, nbcode, ncodemax, mint, anint, dateprev1, dateprev2, firstpass, lastpass); */
1.126 brouard 7507:
7508:
1.227 brouard 7509: /* strcpy(filerespop,"POP_"); */
7510: /* strcat(filerespop,fileresu); */
7511: /* if((ficrespop=fopen(filerespop,"w"))==NULL) { */
7512: /* printf("Problem with forecast resultfile: %s\n", filerespop); */
7513: /* fprintf(ficlog,"Problem with forecast resultfile: %s\n", filerespop); */
7514: /* } */
7515: /* printf("Computing forecasting: result on file '%s' \n", filerespop); */
7516: /* fprintf(ficlog,"Computing forecasting: result on file '%s' \n", filerespop); */
1.126 brouard 7517:
1.227 brouard 7518: /* if (cptcoveff==0) ncodemax[cptcoveff]=1; */
1.126 brouard 7519:
1.227 brouard 7520: /* /\* if (mobilav!=0) { *\/ */
7521: /* /\* mobaverage= ma3x(1, AGESUP,1,NCOVMAX, 1,NCOVMAX); *\/ */
7522: /* /\* if (movingaverage(probs, ageminpar, fage, mobaverage,mobilav)!=0){ *\/ */
7523: /* /\* fprintf(ficlog," Error in movingaverage mobilav=%d\n",mobilav); *\/ */
7524: /* /\* printf(" Error in movingaverage mobilav=%d\n",mobilav); *\/ */
7525: /* /\* } *\/ */
7526: /* /\* } *\/ */
1.126 brouard 7527:
1.227 brouard 7528: /* stepsize=(int) (stepm+YEARM-1)/YEARM; */
7529: /* if (stepm<=12) stepsize=1; */
1.126 brouard 7530:
1.227 brouard 7531: /* agelim=AGESUP; */
1.126 brouard 7532:
1.227 brouard 7533: /* hstepm=1; */
7534: /* hstepm=hstepm/stepm; */
1.218 brouard 7535:
1.227 brouard 7536: /* if (popforecast==1) { */
7537: /* if((ficpop=fopen(popfile,"r"))==NULL) { */
7538: /* printf("Problem with population file : %s\n",popfile);exit(0); */
7539: /* fprintf(ficlog,"Problem with population file : %s\n",popfile);exit(0); */
7540: /* } */
7541: /* popage=ivector(0,AGESUP); */
7542: /* popeffectif=vector(0,AGESUP); */
7543: /* popcount=vector(0,AGESUP); */
1.126 brouard 7544:
1.227 brouard 7545: /* i=1; */
7546: /* while ((c=fscanf(ficpop,"%d %lf\n",&popage[i],&popcount[i])) != EOF) i=i+1; */
1.218 brouard 7547:
1.227 brouard 7548: /* imx=i; */
7549: /* for (i=1; i<imx;i++) popeffectif[popage[i]]=popcount[i]; */
7550: /* } */
1.218 brouard 7551:
1.227 brouard 7552: /* for(cptcov=1,k=0;cptcov<=i2;cptcov++){ */
7553: /* for(cptcod=1;cptcod<=ncodemax[cptcoveff];cptcod++){ */
7554: /* k=k+1; */
7555: /* fprintf(ficrespop,"\n#******"); */
7556: /* for(j=1;j<=cptcoveff;j++) { */
7557: /* fprintf(ficrespop," V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]); */
7558: /* } */
7559: /* fprintf(ficrespop,"******\n"); */
7560: /* fprintf(ficrespop,"# Age"); */
7561: /* for(j=1; j<=nlstate+ndeath;j++) fprintf(ficrespop," P.%d",j); */
7562: /* if (popforecast==1) fprintf(ficrespop," [Population]"); */
1.126 brouard 7563:
1.227 brouard 7564: /* for (cpt=0; cpt<=0;cpt++) { */
7565: /* fprintf(ficrespop,"\n\n# Forecasting at date %.lf/%.lf/%.lf ",jpyram,mpyram,anpyram+cpt); */
1.126 brouard 7566:
1.227 brouard 7567: /* for (agedeb=(fage-((int)calagedatem %12/12.)); agedeb>=(ageminpar-((int)calagedatem %12)/12.); agedeb--){ */
7568: /* nhstepm=(int) rint((agelim-agedeb)*YEARM/stepm); */
7569: /* nhstepm = nhstepm/hstepm; */
1.126 brouard 7570:
1.227 brouard 7571: /* p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm); */
7572: /* oldm=oldms;savm=savms; */
7573: /* hpxij(p3mat,nhstepm,agedeb,hstepm,p,nlstate,stepm,oldm,savm, k); */
1.218 brouard 7574:
1.227 brouard 7575: /* for (h=0; h<=nhstepm; h++){ */
7576: /* if (h==(int) (calagedatem+YEARM*cpt)) { */
7577: /* fprintf(ficrespop,"\n %3.f ",agedeb+h*hstepm/YEARM*stepm); */
7578: /* } */
7579: /* for(j=1; j<=nlstate+ndeath;j++) { */
7580: /* kk1=0.;kk2=0; */
7581: /* for(i=1; i<=nlstate;i++) { */
7582: /* if (mobilav==1) */
7583: /* kk1=kk1+p3mat[i][j][h]*mobaverage[(int)agedeb+1][i][cptcod]; */
7584: /* else { */
7585: /* kk1=kk1+p3mat[i][j][h]*probs[(int)(agedeb+1)][i][cptcod]; */
7586: /* } */
7587: /* } */
7588: /* if (h==(int)(calagedatem+12*cpt)){ */
7589: /* tabpop[(int)(agedeb)][j][cptcod]=kk1; */
7590: /* /\*fprintf(ficrespop," %.3f", kk1); */
7591: /* if (popforecast==1) fprintf(ficrespop," [%.f]", kk1*popeffectif[(int)agedeb+1]);*\/ */
7592: /* } */
7593: /* } */
7594: /* for(i=1; i<=nlstate;i++){ */
7595: /* kk1=0.; */
7596: /* for(j=1; j<=nlstate;j++){ */
7597: /* kk1= kk1+tabpop[(int)(agedeb)][j][cptcod]; */
7598: /* } */
7599: /* tabpopprev[(int)(agedeb)][i][cptcod]=tabpop[(int)(agedeb)][i][cptcod]/kk1*popeffectif[(int)(agedeb+(calagedatem+12*cpt)*hstepm/YEARM*stepm-1)]; */
7600: /* } */
1.218 brouard 7601:
1.227 brouard 7602: /* if (h==(int)(calagedatem+12*cpt)) */
7603: /* for(j=1; j<=nlstate;j++) */
7604: /* fprintf(ficrespop," %15.2f",tabpopprev[(int)(agedeb+1)][j][cptcod]); */
7605: /* } */
7606: /* free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm); */
7607: /* } */
7608: /* } */
1.218 brouard 7609:
1.227 brouard 7610: /* /\******\/ */
1.218 brouard 7611:
1.227 brouard 7612: /* for (cpt=1; cpt<=(anpyram1-anpyram);cpt++) { */
7613: /* fprintf(ficrespop,"\n\n# Forecasting at date %.lf/%.lf/%.lf ",jpyram,mpyram,anpyram+cpt); */
7614: /* for (agedeb=(fage-((int)calagedatem %12/12.)); agedeb>=(ageminpar-((int)calagedatem %12)/12.); agedeb--){ */
7615: /* nhstepm=(int) rint((agelim-agedeb)*YEARM/stepm); */
7616: /* nhstepm = nhstepm/hstepm; */
1.126 brouard 7617:
1.227 brouard 7618: /* p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm); */
7619: /* oldm=oldms;savm=savms; */
7620: /* hpxij(p3mat,nhstepm,agedeb,hstepm,p,nlstate,stepm,oldm,savm, k); */
7621: /* for (h=0; h<=nhstepm; h++){ */
7622: /* if (h==(int) (calagedatem+YEARM*cpt)) { */
7623: /* fprintf(ficresf,"\n %3.f ",agedeb+h*hstepm/YEARM*stepm); */
7624: /* } */
7625: /* for(j=1; j<=nlstate+ndeath;j++) { */
7626: /* kk1=0.;kk2=0; */
7627: /* for(i=1; i<=nlstate;i++) { */
7628: /* kk1=kk1+p3mat[i][j][h]*tabpopprev[(int)agedeb+1][i][cptcod]; */
7629: /* } */
7630: /* if (h==(int)(calagedatem+12*cpt)) fprintf(ficresf," %15.2f", kk1); */
7631: /* } */
7632: /* } */
7633: /* free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm); */
7634: /* } */
7635: /* } */
7636: /* } */
7637: /* } */
1.218 brouard 7638:
1.227 brouard 7639: /* /\* if (mobilav!=0) free_ma3x(mobaverage,1, AGESUP,1,NCOVMAX, 1,NCOVMAX); *\/ */
1.218 brouard 7640:
1.227 brouard 7641: /* if (popforecast==1) { */
7642: /* free_ivector(popage,0,AGESUP); */
7643: /* free_vector(popeffectif,0,AGESUP); */
7644: /* free_vector(popcount,0,AGESUP); */
7645: /* } */
7646: /* free_ma3x(tabpop,1, AGESUP,1,NCOVMAX, 1,NCOVMAX); */
7647: /* free_ma3x(tabpopprev,1, AGESUP,1,NCOVMAX, 1,NCOVMAX); */
7648: /* fclose(ficrespop); */
7649: /* } /\* End of popforecast *\/ */
1.218 brouard 7650:
1.126 brouard 7651: int fileappend(FILE *fichier, char *optionfich)
7652: {
7653: if((fichier=fopen(optionfich,"a"))==NULL) {
7654: printf("Problem with file: %s\n", optionfich);
7655: fprintf(ficlog,"Problem with file: %s\n", optionfich);
7656: return (0);
7657: }
7658: fflush(fichier);
7659: return (1);
7660: }
7661:
7662:
7663: /**************** function prwizard **********************/
7664: void prwizard(int ncovmodel, int nlstate, int ndeath, char model[], FILE *ficparo)
7665: {
7666:
7667: /* Wizard to print covariance matrix template */
7668:
1.164 brouard 7669: char ca[32], cb[32];
7670: int i,j, k, li, lj, lk, ll, jj, npar, itimes;
1.126 brouard 7671: int numlinepar;
7672:
7673: printf("# Parameters nlstate*nlstate*ncov a12*1 + b12 * age + ...\n");
7674: fprintf(ficparo,"# Parameters nlstate*nlstate*ncov a12*1 + b12 * age + ...\n");
7675: for(i=1; i <=nlstate; i++){
7676: jj=0;
7677: for(j=1; j <=nlstate+ndeath; j++){
7678: if(j==i) continue;
7679: jj++;
7680: /*ca[0]= k+'a'-1;ca[1]='\0';*/
7681: printf("%1d%1d",i,j);
7682: fprintf(ficparo,"%1d%1d",i,j);
7683: for(k=1; k<=ncovmodel;k++){
7684: /* printf(" %lf",param[i][j][k]); */
7685: /* fprintf(ficparo," %lf",param[i][j][k]); */
7686: printf(" 0.");
7687: fprintf(ficparo," 0.");
7688: }
7689: printf("\n");
7690: fprintf(ficparo,"\n");
7691: }
7692: }
7693: printf("# Scales (for hessian or gradient estimation)\n");
7694: fprintf(ficparo,"# Scales (for hessian or gradient estimation)\n");
7695: npar= (nlstate+ndeath-1)*nlstate*ncovmodel; /* Number of parameters*/
7696: for(i=1; i <=nlstate; i++){
7697: jj=0;
7698: for(j=1; j <=nlstate+ndeath; j++){
7699: if(j==i) continue;
7700: jj++;
7701: fprintf(ficparo,"%1d%1d",i,j);
7702: printf("%1d%1d",i,j);
7703: fflush(stdout);
7704: for(k=1; k<=ncovmodel;k++){
7705: /* printf(" %le",delti3[i][j][k]); */
7706: /* fprintf(ficparo," %le",delti3[i][j][k]); */
7707: printf(" 0.");
7708: fprintf(ficparo," 0.");
7709: }
7710: numlinepar++;
7711: printf("\n");
7712: fprintf(ficparo,"\n");
7713: }
7714: }
7715: printf("# Covariance matrix\n");
7716: /* # 121 Var(a12)\n\ */
7717: /* # 122 Cov(b12,a12) Var(b12)\n\ */
7718: /* # 131 Cov(a13,a12) Cov(a13,b12, Var(a13)\n\ */
7719: /* # 132 Cov(b13,a12) Cov(b13,b12, Cov(b13,a13) Var(b13)\n\ */
7720: /* # 212 Cov(a21,a12) Cov(a21,b12, Cov(a21,a13) Cov(a21,b13) Var(a21)\n\ */
7721: /* # 212 Cov(b21,a12) Cov(b21,b12, Cov(b21,a13) Cov(b21,b13) Cov(b21,a21) Var(b21)\n\ */
7722: /* # 232 Cov(a23,a12) Cov(a23,b12, Cov(a23,a13) Cov(a23,b13) Cov(a23,a21) Cov(a23,b21) Var(a23)\n\ */
7723: /* # 232 Cov(b23,a12) Cov(b23,b12) ... Var (b23)\n" */
7724: fflush(stdout);
7725: fprintf(ficparo,"# Covariance matrix\n");
7726: /* # 121 Var(a12)\n\ */
7727: /* # 122 Cov(b12,a12) Var(b12)\n\ */
7728: /* # ...\n\ */
7729: /* # 232 Cov(b23,a12) Cov(b23,b12) ... Var (b23)\n" */
7730:
7731: for(itimes=1;itimes<=2;itimes++){
7732: jj=0;
7733: for(i=1; i <=nlstate; i++){
7734: for(j=1; j <=nlstate+ndeath; j++){
7735: if(j==i) continue;
7736: for(k=1; k<=ncovmodel;k++){
7737: jj++;
7738: ca[0]= k+'a'-1;ca[1]='\0';
7739: if(itimes==1){
7740: printf("#%1d%1d%d",i,j,k);
7741: fprintf(ficparo,"#%1d%1d%d",i,j,k);
7742: }else{
7743: printf("%1d%1d%d",i,j,k);
7744: fprintf(ficparo,"%1d%1d%d",i,j,k);
7745: /* printf(" %.5le",matcov[i][j]); */
7746: }
7747: ll=0;
7748: for(li=1;li <=nlstate; li++){
7749: for(lj=1;lj <=nlstate+ndeath; lj++){
7750: if(lj==li) continue;
7751: for(lk=1;lk<=ncovmodel;lk++){
7752: ll++;
7753: if(ll<=jj){
7754: cb[0]= lk +'a'-1;cb[1]='\0';
7755: if(ll<jj){
7756: if(itimes==1){
7757: printf(" Cov(%s%1d%1d,%s%1d%1d)",ca,i,j,cb, li,lj);
7758: fprintf(ficparo," Cov(%s%1d%1d,%s%1d%1d)",ca,i,j,cb, li,lj);
7759: }else{
7760: printf(" 0.");
7761: fprintf(ficparo," 0.");
7762: }
7763: }else{
7764: if(itimes==1){
7765: printf(" Var(%s%1d%1d)",ca,i,j);
7766: fprintf(ficparo," Var(%s%1d%1d)",ca,i,j);
7767: }else{
7768: printf(" 0.");
7769: fprintf(ficparo," 0.");
7770: }
7771: }
7772: }
7773: } /* end lk */
7774: } /* end lj */
7775: } /* end li */
7776: printf("\n");
7777: fprintf(ficparo,"\n");
7778: numlinepar++;
7779: } /* end k*/
7780: } /*end j */
7781: } /* end i */
7782: } /* end itimes */
7783:
7784: } /* end of prwizard */
7785: /******************* Gompertz Likelihood ******************************/
7786: double gompertz(double x[])
7787: {
7788: double A,B,L=0.0,sump=0.,num=0.;
7789: int i,n=0; /* n is the size of the sample */
7790:
1.220 brouard 7791: for (i=1;i<=imx ; i++) {
1.126 brouard 7792: sump=sump+weight[i];
7793: /* sump=sump+1;*/
7794: num=num+1;
7795: }
7796:
7797:
7798: /* for (i=0; i<=imx; i++)
7799: 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]);*/
7800:
7801: for (i=1;i<=imx ; i++)
7802: {
7803: if (cens[i] == 1 && wav[i]>1)
7804: A=-x[1]/(x[2])*(exp(x[2]*(agecens[i]-agegomp))-exp(x[2]*(ageexmed[i]-agegomp)));
7805:
7806: if (cens[i] == 0 && wav[i]>1)
7807: A=-x[1]/(x[2])*(exp(x[2]*(agedc[i]-agegomp))-exp(x[2]*(ageexmed[i]-agegomp)))
7808: +log(x[1]/YEARM)+x[2]*(agedc[i]-agegomp)+log(YEARM);
7809:
7810: /*if (wav[i] > 1 && agecens[i] > 15) {*/ /* ??? */
7811: if (wav[i] > 1 ) { /* ??? */
7812: L=L+A*weight[i];
7813: /* 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]);*/
7814: }
7815: }
7816:
7817: /*printf("x1=%2.9f x2=%2.9f x3=%2.9f L=%f\n",x[1],x[2],x[3],L);*/
7818:
7819: return -2*L*num/sump;
7820: }
7821:
1.136 brouard 7822: #ifdef GSL
7823: /******************* Gompertz_f Likelihood ******************************/
7824: double gompertz_f(const gsl_vector *v, void *params)
7825: {
7826: double A,B,LL=0.0,sump=0.,num=0.;
7827: double *x= (double *) v->data;
7828: int i,n=0; /* n is the size of the sample */
7829:
7830: for (i=0;i<=imx-1 ; i++) {
7831: sump=sump+weight[i];
7832: /* sump=sump+1;*/
7833: num=num+1;
7834: }
7835:
7836:
7837: /* for (i=0; i<=imx; i++)
7838: 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]);*/
7839: printf("x[0]=%lf x[1]=%lf\n",x[0],x[1]);
7840: for (i=1;i<=imx ; i++)
7841: {
7842: if (cens[i] == 1 && wav[i]>1)
7843: A=-x[0]/(x[1])*(exp(x[1]*(agecens[i]-agegomp))-exp(x[1]*(ageexmed[i]-agegomp)));
7844:
7845: if (cens[i] == 0 && wav[i]>1)
7846: A=-x[0]/(x[1])*(exp(x[1]*(agedc[i]-agegomp))-exp(x[1]*(ageexmed[i]-agegomp)))
7847: +log(x[0]/YEARM)+x[1]*(agedc[i]-agegomp)+log(YEARM);
7848:
7849: /*if (wav[i] > 1 && agecens[i] > 15) {*/ /* ??? */
7850: if (wav[i] > 1 ) { /* ??? */
7851: LL=LL+A*weight[i];
7852: /* 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]);*/
7853: }
7854: }
7855:
7856: /*printf("x1=%2.9f x2=%2.9f x3=%2.9f L=%f\n",x[1],x[2],x[3],L);*/
7857: printf("x[0]=%lf x[1]=%lf -2*LL*num/sump=%lf\n",x[0],x[1],-2*LL*num/sump);
7858:
7859: return -2*LL*num/sump;
7860: }
7861: #endif
7862:
1.126 brouard 7863: /******************* Printing html file ***********/
1.201 brouard 7864: void printinghtmlmort(char fileresu[], char title[], char datafile[], int firstpass, \
1.126 brouard 7865: int lastpass, int stepm, int weightopt, char model[],\
7866: int imx, double p[],double **matcov,double agemortsup){
7867: int i,k;
7868:
7869: fprintf(fichtm,"<ul><li><h4>Result files </h4>\n Force of mortality. Parameters of the Gompertz fit (with confidence interval in brackets):<br>");
7870: fprintf(fichtm," mu(age) =%lf*exp(%lf*(age-%d)) per year<br><br>",p[1],p[2],agegomp);
7871: for (i=1;i<=2;i++)
7872: 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 7873: fprintf(fichtm,"<br><br><img src=\"graphmort.svg\">");
1.126 brouard 7874: fprintf(fichtm,"</ul>");
7875:
7876: fprintf(fichtm,"<ul><li><h4>Life table</h4>\n <br>");
7877:
7878: 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>");
7879:
7880: for (k=agegomp;k<(agemortsup-2);k++)
7881: 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]);
7882:
7883:
7884: fflush(fichtm);
7885: }
7886:
7887: /******************* Gnuplot file **************/
1.201 brouard 7888: void printinggnuplotmort(char fileresu[], char optionfilefiname[], double ageminpar, double agemaxpar, double fage , char pathc[], double p[]){
1.126 brouard 7889:
7890: char dirfileres[132],optfileres[132];
1.164 brouard 7891:
1.126 brouard 7892: int ng;
7893:
7894:
7895: /*#ifdef windows */
7896: fprintf(ficgp,"cd \"%s\" \n",pathc);
7897: /*#endif */
7898:
7899:
7900: strcpy(dirfileres,optionfilefiname);
7901: strcpy(optfileres,"vpl");
1.199 brouard 7902: fprintf(ficgp,"set out \"graphmort.svg\"\n ");
1.126 brouard 7903: fprintf(ficgp,"set xlabel \"Age\"\n set ylabel \"Force of mortality (per year)\" \n ");
1.199 brouard 7904: fprintf(ficgp, "set ter svg size 640, 480\n set log y\n");
1.145 brouard 7905: /* fprintf(ficgp, "set size 0.65,0.65\n"); */
1.126 brouard 7906: fprintf(ficgp,"plot [%d:100] %lf*exp(%lf*(x-%d))",agegomp,p[1],p[2],agegomp);
7907:
7908: }
7909:
1.136 brouard 7910: int readdata(char datafile[], int firstobs, int lastobs, int *imax)
7911: {
1.126 brouard 7912:
1.136 brouard 7913: /*-------- data file ----------*/
7914: FILE *fic;
7915: char dummy[]=" ";
1.240 brouard 7916: int i=0, j=0, n=0, iv=0, v;
1.223 brouard 7917: int lstra;
1.136 brouard 7918: int linei, month, year,iout;
7919: char line[MAXLINE], linetmp[MAXLINE];
1.164 brouard 7920: char stra[MAXLINE], strb[MAXLINE];
1.136 brouard 7921: char *stratrunc;
1.223 brouard 7922:
1.240 brouard 7923: DummyV=ivector(1,NCOVMAX); /* 1 to 3 */
7924: FixedV=ivector(1,NCOVMAX); /* 1 to 3 */
1.126 brouard 7925:
1.240 brouard 7926: for(v=1; v <=ncovcol;v++){
7927: DummyV[v]=0;
7928: FixedV[v]=0;
7929: }
7930: for(v=ncovcol+1; v <=ncovcol+nqv;v++){
7931: DummyV[v]=1;
7932: FixedV[v]=0;
7933: }
7934: for(v=ncovcol+nqv+1; v <=ncovcol+nqv+ntv;v++){
7935: DummyV[v]=0;
7936: FixedV[v]=1;
7937: }
7938: for(v=ncovcol+nqv+ntv+1; v <=ncovcol+nqv+ntv+nqtv;v++){
7939: DummyV[v]=1;
7940: FixedV[v]=1;
7941: }
7942: for(v=1; v <=ncovcol+nqv+ntv+nqtv;v++){
7943: printf("Covariate type in the data: V%d, DummyV(V%d)=%d, FixedV(V%d)=%d\n",v,v,DummyV[v],v,FixedV[v]);
7944: 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]);
7945: }
1.126 brouard 7946:
1.136 brouard 7947: if((fic=fopen(datafile,"r"))==NULL) {
1.218 brouard 7948: printf("Problem while opening datafile: %s with errno='%s'\n", datafile,strerror(errno));fflush(stdout);
7949: fprintf(ficlog,"Problem while opening datafile: %s with errno='%s'\n", datafile,strerror(errno));fflush(ficlog);return 1;
1.136 brouard 7950: }
1.126 brouard 7951:
1.136 brouard 7952: i=1;
7953: linei=0;
7954: while ((fgets(line, MAXLINE, fic) != NULL) &&((i >= firstobs) && (i <=lastobs))) {
7955: linei=linei+1;
7956: for(j=strlen(line); j>=0;j--){ /* Untabifies line */
7957: if(line[j] == '\t')
7958: line[j] = ' ';
7959: }
7960: for(j=strlen(line)-1; (line[j]==' ')||(line[j]==10)||(line[j]==13);j--){
7961: ;
7962: };
7963: line[j+1]=0; /* Trims blanks at end of line */
7964: if(line[0]=='#'){
7965: fprintf(ficlog,"Comment line\n%s\n",line);
7966: printf("Comment line\n%s\n",line);
7967: continue;
7968: }
7969: trimbb(linetmp,line); /* Trims multiple blanks in line */
1.164 brouard 7970: strcpy(line, linetmp);
1.223 brouard 7971:
7972: /* Loops on waves */
7973: for (j=maxwav;j>=1;j--){
7974: for (iv=nqtv;iv>=1;iv--){ /* Loop on time varying quantitative variables */
1.238 brouard 7975: cutv(stra, strb, line, ' ');
7976: if(strb[0]=='.') { /* Missing value */
7977: lval=-1;
7978: cotqvar[j][iv][i]=-1; /* 0.0/0.0 */
7979: cotvar[j][ntv+iv][i]=-1; /* For performance reasons */
7980: if(isalpha(strb[1])) { /* .m or .d Really Missing value */
7981: 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);
7982: 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);
7983: return 1;
7984: }
7985: }else{
7986: errno=0;
7987: /* what_kind_of_number(strb); */
7988: dval=strtod(strb,&endptr);
7989: /* if( strb[0]=='\0' || (*endptr != '\0')){ */
7990: /* if(strb != endptr && *endptr == '\0') */
7991: /* dval=dlval; */
7992: /* if (errno == ERANGE && (lval == LONG_MAX || lval == LONG_MIN)) */
7993: if( strb[0]=='\0' || (*endptr != '\0')){
7994: 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);
7995: 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);
7996: return 1;
7997: }
7998: cotqvar[j][iv][i]=dval;
7999: cotvar[j][ntv+iv][i]=dval;
8000: }
8001: strcpy(line,stra);
1.223 brouard 8002: }/* end loop ntqv */
1.225 brouard 8003:
1.223 brouard 8004: for (iv=ntv;iv>=1;iv--){ /* Loop on time varying dummies */
1.238 brouard 8005: cutv(stra, strb, line, ' ');
8006: if(strb[0]=='.') { /* Missing value */
8007: lval=-1;
8008: }else{
8009: errno=0;
8010: lval=strtol(strb,&endptr,10);
8011: /* if (errno == ERANGE && (lval == LONG_MAX || lval == LONG_MIN))*/
8012: if( strb[0]=='\0' || (*endptr != '\0')){
8013: 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);
8014: 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);
8015: return 1;
8016: }
8017: }
8018: if(lval <-1 || lval >1){
8019: printf("Error reading data around '%ld' at line number %d for individual %d, '%s'\n \
1.223 brouard 8020: Should be a value of %d(nth) covariate (0 should be the value for the reference and 1\n \
8021: for the alternative. IMaCh does not build design variables automatically, do it yourself.\n \
1.238 brouard 8022: For example, for multinomial values like 1, 2 and 3,\n \
8023: build V1=0 V2=0 for the reference value (1),\n \
8024: V1=1 V2=0 for (2) \n \
1.223 brouard 8025: and V1=0 V2=1 for (3). V1=1 V2=1 should not exist and the corresponding\n \
1.238 brouard 8026: output of IMaCh is often meaningless.\n \
1.223 brouard 8027: Exiting.\n",lval,linei, i,line,j);
1.238 brouard 8028: fprintf(ficlog,"Error reading data around '%ld' at line number %d for individual %d, '%s'\n \
1.223 brouard 8029: Should be a value of %d(nth) covariate (0 should be the value for the reference and 1\n \
8030: for the alternative. IMaCh does not build design variables automatically, do it yourself.\n \
1.238 brouard 8031: For example, for multinomial values like 1, 2 and 3,\n \
8032: build V1=0 V2=0 for the reference value (1),\n \
8033: V1=1 V2=0 for (2) \n \
1.223 brouard 8034: and V1=0 V2=1 for (3). V1=1 V2=1 should not exist and the corresponding\n \
1.238 brouard 8035: output of IMaCh is often meaningless.\n \
1.223 brouard 8036: Exiting.\n",lval,linei, i,line,j);fflush(ficlog);
1.238 brouard 8037: return 1;
8038: }
8039: cotvar[j][iv][i]=(double)(lval);
8040: strcpy(line,stra);
1.223 brouard 8041: }/* end loop ntv */
1.225 brouard 8042:
1.223 brouard 8043: /* Statuses at wave */
1.137 brouard 8044: cutv(stra, strb, line, ' ');
1.223 brouard 8045: if(strb[0]=='.') { /* Missing value */
1.238 brouard 8046: lval=-1;
1.136 brouard 8047: }else{
1.238 brouard 8048: errno=0;
8049: lval=strtol(strb,&endptr,10);
8050: /* if (errno == ERANGE && (lval == LONG_MAX || lval == LONG_MIN))*/
8051: if( strb[0]=='\0' || (*endptr != '\0')){
8052: 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);
8053: 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);
8054: return 1;
8055: }
1.136 brouard 8056: }
1.225 brouard 8057:
1.136 brouard 8058: s[j][i]=lval;
1.225 brouard 8059:
1.223 brouard 8060: /* Date of Interview */
1.136 brouard 8061: strcpy(line,stra);
8062: cutv(stra, strb,line,' ');
1.169 brouard 8063: if( (iout=sscanf(strb,"%d/%d",&month, &year)) != 0){
1.136 brouard 8064: }
1.169 brouard 8065: else if( (iout=sscanf(strb,"%s.",dummy)) != 0){
1.225 brouard 8066: month=99;
8067: year=9999;
1.136 brouard 8068: }else{
1.225 brouard 8069: 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);
8070: 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);
8071: return 1;
1.136 brouard 8072: }
8073: anint[j][i]= (double) year;
8074: mint[j][i]= (double)month;
8075: strcpy(line,stra);
1.223 brouard 8076: } /* End loop on waves */
1.225 brouard 8077:
1.223 brouard 8078: /* Date of death */
1.136 brouard 8079: cutv(stra, strb,line,' ');
1.169 brouard 8080: if( (iout=sscanf(strb,"%d/%d",&month, &year)) != 0){
1.136 brouard 8081: }
1.169 brouard 8082: else if( (iout=sscanf(strb,"%s.",dummy)) != 0){
1.136 brouard 8083: month=99;
8084: year=9999;
8085: }else{
1.141 brouard 8086: 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 8087: 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);
8088: return 1;
1.136 brouard 8089: }
8090: andc[i]=(double) year;
8091: moisdc[i]=(double) month;
8092: strcpy(line,stra);
8093:
1.223 brouard 8094: /* Date of birth */
1.136 brouard 8095: cutv(stra, strb,line,' ');
1.169 brouard 8096: if( (iout=sscanf(strb,"%d/%d",&month, &year)) != 0){
1.136 brouard 8097: }
1.169 brouard 8098: else if( (iout=sscanf(strb,"%s.", dummy)) != 0){
1.136 brouard 8099: month=99;
8100: year=9999;
8101: }else{
1.141 brouard 8102: 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);
8103: 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 8104: return 1;
1.136 brouard 8105: }
8106: if (year==9999) {
1.141 brouard 8107: 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);
8108: 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 8109: return 1;
8110:
1.136 brouard 8111: }
8112: annais[i]=(double)(year);
8113: moisnais[i]=(double)(month);
8114: strcpy(line,stra);
1.225 brouard 8115:
1.223 brouard 8116: /* Sample weight */
1.136 brouard 8117: cutv(stra, strb,line,' ');
8118: errno=0;
8119: dval=strtod(strb,&endptr);
8120: if( strb[0]=='\0' || (*endptr != '\0')){
1.141 brouard 8121: printf("Error reading data around '%f' at line number %d, \"%s\" for individual %d\nShould be a weight. Exiting.\n",dval, i,line,linei);
8122: 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 8123: fflush(ficlog);
8124: return 1;
8125: }
8126: weight[i]=dval;
8127: strcpy(line,stra);
1.225 brouard 8128:
1.223 brouard 8129: for (iv=nqv;iv>=1;iv--){ /* Loop on fixed quantitative variables */
8130: cutv(stra, strb, line, ' ');
8131: if(strb[0]=='.') { /* Missing value */
1.225 brouard 8132: lval=-1;
1.223 brouard 8133: }else{
1.225 brouard 8134: errno=0;
8135: /* what_kind_of_number(strb); */
8136: dval=strtod(strb,&endptr);
8137: /* if(strb != endptr && *endptr == '\0') */
8138: /* dval=dlval; */
8139: /* if (errno == ERANGE && (lval == LONG_MAX || lval == LONG_MIN)) */
8140: if( strb[0]=='\0' || (*endptr != '\0')){
8141: 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);
8142: 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);
8143: return 1;
8144: }
8145: coqvar[iv][i]=dval;
1.226 brouard 8146: covar[ncovcol+iv][i]=dval; /* including qvar in standard covar for performance reasons */
1.223 brouard 8147: }
8148: strcpy(line,stra);
8149: }/* end loop nqv */
1.136 brouard 8150:
1.223 brouard 8151: /* Covariate values */
1.136 brouard 8152: for (j=ncovcol;j>=1;j--){
8153: cutv(stra, strb,line,' ');
1.223 brouard 8154: if(strb[0]=='.') { /* Missing covariate value */
1.225 brouard 8155: lval=-1;
1.136 brouard 8156: }else{
1.225 brouard 8157: errno=0;
8158: lval=strtol(strb,&endptr,10);
8159: if( strb[0]=='\0' || (*endptr != '\0')){
8160: 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);
8161: 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);
8162: return 1;
8163: }
1.136 brouard 8164: }
8165: if(lval <-1 || lval >1){
1.225 brouard 8166: printf("Error reading data around '%ld' at line number %d for individual %d, '%s'\n \
1.136 brouard 8167: Should be a value of %d(nth) covariate (0 should be the value for the reference and 1\n \
8168: for the alternative. IMaCh does not build design variables automatically, do it yourself.\n \
1.225 brouard 8169: For example, for multinomial values like 1, 2 and 3,\n \
8170: build V1=0 V2=0 for the reference value (1),\n \
8171: V1=1 V2=0 for (2) \n \
1.136 brouard 8172: and V1=0 V2=1 for (3). V1=1 V2=1 should not exist and the corresponding\n \
1.225 brouard 8173: output of IMaCh is often meaningless.\n \
1.136 brouard 8174: Exiting.\n",lval,linei, i,line,j);
1.225 brouard 8175: fprintf(ficlog,"Error reading data around '%ld' at line number %d for individual %d, '%s'\n \
1.136 brouard 8176: Should be a value of %d(nth) covariate (0 should be the value for the reference and 1\n \
8177: for the alternative. IMaCh does not build design variables automatically, do it yourself.\n \
1.225 brouard 8178: For example, for multinomial values like 1, 2 and 3,\n \
8179: build V1=0 V2=0 for the reference value (1),\n \
8180: V1=1 V2=0 for (2) \n \
1.136 brouard 8181: and V1=0 V2=1 for (3). V1=1 V2=1 should not exist and the corresponding\n \
1.225 brouard 8182: output of IMaCh is often meaningless.\n \
1.136 brouard 8183: Exiting.\n",lval,linei, i,line,j);fflush(ficlog);
1.225 brouard 8184: return 1;
1.136 brouard 8185: }
8186: covar[j][i]=(double)(lval);
8187: strcpy(line,stra);
8188: }
8189: lstra=strlen(stra);
1.225 brouard 8190:
1.136 brouard 8191: if(lstra > 9){ /* More than 2**32 or max of what printf can write with %ld */
8192: stratrunc = &(stra[lstra-9]);
8193: num[i]=atol(stratrunc);
8194: }
8195: else
8196: num[i]=atol(stra);
8197: /*if((s[2][i]==2) && (s[3][i]==-1)&&(s[4][i]==9)){
8198: 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;}*/
8199:
8200: i=i+1;
8201: } /* End loop reading data */
1.225 brouard 8202:
1.136 brouard 8203: *imax=i-1; /* Number of individuals */
8204: fclose(fic);
1.225 brouard 8205:
1.136 brouard 8206: return (0);
1.164 brouard 8207: /* endread: */
1.225 brouard 8208: printf("Exiting readdata: ");
8209: fclose(fic);
8210: return (1);
1.223 brouard 8211: }
1.126 brouard 8212:
1.234 brouard 8213: void removefirstspace(char **stri){/*, char stro[]) {*/
1.230 brouard 8214: char *p1 = *stri, *p2 = *stri;
1.235 brouard 8215: while (*p2 == ' ')
1.234 brouard 8216: p2++;
8217: /* while ((*p1++ = *p2++) !=0) */
8218: /* ; */
8219: /* do */
8220: /* while (*p2 == ' ') */
8221: /* p2++; */
8222: /* while (*p1++ == *p2++); */
8223: *stri=p2;
1.145 brouard 8224: }
8225:
1.235 brouard 8226: int decoderesult ( char resultline[], int nres)
1.230 brouard 8227: /**< This routine decode one result line and returns the combination # of dummy covariates only **/
8228: {
1.235 brouard 8229: int j=0, k=0, k1=0, k2=0, k3=0, k4=0, match=0, k2q=0, k3q=0, k4q=0;
1.230 brouard 8230: char resultsav[MAXLINE];
1.234 brouard 8231: int resultmodel[MAXLINE];
8232: int modelresult[MAXLINE];
1.230 brouard 8233: char stra[80], strb[80], strc[80], strd[80],stre[80];
8234:
1.234 brouard 8235: removefirstspace(&resultline);
1.233 brouard 8236: printf("decoderesult:%s\n",resultline);
1.230 brouard 8237:
8238: if (strstr(resultline,"v") !=0){
8239: printf("Error. 'v' must be in upper case 'V' result: %s ",resultline);
8240: fprintf(ficlog,"Error. 'v' must be in upper case result: %s ",resultline);fflush(ficlog);
8241: return 1;
8242: }
8243: trimbb(resultsav, resultline);
8244: if (strlen(resultsav) >1){
8245: j=nbocc(resultsav,'='); /**< j=Number of covariate values'=' */
8246: }
1.234 brouard 8247: if( j != cptcovs ){ /* Be careful if a variable is in a product but not single */
8248: 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);
8249: 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);
8250: }
8251: for(k=1; k<=j;k++){ /* Loop on any covariate of the result line */
8252: if(nbocc(resultsav,'=') >1){
8253: cutl(stra,strb,resultsav,' '); /* keeps in strb after the first ' '
8254: resultsav= V4=1 V5=25.1 V3=0 strb=V3=0 stra= V4=1 V5=25.1 */
8255: cutl(strc,strd,strb,'='); /* strb:V4=1 strc=1 strd=V4 */
8256: }else
8257: cutl(strc,strd,resultsav,'=');
1.230 brouard 8258: Tvalsel[k]=atof(strc); /* 1 */
1.234 brouard 8259:
1.230 brouard 8260: cutl(strc,stre,strd,'V'); /* strd='V4' strc=4 stre='V' */;
8261: Tvarsel[k]=atoi(strc);
8262: /* Typevarsel[k]=1; /\* 1 for age product *\/ */
8263: /* cptcovsel++; */
8264: if (nbocc(stra,'=') >0)
8265: strcpy(resultsav,stra); /* and analyzes it */
8266: }
1.235 brouard 8267: /* Checking for missing or useless values in comparison of current model needs */
1.236 brouard 8268: for(k1=1; k1<= cptcovt ;k1++){ /* model line V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
8269: if(Typevar[k1]==0){ /* Single covariate in model */
1.234 brouard 8270: match=0;
1.236 brouard 8271: for(k2=1; k2 <=j;k2++){/* result line V4=1 V5=24.1 V3=1 V2=8 V1=0 */
1.237 brouard 8272: if(Tvar[k1]==Tvarsel[k2]) {/* Tvar[1]=5 == Tvarsel[2]=5 */
1.236 brouard 8273: modelresult[k2]=k1;/* modelresult[2]=1 modelresult[1]=2 modelresult[3]=3 modelresult[6]=4 modelresult[9]=5 */
1.234 brouard 8274: match=1;
8275: break;
8276: }
8277: }
8278: if(match == 0){
8279: printf("Error in result line: %d value missing; result: %s, model=%s\n",k1, resultline, model);
8280: }
8281: }
8282: }
1.235 brouard 8283: /* Checking for missing or useless values in comparison of current model needs */
8284: for(k2=1; k2 <=j;k2++){ /* result line V4=1 V5=24.1 V3=1 V2=8 V1=0 */
1.234 brouard 8285: match=0;
1.235 brouard 8286: for(k1=1; k1<= cptcovt ;k1++){ /* model line V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
8287: if(Typevar[k1]==0){ /* Single */
1.237 brouard 8288: if(Tvar[k1]==Tvarsel[k2]) { /* Tvar[2]=4 == Tvarsel[1]=4 */
1.235 brouard 8289: resultmodel[k1]=k2; /* resultmodel[2]=1 resultmodel[1]=2 resultmodel[3]=3 resultmodel[6]=4 resultmodel[9]=5 */
1.234 brouard 8290: ++match;
8291: }
8292: }
8293: }
8294: if(match == 0){
8295: printf("Error in result line: %d value missing; result: %s, model=%s\n",k1, resultline, model);
8296: }else if(match > 1){
8297: printf("Error in result line: %d doubled; result: %s, model=%s\n",k2, resultline, model);
8298: }
8299: }
1.235 brouard 8300:
1.234 brouard 8301: /* We need to deduce which combination number is chosen and save quantitative values */
1.235 brouard 8302: /* model line V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
8303: /* result line V4=1 V5=25.1 V3=0 V2=8 V1=1 */
8304: /* should give a combination of dummy V4=1, V3=0, V1=1 => V4*2**(0) + V3*2**(1) + V1*2**(2) = 5 + (1offset) = 6*/
8305: /* result line V4=1 V5=24.1 V3=1 V2=8 V1=0 */
8306: /* should give a combination of dummy V4=1, V3=1, V1=0 => V4*2**(0) + V3*2**(1) + V1*2**(2) = 3 + (1offset) = 4*/
8307: /* 1 0 0 0 */
8308: /* 2 1 0 0 */
8309: /* 3 0 1 0 */
8310: /* 4 1 1 0 */ /* V4=1, V3=1, V1=0 */
8311: /* 5 0 0 1 */
8312: /* 6 1 0 1 */ /* V4=1, V3=0, V1=1 */
8313: /* 7 0 1 1 */
8314: /* 8 1 1 1 */
1.237 brouard 8315: /* V(Tvresult)=Tresult V4=1 V3=0 V1=1 Tresult[nres=1][2]=0 */
8316: /* V(Tvqresult)=Tqresult V5=25.1 V2=8 Tqresult[nres=1][1]=25.1 */
8317: /* V5*age V5 known which value for nres? */
8318: /* Tqinvresult[2]=8 Tqinvresult[1]=25.1 */
1.235 brouard 8319: for(k1=1, k=0, k4=0, k4q=0; k1 <=cptcovt;k1++){ /* model line */
8320: if( Dummy[k1]==0 && Typevar[k1]==0 ){ /* Single dummy */
1.237 brouard 8321: k3= resultmodel[k1]; /* resultmodel[2(V4)] = 1=k3 */
1.235 brouard 8322: k2=(int)Tvarsel[k3]; /* Tvarsel[resultmodel[2]]= Tvarsel[1] = 4=k2 */
8323: k+=Tvalsel[k3]*pow(2,k4); /* Tvalsel[1]=1 */
1.237 brouard 8324: Tresult[nres][k4+1]=Tvalsel[k3];/* Tresult[nres][1]=1(V4=1) Tresult[nres][2]=0(V3=0) */
8325: Tvresult[nres][k4+1]=(int)Tvarsel[k3];/* Tvresult[nres][1]=4 Tvresult[nres][3]=1 */
8326: Tinvresult[nres][(int)Tvarsel[k3]]=Tvalsel[k3]; /* Tinvresult[nres][4]=1 */
1.235 brouard 8327: printf("Decoderesult Dummy k=%d, V(k2=V%d)= Tvalsel[%d]=%d, 2**(%d)\n",k, k2, k3, (int)Tvalsel[k3], k4);
8328: k4++;;
8329: } else if( Dummy[k1]==1 && Typevar[k1]==0 ){ /* Single quantitative */
8330: k3q= resultmodel[k1]; /* resultmodel[2] = 1=k3 */
8331: k2q=(int)Tvarsel[k3q]; /* Tvarsel[resultmodel[2]]= Tvarsel[1] = 4=k2 */
1.237 brouard 8332: Tqresult[nres][k4q+1]=Tvalsel[k3q]; /* Tqresult[nres][1]=25.1 */
8333: Tvqresult[nres][k4q+1]=(int)Tvarsel[k3q]; /* Tvqresult[nres][1]=5 */
8334: Tqinvresult[nres][(int)Tvarsel[k3q]]=Tvalsel[k3q]; /* Tqinvresult[nres][5]=25.1 */
1.235 brouard 8335: printf("Decoderesult Quantitative nres=%d, V(k2q=V%d)= Tvalsel[%d]=%d, Tvarsel[%d]=%f\n",nres, k2q, k3q, Tvarsel[k3q], k3q, Tvalsel[k3q]);
8336: k4q++;;
8337: }
8338: }
1.234 brouard 8339:
1.235 brouard 8340: TKresult[nres]=++k; /* Combination for the nresult and the model */
1.230 brouard 8341: return (0);
8342: }
1.235 brouard 8343:
1.230 brouard 8344: int decodemodel( char model[], int lastobs)
8345: /**< This routine decodes the model and returns:
1.224 brouard 8346: * Model V1+V2+V3+V8+V7*V8+V5*V6+V8*age+V3*age+age*age
8347: * - nagesqr = 1 if age*age in the model, otherwise 0.
8348: * - cptcovt total number of covariates of the model nbocc(+)+1 = 8 excepting constant and age and age*age
8349: * - cptcovn or number of covariates k of the models excluding age*products =6 and age*age
8350: * - cptcovage number of covariates with age*products =2
8351: * - cptcovs number of simple covariates
8352: * - 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
8353: * which is a new column after the 9 (ncovcol) variables.
8354: * - if k is a product Vn*Vm covar[k][i] is filled with correct values for each individual
8355: * - Tprod[l] gives the kth covariates of the product Vn*Vm l=1 to cptcovprod-cptcovage
8356: * Tprod[1]@2 {5, 6}: position of first product V7*V8 is 5, and second V5*V6 is 6.
8357: * - Tvard[k] p Tvard[1][1]@4 {7, 8, 5, 6} for V7*V8 and V5*V6 .
8358: */
1.136 brouard 8359: {
1.238 brouard 8360: int i, j, k, ks, v;
1.227 brouard 8361: int j1, k1, k2, k3, k4;
1.136 brouard 8362: char modelsav[80];
1.145 brouard 8363: char stra[80], strb[80], strc[80], strd[80],stre[80];
1.187 brouard 8364: char *strpt;
1.136 brouard 8365:
1.145 brouard 8366: /*removespace(model);*/
1.136 brouard 8367: if (strlen(model) >1){ /* If there is at least 1 covariate */
1.145 brouard 8368: j=0, j1=0, k1=0, k2=-1, ks=0, cptcovn=0;
1.137 brouard 8369: if (strstr(model,"AGE") !=0){
1.192 brouard 8370: printf("Error. AGE must be in lower case 'age' model=1+age+%s. ",model);
8371: fprintf(ficlog,"Error. AGE must be in lower case model=1+age+%s. ",model);fflush(ficlog);
1.136 brouard 8372: return 1;
8373: }
1.141 brouard 8374: if (strstr(model,"v") !=0){
8375: printf("Error. 'v' must be in upper case 'V' model=%s ",model);
8376: fprintf(ficlog,"Error. 'v' must be in upper case model=%s ",model);fflush(ficlog);
8377: return 1;
8378: }
1.187 brouard 8379: strcpy(modelsav,model);
8380: if ((strpt=strstr(model,"age*age")) !=0){
8381: printf(" strpt=%s, model=%s\n",strpt, model);
8382: if(strpt != model){
1.234 brouard 8383: printf("Error in model: 'model=%s'; 'age*age' should in first place before other covariates\n \
1.192 brouard 8384: 'model=1+age+age*age+V1.' or 'model=1+age+age*age+V1+V1*age.', please swap as well as \n \
1.187 brouard 8385: corresponding column of parameters.\n",model);
1.234 brouard 8386: fprintf(ficlog,"Error in model: 'model=%s'; 'age*age' should in first place before other covariates\n \
1.192 brouard 8387: 'model=1+age+age*age+V1.' or 'model=1+age+age*age+V1+V1*age.', please swap as well as \n \
1.187 brouard 8388: corresponding column of parameters.\n",model); fflush(ficlog);
1.234 brouard 8389: return 1;
1.225 brouard 8390: }
1.187 brouard 8391: nagesqr=1;
8392: if (strstr(model,"+age*age") !=0)
1.234 brouard 8393: substrchaine(modelsav, model, "+age*age");
1.187 brouard 8394: else if (strstr(model,"age*age+") !=0)
1.234 brouard 8395: substrchaine(modelsav, model, "age*age+");
1.187 brouard 8396: else
1.234 brouard 8397: substrchaine(modelsav, model, "age*age");
1.187 brouard 8398: }else
8399: nagesqr=0;
8400: if (strlen(modelsav) >1){
8401: j=nbocc(modelsav,'+'); /**< j=Number of '+' */
8402: j1=nbocc(modelsav,'*'); /**< j1=Number of '*' */
1.224 brouard 8403: cptcovs=j+1-j1; /**< Number of simple covariates V1+V1*age+V3 +V3*V4+age*age=> V1 + V3 =5-3=2 */
1.187 brouard 8404: cptcovt= j+1; /* Number of total covariates in the model, not including
1.225 brouard 8405: * cst, age and age*age
8406: * V1+V1*age+ V3 + V3*V4+age*age=> 3+1=4*/
8407: /* including age products which are counted in cptcovage.
8408: * but the covariates which are products must be treated
8409: * separately: ncovn=4- 2=2 (V1+V3). */
1.187 brouard 8410: cptcovprod=j1; /**< Number of products V1*V2 +v3*age = 2 */
8411: cptcovprodnoage=0; /**< Number of covariate products without age: V3*V4 =1 */
1.225 brouard 8412:
8413:
1.187 brouard 8414: /* Design
8415: * V1 V2 V3 V4 V5 V6 V7 V8 V9 Weight
8416: * < ncovcol=8 >
8417: * Model V2 + V1 + V3*age + V3 + V5*V6 + V7*V8 + V8*age + V8
8418: * k= 1 2 3 4 5 6 7 8
8419: * cptcovn number of covariates (not including constant and age ) = # of + plus 1 = 7+1=8
8420: * covar[k,i], value of kth covariate if not including age for individual i:
1.224 brouard 8421: * covar[1][i]= (V1), covar[4][i]=(V4), covar[8][i]=(V8)
8422: * Tvar[k] # of the kth covariate: Tvar[1]=2 Tvar[2]=1 Tvar[4]=3 Tvar[8]=8
1.187 brouard 8423: * if multiplied by age: V3*age Tvar[3=V3*age]=3 (V3) Tvar[7]=8 and
8424: * Tage[++cptcovage]=k
8425: * if products, new covar are created after ncovcol with k1
8426: * Tvar[k]=ncovcol+k1; # of the kth covariate product: Tvar[5]=ncovcol+1=10 Tvar[6]=ncovcol+1=11
8427: * Tprod[k1]=k; Tprod[1]=5 Tprod[2]= 6; gives the position of the k1th product
8428: * 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
8429: * Tvar[cptcovn+k2]=Tvard[k1][1];Tvar[cptcovn+k2+1]=Tvard[k1][2];
8430: * Tvar[8+1]=5;Tvar[8+2]=6;Tvar[8+3]=7;Tvar[8+4]=8 inverted
8431: * V1 V2 V3 V4 V5 V6 V7 V8 V9 V10 V11
8432: * < ncovcol=8 >
8433: * Model V2 + V1 + V3*age + V3 + V5*V6 + V7*V8 + V8*age + V8 d1 d1 d2 d2
8434: * k= 1 2 3 4 5 6 7 8 9 10 11 12
8435: * Tvar[k]= 2 1 3 3 10 11 8 8 5 6 7 8
8436: * p Tvar[1]@12={2, 1, 3, 3, 11, 10, 8, 8, 7, 8, 5, 6}
8437: * p Tprod[1]@2={ 6, 5}
8438: *p Tvard[1][1]@4= {7, 8, 5, 6}
8439: * covar[k][i]= V2 V1 ? V3 V5*V6? V7*V8? ? V8
8440: * cov[Tage[kk]+2]=covar[Tvar[Tage[kk]]][i]*cov[2];
8441: *How to reorganize?
8442: * Model V1 + V2 + V3 + V8 + V5*V6 + V7*V8 + V3*age + V8*age
8443: * Tvars {2, 1, 3, 3, 11, 10, 8, 8, 7, 8, 5, 6}
8444: * {2, 1, 4, 8, 5, 6, 3, 7}
8445: * Struct []
8446: */
1.225 brouard 8447:
1.187 brouard 8448: /* This loop fills the array Tvar from the string 'model'.*/
8449: /* j is the number of + signs in the model V1+V2+V3 j=2 i=3 to 1 */
8450: /* modelsav=V2+V1+V4+age*V3 strb=age*V3 stra=V2+V1+V4 */
8451: /* k=4 (age*V3) Tvar[k=4]= 3 (from V3) Tage[cptcovage=1]=4 */
8452: /* k=3 V4 Tvar[k=3]= 4 (from V4) */
8453: /* k=2 V1 Tvar[k=2]= 1 (from V1) */
8454: /* k=1 Tvar[1]=2 (from V2) */
8455: /* k=5 Tvar[5] */
8456: /* for (k=1; k<=cptcovn;k++) { */
1.198 brouard 8457: /* cov[2+k]=nbcode[Tvar[k]][codtabm(ij,Tvar[k])]; */
1.187 brouard 8458: /* } */
1.198 brouard 8459: /* for (k=1; k<=cptcovage;k++) cov[2+Tage[k]]=nbcode[Tvar[Tage[k]]][codtabm(ij,Tvar[Tage[k])]]*cov[2]; */
1.187 brouard 8460: /*
8461: * Treating invertedly V2+V1+V3*age+V2*V4 is as if written V2*V4 +V3*age + V1 + V2 */
1.227 brouard 8462: for(k=cptcovt; k>=1;k--){ /**< Number of covariates not including constant and age, neither age*age*/
8463: Tvar[k]=0; Tprod[k]=0; Tposprod[k]=0;
8464: }
1.187 brouard 8465: cptcovage=0;
8466: for(k=1; k<=cptcovt;k++){ /* Loop on total covariates of the model */
1.234 brouard 8467: cutl(stra,strb,modelsav,'+'); /* keeps in strb after the first '+'
1.225 brouard 8468: modelsav==V2+V1+V4+V3*age strb=V3*age stra=V2+V1+V4 */
1.234 brouard 8469: if (nbocc(modelsav,'+')==0) strcpy(strb,modelsav); /* and analyzes it */
8470: /* printf("i=%d a=%s b=%s sav=%s\n",i, stra,strb,modelsav);*/
8471: /*scanf("%d",i);*/
8472: if (strchr(strb,'*')) { /**< Model includes a product V2+V1+V4+V3*age strb=V3*age */
8473: cutl(strc,strd,strb,'*'); /**< strd*strc Vm*Vn: strb=V3*age(input) strc=age strd=V3 ; V3*V2 strc=V2, strd=V3 */
8474: if (strcmp(strc,"age")==0) { /**< Model includes age: Vn*age */
8475: /* covar is not filled and then is empty */
8476: cptcovprod--;
8477: cutl(stre,strb,strd,'V'); /* strd=V3(input): stre="3" */
8478: Tvar[k]=atoi(stre); /* V2+V1+V4+V3*age Tvar[4]=3 ; V1+V2*age Tvar[2]=2; V1+V1*age Tvar[2]=1 */
8479: Typevar[k]=1; /* 1 for age product */
8480: cptcovage++; /* Sums the number of covariates which include age as a product */
8481: Tage[cptcovage]=k; /* Tvar[4]=3, Tage[1] = 4 or V1+V1*age Tvar[2]=1, Tage[1]=2 */
8482: /*printf("stre=%s ", stre);*/
8483: } else if (strcmp(strd,"age")==0) { /* or age*Vn */
8484: cptcovprod--;
8485: cutl(stre,strb,strc,'V');
8486: Tvar[k]=atoi(stre);
8487: Typevar[k]=1; /* 1 for age product */
8488: cptcovage++;
8489: Tage[cptcovage]=k;
8490: } else { /* Age is not in the model product V2+V1+V1*V4+V3*age+V3*V2 strb=V3*V2*/
8491: /* loops on k1=1 (V3*V2) and k1=2 V4*V3 */
8492: cptcovn++;
8493: cptcovprodnoage++;k1++;
8494: cutl(stre,strb,strc,'V'); /* strc= Vn, stre is n; strb=V3*V2 stre=3 strc=*/
8495: Tvar[k]=ncovcol+nqv+ntv+nqtv+k1; /* For model-covariate k tells which data-covariate to use but
8496: because this model-covariate is a construction we invent a new column
8497: which is after existing variables ncovcol+nqv+ntv+nqtv + k1
8498: If already ncovcol=4 and model=V2+V1+V1*V4+age*V3+V3*V2
8499: Tvar[3=V1*V4]=4+1 Tvar[5=V3*V2]=4 + 2= 6, etc */
8500: Typevar[k]=2; /* 2 for double fixed dummy covariates */
8501: cutl(strc,strb,strd,'V'); /* strd was Vm, strc is m */
8502: Tprod[k1]=k; /* Tprod[1]=3(=V1*V4) for V2+V1+V1*V4+age*V3+V3*V2 */
8503: Tposprod[k]=k1; /* Tpsprod[3]=1, Tposprod[2]=5 */
8504: Tvard[k1][1] =atoi(strc); /* m 1 for V1*/
8505: Tvard[k1][2] =atoi(stre); /* n 4 for V4*/
8506: k2=k2+2; /* k2 is initialize to -1, We want to store the n and m in Vn*Vm at the end of Tvar */
8507: /* Tvar[cptcovt+k2]=Tvard[k1][1]; /\* Tvar[(cptcovt=4+k2=1)=5]= 1 (V1) *\/ */
8508: /* Tvar[cptcovt+k2+1]=Tvard[k1][2]; /\* Tvar[(cptcovt=4+(k2=1)+1)=6]= 4 (V4) *\/ */
1.225 brouard 8509: /*ncovcol=4 and model=V2+V1+V1*V4+age*V3+V3*V2, Tvar[3]=5, Tvar[4]=6, cptcovt=5 */
1.234 brouard 8510: /* 1 2 3 4 5 | Tvar[5+1)=1, Tvar[7]=2 */
8511: for (i=1; i<=lastobs;i++){
8512: /* Computes the new covariate which is a product of
8513: covar[n][i]* covar[m][i] and stores it at ncovol+k1 May not be defined */
8514: covar[ncovcol+k1][i]=covar[atoi(stre)][i]*covar[atoi(strc)][i];
8515: }
8516: } /* End age is not in the model */
8517: } /* End if model includes a product */
8518: else { /* no more sum */
8519: /*printf("d=%s c=%s b=%s\n", strd,strc,strb);*/
8520: /* scanf("%d",i);*/
8521: cutl(strd,strc,strb,'V');
8522: ks++; /**< Number of simple covariates dummy or quantitative, fixe or varying */
8523: cptcovn++; /** V4+V3+V5: V4 and V3 timevarying dummy covariates, V5 timevarying quantitative */
8524: Tvar[k]=atoi(strd);
8525: Typevar[k]=0; /* 0 for simple covariates */
8526: }
8527: strcpy(modelsav,stra); /* modelsav=V2+V1+V4 stra=V2+V1+V4 */
1.223 brouard 8528: /*printf("a=%s b=%s sav=%s\n", stra,strb,modelsav);
1.225 brouard 8529: scanf("%d",i);*/
1.187 brouard 8530: } /* end of loop + on total covariates */
8531: } /* end if strlen(modelsave == 0) age*age might exist */
8532: } /* end if strlen(model == 0) */
1.136 brouard 8533:
8534: /*The number n of Vn is stored in Tvar. cptcovage =number of age covariate. Tage gives the position of age. cptcovprod= number of products.
8535: If model=V1+V1*age then Tvar[1]=1 Tvar[2]=1 cptcovage=1 Tage[1]=2 cptcovprod=0*/
1.225 brouard 8536:
1.136 brouard 8537: /* printf("tvar1=%d tvar2=%d tvar3=%d cptcovage=%d Tage=%d",Tvar[1],Tvar[2],Tvar[3],cptcovage,Tage[1]);
1.225 brouard 8538: printf("cptcovprod=%d ", cptcovprod);
8539: fprintf(ficlog,"cptcovprod=%d ", cptcovprod);
8540: scanf("%d ",i);*/
8541:
8542:
1.230 brouard 8543: /* Until here, decodemodel knows only the grammar (simple, product, age*) of the model but not what kind
8544: of variable (dummy vs quantitative, fixed vs time varying) is behind. But we know the # of each. */
1.226 brouard 8545: /* ncovcol= 1, nqv=1 | ntv=2, nqtv= 1 = 5 possible variables data: 2 fixed 3, varying
8546: model= V5 + V4 +V3 + V4*V3 + V5*age + V2 + V1*V2 + V1*age + V5*age, V1 is not used saving its place
8547: k = 1 2 3 4 5 6 7 8 9
8548: Tvar[k]= 5 4 3 1+1+2+1+1=6 5 2 7 1 5
8549: Typevar[k]= 0 0 0 2 1 0 2 1 1
1.227 brouard 8550: Fixed[k] 1 1 1 1 3 0 0 or 2 2 3
8551: Dummy[k] 1 0 0 0 3 1 1 2 3
8552: Tmodelind[combination of covar]=k;
1.225 brouard 8553: */
8554: /* Dispatching between quantitative and time varying covariates */
1.226 brouard 8555: /* If Tvar[k] >ncovcol it is a product */
1.225 brouard 8556: /* 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 8557: /* Computing effective variables, ie used by the model, that is from the cptcovt variables */
1.227 brouard 8558: printf("Model=%s\n\
8559: Typevar: 0 for simple covariate (dummy, quantitative, fixed or varying), 1 for age product, 2 for product \n\
8560: Fixed[k] 0=fixed (product or simple), 1 varying, 2 fixed with age product, 3 varying with age product \n\
8561: 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);
8562: fprintf(ficlog,"Model=%s\n\
8563: Typevar: 0 for simple covariate (dummy, quantitative, fixed or varying), 1 for age product, 2 for product \n\
8564: Fixed[k] 0=fixed (product or simple), 1 varying, 2 fixed with age product, 3 varying with age product \n\
8565: 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 8566: for(k=1;k<=cptcovt; k++){ Fixed[k]=0; Dummy[k]=0;}
1.234 brouard 8567: 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 */
8568: if (Tvar[k] <=ncovcol && Typevar[k]==0 ){ /* Simple fixed dummy (<=ncovcol) covariates */
1.227 brouard 8569: Fixed[k]= 0;
8570: Dummy[k]= 0;
1.225 brouard 8571: ncoveff++;
1.232 brouard 8572: ncovf++;
1.234 brouard 8573: nsd++;
8574: modell[k].maintype= FTYPE;
8575: TvarsD[nsd]=Tvar[k];
8576: TvarsDind[nsd]=k;
8577: TvarF[ncovf]=Tvar[k];
8578: TvarFind[ncovf]=k;
8579: TvarFD[ncoveff]=Tvar[k]; /* TvarFD[1]=V1 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
8580: TvarFDind[ncoveff]=k; /* TvarFDind[1]=9 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
8581: }else if( Tvar[k] <=ncovcol && Typevar[k]==2){ /* Product of fixed dummy (<=ncovcol) covariates */
8582: Fixed[k]= 0;
8583: Dummy[k]= 0;
8584: ncoveff++;
8585: ncovf++;
8586: modell[k].maintype= FTYPE;
8587: TvarF[ncovf]=Tvar[k];
8588: TvarFind[ncovf]=k;
1.230 brouard 8589: TvarFD[ncoveff]=Tvar[k]; /* TvarFD[1]=V1 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1.231 brouard 8590: TvarFDind[ncoveff]=k; /* TvarFDind[1]=9 in V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 */
1.240 brouard 8591: }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 8592: Fixed[k]= 0;
8593: Dummy[k]= 1;
1.230 brouard 8594: nqfveff++;
1.234 brouard 8595: modell[k].maintype= FTYPE;
8596: modell[k].subtype= FQ;
8597: nsq++;
8598: TvarsQ[nsq]=Tvar[k];
8599: TvarsQind[nsq]=k;
1.232 brouard 8600: ncovf++;
1.234 brouard 8601: TvarF[ncovf]=Tvar[k];
8602: TvarFind[ncovf]=k;
1.231 brouard 8603: 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 8604: 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 8605: }else if( Tvar[k] <=ncovcol+nqv+ntv && Typevar[k]==0){/* Only simple time varying dummy variables */
1.227 brouard 8606: Fixed[k]= 1;
8607: Dummy[k]= 0;
1.225 brouard 8608: ntveff++; /* Only simple time varying dummy variable */
1.234 brouard 8609: modell[k].maintype= VTYPE;
8610: modell[k].subtype= VD;
8611: nsd++;
8612: TvarsD[nsd]=Tvar[k];
8613: TvarsDind[nsd]=k;
8614: ncovv++; /* Only simple time varying variables */
8615: TvarV[ncovv]=Tvar[k];
1.242 brouard 8616: 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 8617: 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 */
8618: 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 8619: 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);
8620: printf("Quasi TmodelInvind[%d]=%d\n",k,Tvar[k]- ncovcol-nqv);
1.231 brouard 8621: }else if( Tvar[k] <=ncovcol+nqv+ntv+nqtv && Typevar[k]==0){ /* Only simple time varying quantitative variable V5*/
1.234 brouard 8622: Fixed[k]= 1;
8623: Dummy[k]= 1;
8624: nqtveff++;
8625: modell[k].maintype= VTYPE;
8626: modell[k].subtype= VQ;
8627: ncovv++; /* Only simple time varying variables */
8628: nsq++;
8629: TvarsQ[nsq]=Tvar[k];
8630: TvarsQind[nsq]=k;
8631: TvarV[ncovv]=Tvar[k];
1.242 brouard 8632: 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 8633: 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 */
8634: 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 8635: TmodelInvQind[nqtveff]=Tvar[k]- ncovcol-nqv-ntv;/* Only simple time varying quantitative variable */
8636: /* Tmodeliqind[k]=nqtveff;/\* Only simple time varying quantitative variable *\/ */
8637: 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 8638: printf("Quasi TmodelInvQind[%d]=%d\n",k,Tvar[k]- ncovcol-nqv-ntv);
1.227 brouard 8639: }else if (Typevar[k] == 1) { /* product with age */
1.234 brouard 8640: ncova++;
8641: TvarA[ncova]=Tvar[k];
8642: TvarAind[ncova]=k;
1.231 brouard 8643: if (Tvar[k] <=ncovcol ){ /* Product age with fixed dummy covariatee */
1.240 brouard 8644: Fixed[k]= 2;
8645: Dummy[k]= 2;
8646: modell[k].maintype= ATYPE;
8647: modell[k].subtype= APFD;
8648: /* ncoveff++; */
1.227 brouard 8649: }else if( Tvar[k] <=ncovcol+nqv) { /* Remind that product Vn*Vm are added in k*/
1.240 brouard 8650: Fixed[k]= 2;
8651: Dummy[k]= 3;
8652: modell[k].maintype= ATYPE;
8653: modell[k].subtype= APFQ; /* Product age * fixed quantitative */
8654: /* nqfveff++; /\* Only simple fixed quantitative variable *\/ */
1.227 brouard 8655: }else if( Tvar[k] <=ncovcol+nqv+ntv ){
1.240 brouard 8656: Fixed[k]= 3;
8657: Dummy[k]= 2;
8658: modell[k].maintype= ATYPE;
8659: modell[k].subtype= APVD; /* Product age * varying dummy */
8660: /* ntveff++; /\* Only simple time varying dummy variable *\/ */
1.227 brouard 8661: }else if( Tvar[k] <=ncovcol+nqv+ntv+nqtv){
1.240 brouard 8662: Fixed[k]= 3;
8663: Dummy[k]= 3;
8664: modell[k].maintype= ATYPE;
8665: modell[k].subtype= APVQ; /* Product age * varying quantitative */
8666: /* nqtveff++;/\* Only simple time varying quantitative variable *\/ */
1.227 brouard 8667: }
8668: }else if (Typevar[k] == 2) { /* product without age */
8669: k1=Tposprod[k];
8670: if(Tvard[k1][1] <=ncovcol){
1.240 brouard 8671: if(Tvard[k1][2] <=ncovcol){
8672: Fixed[k]= 1;
8673: Dummy[k]= 0;
8674: modell[k].maintype= FTYPE;
8675: modell[k].subtype= FPDD; /* Product fixed dummy * fixed dummy */
8676: ncovf++; /* Fixed variables without age */
8677: TvarF[ncovf]=Tvar[k];
8678: TvarFind[ncovf]=k;
8679: }else if(Tvard[k1][2] <=ncovcol+nqv){
8680: Fixed[k]= 0; /* or 2 ?*/
8681: Dummy[k]= 1;
8682: modell[k].maintype= FTYPE;
8683: modell[k].subtype= FPDQ; /* Product fixed dummy * fixed quantitative */
8684: ncovf++; /* Varying variables without age */
8685: TvarF[ncovf]=Tvar[k];
8686: TvarFind[ncovf]=k;
8687: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv){
8688: Fixed[k]= 1;
8689: Dummy[k]= 0;
8690: modell[k].maintype= VTYPE;
8691: modell[k].subtype= VPDD; /* Product fixed dummy * varying dummy */
8692: ncovv++; /* Varying variables without age */
8693: TvarV[ncovv]=Tvar[k];
8694: TvarVind[ncovv]=k;
8695: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv+nqtv){
8696: Fixed[k]= 1;
8697: Dummy[k]= 1;
8698: modell[k].maintype= VTYPE;
8699: modell[k].subtype= VPDQ; /* Product fixed dummy * varying quantitative */
8700: ncovv++; /* Varying variables without age */
8701: TvarV[ncovv]=Tvar[k];
8702: TvarVind[ncovv]=k;
8703: }
1.227 brouard 8704: }else if(Tvard[k1][1] <=ncovcol+nqv){
1.240 brouard 8705: if(Tvard[k1][2] <=ncovcol){
8706: Fixed[k]= 0; /* or 2 ?*/
8707: Dummy[k]= 1;
8708: modell[k].maintype= FTYPE;
8709: modell[k].subtype= FPDQ; /* Product fixed quantitative * fixed dummy */
8710: ncovf++; /* Fixed variables without age */
8711: TvarF[ncovf]=Tvar[k];
8712: TvarFind[ncovf]=k;
8713: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv){
8714: Fixed[k]= 1;
8715: Dummy[k]= 1;
8716: modell[k].maintype= VTYPE;
8717: modell[k].subtype= VPDQ; /* Product fixed quantitative * varying dummy */
8718: ncovv++; /* Varying variables without age */
8719: TvarV[ncovv]=Tvar[k];
8720: TvarVind[ncovv]=k;
8721: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv+nqtv){
8722: Fixed[k]= 1;
8723: Dummy[k]= 1;
8724: modell[k].maintype= VTYPE;
8725: modell[k].subtype= VPQQ; /* Product fixed quantitative * varying quantitative */
8726: ncovv++; /* Varying variables without age */
8727: TvarV[ncovv]=Tvar[k];
8728: TvarVind[ncovv]=k;
8729: ncovv++; /* Varying variables without age */
8730: TvarV[ncovv]=Tvar[k];
8731: TvarVind[ncovv]=k;
8732: }
1.227 brouard 8733: }else if(Tvard[k1][1] <=ncovcol+nqv+ntv){
1.240 brouard 8734: if(Tvard[k1][2] <=ncovcol){
8735: Fixed[k]= 1;
8736: Dummy[k]= 1;
8737: modell[k].maintype= VTYPE;
8738: modell[k].subtype= VPDD; /* Product time varying dummy * fixed dummy */
8739: ncovv++; /* Varying variables without age */
8740: TvarV[ncovv]=Tvar[k];
8741: TvarVind[ncovv]=k;
8742: }else if(Tvard[k1][2] <=ncovcol+nqv){
8743: Fixed[k]= 1;
8744: Dummy[k]= 1;
8745: modell[k].maintype= VTYPE;
8746: modell[k].subtype= VPDQ; /* Product time varying dummy * fixed quantitative */
8747: ncovv++; /* Varying variables without age */
8748: TvarV[ncovv]=Tvar[k];
8749: TvarVind[ncovv]=k;
8750: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv){
8751: Fixed[k]= 1;
8752: Dummy[k]= 0;
8753: modell[k].maintype= VTYPE;
8754: modell[k].subtype= VPDD; /* Product time varying dummy * time varying dummy */
8755: ncovv++; /* Varying variables without age */
8756: TvarV[ncovv]=Tvar[k];
8757: TvarVind[ncovv]=k;
8758: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv+nqtv){
8759: Fixed[k]= 1;
8760: Dummy[k]= 1;
8761: modell[k].maintype= VTYPE;
8762: modell[k].subtype= VPDQ; /* Product time varying dummy * time varying quantitative */
8763: ncovv++; /* Varying variables without age */
8764: TvarV[ncovv]=Tvar[k];
8765: TvarVind[ncovv]=k;
8766: }
1.227 brouard 8767: }else if(Tvard[k1][1] <=ncovcol+nqv+ntv+nqtv){
1.240 brouard 8768: if(Tvard[k1][2] <=ncovcol){
8769: Fixed[k]= 1;
8770: Dummy[k]= 1;
8771: modell[k].maintype= VTYPE;
8772: modell[k].subtype= VPDQ; /* Product time varying quantitative * fixed dummy */
8773: ncovv++; /* Varying variables without age */
8774: TvarV[ncovv]=Tvar[k];
8775: TvarVind[ncovv]=k;
8776: }else if(Tvard[k1][2] <=ncovcol+nqv){
8777: Fixed[k]= 1;
8778: Dummy[k]= 1;
8779: modell[k].maintype= VTYPE;
8780: modell[k].subtype= VPQQ; /* Product time varying quantitative * fixed quantitative */
8781: ncovv++; /* Varying variables without age */
8782: TvarV[ncovv]=Tvar[k];
8783: TvarVind[ncovv]=k;
8784: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv){
8785: Fixed[k]= 1;
8786: Dummy[k]= 1;
8787: modell[k].maintype= VTYPE;
8788: modell[k].subtype= VPDQ; /* Product time varying quantitative * time varying dummy */
8789: ncovv++; /* Varying variables without age */
8790: TvarV[ncovv]=Tvar[k];
8791: TvarVind[ncovv]=k;
8792: }else if(Tvard[k1][2] <=ncovcol+nqv+ntv+nqtv){
8793: Fixed[k]= 1;
8794: Dummy[k]= 1;
8795: modell[k].maintype= VTYPE;
8796: modell[k].subtype= VPQQ; /* Product time varying quantitative * time varying quantitative */
8797: ncovv++; /* Varying variables without age */
8798: TvarV[ncovv]=Tvar[k];
8799: TvarVind[ncovv]=k;
8800: }
1.227 brouard 8801: }else{
1.240 brouard 8802: printf("Error unknown type of covariate: Tvard[%d][1]=%d,Tvard[%d][2]=%d\n",k1,Tvard[k1][1],k1,Tvard[k1][2]);
8803: fprintf(ficlog,"Error unknown type of covariate: Tvard[%d][1]=%d,Tvard[%d][2]=%d\n",k1,Tvard[k1][1],k1,Tvard[k1][2]);
8804: } /*end k1*/
1.225 brouard 8805: }else{
1.226 brouard 8806: printf("Error, current version can't treat for performance reasons, Tvar[%d]=%d, Typevar[%d]=%d\n", k, Tvar[k], k, Typevar[k]);
8807: 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 8808: }
1.227 brouard 8809: 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 8810: printf(" modell[%d].maintype=%d, modell[%d].subtype=%d\n",k,modell[k].maintype,k,modell[k].subtype);
1.227 brouard 8811: 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]);
8812: }
8813: /* Searching for doublons in the model */
8814: for(k1=1; k1<= cptcovt;k1++){
8815: for(k2=1; k2 <k1;k2++){
8816: if((Typevar[k1]==Typevar[k2]) && (Fixed[Tvar[k1]]==Fixed[Tvar[k2]]) && (Dummy[Tvar[k1]]==Dummy[Tvar[k2]] )){
1.234 brouard 8817: if((Typevar[k1] == 0 || Typevar[k1] == 1)){ /* Simple or age product */
8818: if(Tvar[k1]==Tvar[k2]){
8819: 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]]);
8820: 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);
8821: return(1);
8822: }
8823: }else if (Typevar[k1] ==2){
8824: k3=Tposprod[k1];
8825: k4=Tposprod[k2];
8826: 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])) ){
8827: 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]]);
8828: 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);
8829: return(1);
8830: }
8831: }
1.227 brouard 8832: }
8833: }
1.225 brouard 8834: }
8835: printf("ncoveff=%d, nqfveff=%d, ntveff=%d, nqtveff=%d, cptcovn=%d\n",ncoveff,nqfveff,ntveff,nqtveff,cptcovn);
8836: fprintf(ficlog,"ncoveff=%d, nqfveff=%d, ntveff=%d, nqtveff=%d, cptcovn=%d\n",ncoveff,nqfveff,ntveff,nqtveff,cptcovn);
1.234 brouard 8837: printf("ncovf=%d, ncovv=%d, ncova=%d, nsd=%d, nsq=%d\n",ncovf,ncovv,ncova,nsd,nsq);
8838: fprintf(ficlog,"ncovf=%d, ncovv=%d, ncova=%d, nsd=%d, nsq=%d\n",ncovf,ncovv,ncova,nsd, nsq);
1.137 brouard 8839: return (0); /* with covar[new additional covariate if product] and Tage if age */
1.164 brouard 8840: /*endread:*/
1.225 brouard 8841: printf("Exiting decodemodel: ");
8842: return (1);
1.136 brouard 8843: }
8844:
1.169 brouard 8845: int calandcheckages(int imx, int maxwav, double *agemin, double *agemax, int *nberr, int *nbwarn )
1.248 ! brouard 8846: {/* Check ages at death */
1.136 brouard 8847: int i, m;
1.218 brouard 8848: int firstone=0;
8849:
1.136 brouard 8850: for (i=1; i<=imx; i++) {
8851: for(m=2; (m<= maxwav); m++) {
8852: if (((int)mint[m][i]== 99) && (s[m][i] <= nlstate)){
8853: anint[m][i]=9999;
1.216 brouard 8854: if (s[m][i] != -2) /* Keeping initial status of unknown vital status */
8855: s[m][i]=-1;
1.136 brouard 8856: }
8857: if((int)moisdc[i]==99 && (int)andc[i]==9999 && s[m][i]>nlstate){
1.169 brouard 8858: *nberr = *nberr + 1;
1.218 brouard 8859: if(firstone == 0){
8860: firstone=1;
8861: printf("Error! Date of death (month %2d and year %4d) of individual %ld on line %d was unknown, you must set an arbitrary year of death or he/she is skipped and results can be biased (%d) because status is a death state %d at wave %d. Wave dropped.\nOther similar cases in log file\n",(int)moisdc[i],(int)andc[i],num[i],i, *nberr,s[m][i],m);
8862: }
8863: fprintf(ficlog,"Error! Date of death (month %2d and year %4d) of individual %ld on line %d was unknown, you must set an arbitrary year of death or he/she is skipped and results can be biased (%d) because status is a death state %d at wave %d. Wave dropped.\n",(int)moisdc[i],(int)andc[i],num[i],i, *nberr,s[m][i],m);
1.136 brouard 8864: s[m][i]=-1;
8865: }
8866: if((int)moisdc[i]==99 && (int)andc[i]!=9999 && s[m][i]>nlstate){
1.169 brouard 8867: (*nberr)++;
1.136 brouard 8868: printf("Error! Month of death of individual %ld on line %d was unknown %2d, you should set it otherwise the information on the death is skipped and results are biased.\n",num[i],i,(int)moisdc[i]);
8869: fprintf(ficlog,"Error! Month of death of individual %ld on line %d was unknown %f, you should set it otherwise the information on the death is skipped and results are biased.\n",num[i],i,moisdc[i]);
8870: s[m][i]=-1; /* We prefer to skip it (and to skip it in version 0.8a1 too */
8871: }
8872: }
8873: }
8874:
8875: for (i=1; i<=imx; i++) {
8876: agedc[i]=(moisdc[i]/12.+andc[i])-(moisnais[i]/12.+annais[i]);
8877: for(m=firstpass; (m<= lastpass); m++){
1.214 brouard 8878: 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 8879: if (s[m][i] >= nlstate+1) {
1.169 brouard 8880: if(agedc[i]>0){
8881: if((int)moisdc[i]!=99 && (int)andc[i]!=9999){
1.136 brouard 8882: agev[m][i]=agedc[i];
1.214 brouard 8883: /*if(moisdc[i]==99 && andc[i]==9999) s[m][i]=-1;*/
1.169 brouard 8884: }else {
1.136 brouard 8885: if ((int)andc[i]!=9999){
8886: nbwarn++;
8887: printf("Warning negative age at death: %ld line:%d\n",num[i],i);
8888: fprintf(ficlog,"Warning negative age at death: %ld line:%d\n",num[i],i);
8889: agev[m][i]=-1;
8890: }
8891: }
1.169 brouard 8892: } /* agedc > 0 */
1.214 brouard 8893: } /* end if */
1.136 brouard 8894: else if(s[m][i] !=9){ /* Standard case, age in fractional
8895: years but with the precision of a month */
8896: agev[m][i]=(mint[m][i]/12.+1./24.+anint[m][i])-(moisnais[i]/12.+1./24.+annais[i]);
8897: if((int)mint[m][i]==99 || (int)anint[m][i]==9999)
8898: agev[m][i]=1;
8899: else if(agev[m][i] < *agemin){
8900: *agemin=agev[m][i];
8901: printf(" Min anint[%d][%d]=%.2f annais[%d]=%.2f, agemin=%.2f\n",m,i,anint[m][i], i,annais[i], *agemin);
8902: }
8903: else if(agev[m][i] >*agemax){
8904: *agemax=agev[m][i];
1.156 brouard 8905: /* printf(" Max anint[%d][%d]=%.0f annais[%d]=%.0f, agemax=%.2f\n",m,i,anint[m][i], i,annais[i], *agemax);*/
1.136 brouard 8906: }
8907: /*agev[m][i]=anint[m][i]-annais[i];*/
8908: /* agev[m][i] = age[i]+2*m;*/
1.214 brouard 8909: } /* en if 9*/
1.136 brouard 8910: else { /* =9 */
1.214 brouard 8911: /* printf("Debug num[%d]=%ld s[%d][%d]=%d\n",i,num[i], m,i, s[m][i]); */
1.136 brouard 8912: agev[m][i]=1;
8913: s[m][i]=-1;
8914: }
8915: }
1.214 brouard 8916: else if(s[m][i]==0) /*= 0 Unknown */
1.136 brouard 8917: agev[m][i]=1;
1.214 brouard 8918: else{
8919: printf("Warning, num[%d]=%ld, s[%d][%d]=%d\n", i, num[i], m, i,s[m][i]);
8920: fprintf(ficlog, "Warning, num[%d]=%ld, s[%d][%d]=%d\n", i, num[i], m, i,s[m][i]);
8921: agev[m][i]=0;
8922: }
8923: } /* End for lastpass */
8924: }
1.136 brouard 8925:
8926: for (i=1; i<=imx; i++) {
8927: for(m=firstpass; (m<=lastpass); m++){
8928: if (s[m][i] > (nlstate+ndeath)) {
1.169 brouard 8929: (*nberr)++;
1.136 brouard 8930: 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);
8931: 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);
8932: return 1;
8933: }
8934: }
8935: }
8936:
8937: /*for (i=1; i<=imx; i++){
8938: for (m=firstpass; (m<lastpass); m++){
8939: printf("%ld %d %.lf %d %d\n", num[i],(covar[1][i]),agev[m][i],s[m][i],s[m+1][i]);
8940: }
8941:
8942: }*/
8943:
8944:
1.139 brouard 8945: printf("Total number of individuals= %d, Agemin = %.2f, Agemax= %.2f\n\n", imx, *agemin, *agemax);
8946: fprintf(ficlog,"Total number of individuals= %d, Agemin = %.2f, Agemax= %.2f\n\n", imx, *agemin, *agemax);
1.136 brouard 8947:
8948: return (0);
1.164 brouard 8949: /* endread:*/
1.136 brouard 8950: printf("Exiting calandcheckages: ");
8951: return (1);
8952: }
8953:
1.172 brouard 8954: #if defined(_MSC_VER)
8955: /*printf("Visual C++ compiler: %s \n;", _MSC_FULL_VER);*/
8956: /*fprintf(ficlog, "Visual C++ compiler: %s \n;", _MSC_FULL_VER);*/
8957: //#include "stdafx.h"
8958: //#include <stdio.h>
8959: //#include <tchar.h>
8960: //#include <windows.h>
8961: //#include <iostream>
8962: typedef BOOL(WINAPI *LPFN_ISWOW64PROCESS) (HANDLE, PBOOL);
8963:
8964: LPFN_ISWOW64PROCESS fnIsWow64Process;
8965:
8966: BOOL IsWow64()
8967: {
8968: BOOL bIsWow64 = FALSE;
8969:
8970: //typedef BOOL (APIENTRY *LPFN_ISWOW64PROCESS)
8971: // (HANDLE, PBOOL);
8972:
8973: //LPFN_ISWOW64PROCESS fnIsWow64Process;
8974:
8975: HMODULE module = GetModuleHandle(_T("kernel32"));
8976: const char funcName[] = "IsWow64Process";
8977: fnIsWow64Process = (LPFN_ISWOW64PROCESS)
8978: GetProcAddress(module, funcName);
8979:
8980: if (NULL != fnIsWow64Process)
8981: {
8982: if (!fnIsWow64Process(GetCurrentProcess(),
8983: &bIsWow64))
8984: //throw std::exception("Unknown error");
8985: printf("Unknown error\n");
8986: }
8987: return bIsWow64 != FALSE;
8988: }
8989: #endif
1.177 brouard 8990:
1.191 brouard 8991: void syscompilerinfo(int logged)
1.167 brouard 8992: {
8993: /* #include "syscompilerinfo.h"*/
1.185 brouard 8994: /* command line Intel compiler 32bit windows, XP compatible:*/
8995: /* /GS /W3 /Gy
8996: /Zc:wchar_t /Zi /O2 /Fd"Release\vc120.pdb" /D "WIN32" /D "NDEBUG" /D
8997: "_CONSOLE" /D "_LIB" /D "_USING_V110_SDK71_" /D "_UNICODE" /D
8998: "UNICODE" /Qipo /Zc:forScope /Gd /Oi /MT /Fa"Release\" /EHsc /nologo
1.186 brouard 8999: /Fo"Release\" /Qprof-dir "Release\" /Fp"Release\IMaCh.pch"
9000: */
9001: /* 64 bits */
1.185 brouard 9002: /*
9003: /GS /W3 /Gy
9004: /Zc:wchar_t /Zi /O2 /Fd"x64\Release\vc120.pdb" /D "WIN32" /D "NDEBUG"
9005: /D "_CONSOLE" /D "_LIB" /D "_UNICODE" /D "UNICODE" /Qipo /Zc:forScope
9006: /Oi /MD /Fa"x64\Release\" /EHsc /nologo /Fo"x64\Release\" /Qprof-dir
9007: "x64\Release\" /Fp"x64\Release\IMaCh.pch" */
9008: /* Optimization are useless and O3 is slower than O2 */
9009: /*
9010: /GS /W3 /Gy /Zc:wchar_t /Zi /O3 /Fd"x64\Release\vc120.pdb" /D "WIN32"
9011: /D "NDEBUG" /D "_CONSOLE" /D "_LIB" /D "_UNICODE" /D "UNICODE" /Qipo
9012: /Zc:forScope /Oi /MD /Fa"x64\Release\" /EHsc /nologo /Qparallel
9013: /Fo"x64\Release\" /Qprof-dir "x64\Release\" /Fp"x64\Release\IMaCh.pch"
9014: */
1.186 brouard 9015: /* Link is */ /* /OUT:"visual studio
1.185 brouard 9016: 2013\Projects\IMaCh\Release\IMaCh.exe" /MANIFEST /NXCOMPAT
9017: /PDB:"visual studio
9018: 2013\Projects\IMaCh\Release\IMaCh.pdb" /DYNAMICBASE
9019: "kernel32.lib" "user32.lib" "gdi32.lib" "winspool.lib"
9020: "comdlg32.lib" "advapi32.lib" "shell32.lib" "ole32.lib"
9021: "oleaut32.lib" "uuid.lib" "odbc32.lib" "odbccp32.lib"
9022: /MACHINE:X86 /OPT:REF /SAFESEH /INCREMENTAL:NO
9023: /SUBSYSTEM:CONSOLE",5.01" /MANIFESTUAC:"level='asInvoker'
9024: uiAccess='false'"
9025: /ManifestFile:"Release\IMaCh.exe.intermediate.manifest" /OPT:ICF
9026: /NOLOGO /TLBID:1
9027: */
1.177 brouard 9028: #if defined __INTEL_COMPILER
1.178 brouard 9029: #if defined(__GNUC__)
9030: struct utsname sysInfo; /* For Intel on Linux and OS/X */
9031: #endif
1.177 brouard 9032: #elif defined(__GNUC__)
1.179 brouard 9033: #ifndef __APPLE__
1.174 brouard 9034: #include <gnu/libc-version.h> /* Only on gnu */
1.179 brouard 9035: #endif
1.177 brouard 9036: struct utsname sysInfo;
1.178 brouard 9037: int cross = CROSS;
9038: if (cross){
9039: printf("Cross-");
1.191 brouard 9040: if(logged) fprintf(ficlog, "Cross-");
1.178 brouard 9041: }
1.174 brouard 9042: #endif
9043:
1.171 brouard 9044: #include <stdint.h>
1.178 brouard 9045:
1.191 brouard 9046: printf("Compiled with:");if(logged)fprintf(ficlog,"Compiled with:");
1.169 brouard 9047: #if defined(__clang__)
1.191 brouard 9048: printf(" Clang/LLVM");if(logged)fprintf(ficlog," Clang/LLVM"); /* Clang/LLVM. ---------------------------------------------- */
1.169 brouard 9049: #endif
9050: #if defined(__ICC) || defined(__INTEL_COMPILER)
1.191 brouard 9051: printf(" Intel ICC/ICPC");if(logged)fprintf(ficlog," Intel ICC/ICPC");/* Intel ICC/ICPC. ------------------------------------------ */
1.169 brouard 9052: #endif
9053: #if defined(__GNUC__) || defined(__GNUG__)
1.191 brouard 9054: printf(" GNU GCC/G++");if(logged)fprintf(ficlog," GNU GCC/G++");/* GNU GCC/G++. --------------------------------------------- */
1.169 brouard 9055: #endif
9056: #if defined(__HP_cc) || defined(__HP_aCC)
1.191 brouard 9057: printf(" Hewlett-Packard C/aC++");if(logged)fprintf(fcilog," Hewlett-Packard C/aC++"); /* Hewlett-Packard C/aC++. ---------------------------------- */
1.169 brouard 9058: #endif
9059: #if defined(__IBMC__) || defined(__IBMCPP__)
1.191 brouard 9060: printf(" IBM XL C/C++"); if(logged) fprintf(ficlog," IBM XL C/C++");/* IBM XL C/C++. -------------------------------------------- */
1.169 brouard 9061: #endif
9062: #if defined(_MSC_VER)
1.191 brouard 9063: printf(" Microsoft Visual Studio");if(logged)fprintf(ficlog," Microsoft Visual Studio");/* Microsoft Visual Studio. --------------------------------- */
1.169 brouard 9064: #endif
9065: #if defined(__PGI)
1.191 brouard 9066: printf(" Portland Group PGCC/PGCPP");if(logged) fprintf(ficlog," Portland Group PGCC/PGCPP");/* Portland Group PGCC/PGCPP. ------------------------------- */
1.169 brouard 9067: #endif
9068: #if defined(__SUNPRO_C) || defined(__SUNPRO_CC)
1.191 brouard 9069: printf(" Oracle Solaris Studio");if(logged)fprintf(ficlog," Oracle Solaris Studio\n");/* Oracle Solaris Studio. ----------------------------------- */
1.167 brouard 9070: #endif
1.191 brouard 9071: printf(" for "); if (logged) fprintf(ficlog, " for ");
1.169 brouard 9072:
1.167 brouard 9073: // http://stackoverflow.com/questions/4605842/how-to-identify-platform-compiler-from-preprocessor-macros
9074: #ifdef _WIN32 // note the underscore: without it, it's not msdn official!
9075: // Windows (x64 and x86)
1.191 brouard 9076: printf("Windows (x64 and x86) ");if(logged) fprintf(ficlog,"Windows (x64 and x86) ");
1.167 brouard 9077: #elif __unix__ // all unices, not all compilers
9078: // Unix
1.191 brouard 9079: printf("Unix ");if(logged) fprintf(ficlog,"Unix ");
1.167 brouard 9080: #elif __linux__
9081: // linux
1.191 brouard 9082: printf("linux ");if(logged) fprintf(ficlog,"linux ");
1.167 brouard 9083: #elif __APPLE__
1.174 brouard 9084: // Mac OS, not sure if this is covered by __posix__ and/or __unix__ though..
1.191 brouard 9085: printf("Mac OS ");if(logged) fprintf(ficlog,"Mac OS ");
1.167 brouard 9086: #endif
9087:
9088: /* __MINGW32__ */
9089: /* __CYGWIN__ */
9090: /* __MINGW64__ */
9091: // http://msdn.microsoft.com/en-us/library/b0084kay.aspx
9092: /* _MSC_VER //the Visual C++ compiler is 17.00.51106.1, the _MSC_VER macro evaluates to 1700. Type cl /? */
9093: /* _MSC_FULL_VER //the Visual C++ compiler is 15.00.20706.01, the _MSC_FULL_VER macro evaluates to 150020706 */
9094: /* _WIN64 // Defined for applications for Win64. */
9095: /* _M_X64 // Defined for compilations that target x64 processors. */
9096: /* _DEBUG // Defined when you compile with /LDd, /MDd, and /MTd. */
1.171 brouard 9097:
1.167 brouard 9098: #if UINTPTR_MAX == 0xffffffff
1.191 brouard 9099: printf(" 32-bit"); if(logged) fprintf(ficlog," 32-bit");/* 32-bit */
1.167 brouard 9100: #elif UINTPTR_MAX == 0xffffffffffffffff
1.191 brouard 9101: printf(" 64-bit"); if(logged) fprintf(ficlog," 64-bit");/* 64-bit */
1.167 brouard 9102: #else
1.191 brouard 9103: printf(" wtf-bit"); if(logged) fprintf(ficlog," wtf-bit");/* wtf */
1.167 brouard 9104: #endif
9105:
1.169 brouard 9106: #if defined(__GNUC__)
9107: # if defined(__GNUC_PATCHLEVEL__)
9108: # define __GNUC_VERSION__ (__GNUC__ * 10000 \
9109: + __GNUC_MINOR__ * 100 \
9110: + __GNUC_PATCHLEVEL__)
9111: # else
9112: # define __GNUC_VERSION__ (__GNUC__ * 10000 \
9113: + __GNUC_MINOR__ * 100)
9114: # endif
1.174 brouard 9115: printf(" using GNU C version %d.\n", __GNUC_VERSION__);
1.191 brouard 9116: if(logged) fprintf(ficlog, " using GNU C version %d.\n", __GNUC_VERSION__);
1.176 brouard 9117:
9118: if (uname(&sysInfo) != -1) {
9119: printf("Running on: %s %s %s %s %s\n",sysInfo.sysname, sysInfo.nodename, sysInfo.release, sysInfo.version, sysInfo.machine);
1.191 brouard 9120: 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 9121: }
9122: else
9123: perror("uname() error");
1.179 brouard 9124: //#ifndef __INTEL_COMPILER
9125: #if !defined (__INTEL_COMPILER) && !defined(__APPLE__)
1.174 brouard 9126: printf("GNU libc version: %s\n", gnu_get_libc_version());
1.191 brouard 9127: if(logged) fprintf(ficlog,"GNU libc version: %s\n", gnu_get_libc_version());
1.177 brouard 9128: #endif
1.169 brouard 9129: #endif
1.172 brouard 9130:
9131: // void main()
9132: // {
1.169 brouard 9133: #if defined(_MSC_VER)
1.174 brouard 9134: if (IsWow64()){
1.191 brouard 9135: printf("\nThe program (probably compiled for 32bit) is running under WOW64 (64bit) emulation.\n");
9136: if (logged) fprintf(ficlog, "\nThe program (probably compiled for 32bit) is running under WOW64 (64bit) emulation.\n");
1.174 brouard 9137: }
9138: else{
1.191 brouard 9139: printf("\nThe program is not running under WOW64 (i.e probably on a 64bit Windows).\n");
9140: if (logged) fprintf(ficlog, "\nThe programm is not running under WOW64 (i.e probably on a 64bit Windows).\n");
1.174 brouard 9141: }
1.172 brouard 9142: // printf("\nPress Enter to continue...");
9143: // getchar();
9144: // }
9145:
1.169 brouard 9146: #endif
9147:
1.167 brouard 9148:
1.219 brouard 9149: }
1.136 brouard 9150:
1.219 brouard 9151: int prevalence_limit(double *p, double **prlim, double ageminpar, double agemaxpar, double ftolpl, int *ncvyearp){
1.180 brouard 9152: /*--------------- Prevalence limit (period or stable prevalence) --------------*/
1.235 brouard 9153: int i, j, k, i1, k4=0, nres=0 ;
1.202 brouard 9154: /* double ftolpl = 1.e-10; */
1.180 brouard 9155: double age, agebase, agelim;
1.203 brouard 9156: double tot;
1.180 brouard 9157:
1.202 brouard 9158: strcpy(filerespl,"PL_");
9159: strcat(filerespl,fileresu);
9160: if((ficrespl=fopen(filerespl,"w"))==NULL) {
9161: printf("Problem with period (stable) prevalence resultfile: %s\n", filerespl);return 1;
9162: fprintf(ficlog,"Problem with period (stable) prevalence resultfile: %s\n", filerespl);return 1;
9163: }
1.227 brouard 9164: printf("\nComputing period (stable) prevalence: result on file '%s' \n", filerespl);
9165: fprintf(ficlog,"\nComputing period (stable) prevalence: result on file '%s' \n", filerespl);
1.202 brouard 9166: pstamp(ficrespl);
1.203 brouard 9167: fprintf(ficrespl,"# Period (stable) prevalence. Precision given by ftolpl=%g \n", ftolpl);
1.202 brouard 9168: fprintf(ficrespl,"#Age ");
9169: for(i=1; i<=nlstate;i++) fprintf(ficrespl,"%d-%d ",i,i);
9170: fprintf(ficrespl,"\n");
1.180 brouard 9171:
1.219 brouard 9172: /* prlim=matrix(1,nlstate,1,nlstate);*/ /* back in main */
1.180 brouard 9173:
1.219 brouard 9174: agebase=ageminpar;
9175: agelim=agemaxpar;
1.180 brouard 9176:
1.227 brouard 9177: /* i1=pow(2,ncoveff); */
1.234 brouard 9178: i1=pow(2,cptcoveff); /* Number of combination of dummy covariates */
1.219 brouard 9179: if (cptcovn < 1){i1=1;}
1.180 brouard 9180:
1.238 brouard 9181: for(k=1; k<=i1;k++){ /* For each combination k of dummy covariates in the model */
9182: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
9183: if(TKresult[nres]!= k)
9184: continue;
1.235 brouard 9185:
1.238 brouard 9186: /* for(cptcov=1,k=0;cptcov<=i1;cptcov++){ */
9187: /* for(cptcov=1,k=0;cptcov<=1;cptcov++){ */
9188: //for(cptcod=1;cptcod<=ncodemax[cptcov];cptcod++){
9189: /* k=k+1; */
9190: /* to clean */
9191: //printf("cptcov=%d cptcod=%d codtab=%d\n",cptcov, cptcod,codtabm(cptcod,cptcov));
9192: fprintf(ficrespl,"#******");
9193: printf("#******");
9194: fprintf(ficlog,"#******");
9195: for(j=1;j<=cptcoveff ;j++) {/* all covariates */
9196: fprintf(ficrespl," V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]); /* Here problem for varying dummy*/
9197: printf(" V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9198: fprintf(ficlog," V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9199: }
9200: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
9201: printf(" V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
9202: fprintf(ficrespl," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
9203: fprintf(ficlog," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
9204: }
9205: fprintf(ficrespl,"******\n");
9206: printf("******\n");
9207: fprintf(ficlog,"******\n");
9208: if(invalidvarcomb[k]){
9209: printf("\nCombination (%d) ignored because no case \n",k);
9210: fprintf(ficrespl,"#Combination (%d) ignored because no case \n",k);
9211: fprintf(ficlog,"\nCombination (%d) ignored because no case \n",k);
9212: continue;
9213: }
1.219 brouard 9214:
1.238 brouard 9215: fprintf(ficrespl,"#Age ");
9216: for(j=1;j<=cptcoveff;j++) {
9217: fprintf(ficrespl,"V%d %d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9218: }
9219: for(i=1; i<=nlstate;i++) fprintf(ficrespl," %d-%d ",i,i);
9220: fprintf(ficrespl,"Total Years_to_converge\n");
1.227 brouard 9221:
1.238 brouard 9222: for (age=agebase; age<=agelim; age++){
9223: /* for (age=agebase; age<=agebase; age++){ */
9224: prevalim(prlim, nlstate, p, age, oldm, savm, ftolpl, ncvyearp, k, nres);
9225: fprintf(ficrespl,"%.0f ",age );
9226: for(j=1;j<=cptcoveff;j++)
9227: fprintf(ficrespl,"%d %d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9228: tot=0.;
9229: for(i=1; i<=nlstate;i++){
9230: tot += prlim[i][i];
9231: fprintf(ficrespl," %.5f", prlim[i][i]);
9232: }
9233: fprintf(ficrespl," %.3f %d\n", tot, *ncvyearp);
9234: } /* Age */
9235: /* was end of cptcod */
9236: } /* cptcov */
9237: } /* nres */
1.219 brouard 9238: return 0;
1.180 brouard 9239: }
9240:
1.218 brouard 9241: 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){
9242: /*--------------- Back Prevalence limit (period or stable prevalence) --------------*/
9243:
9244: /* Computes the back prevalence limit for any combination of covariate values
9245: * at any age between ageminpar and agemaxpar
9246: */
1.235 brouard 9247: int i, j, k, i1, nres=0 ;
1.217 brouard 9248: /* double ftolpl = 1.e-10; */
9249: double age, agebase, agelim;
9250: double tot;
1.218 brouard 9251: /* double ***mobaverage; */
9252: /* double **dnewm, **doldm, **dsavm; /\* for use *\/ */
1.217 brouard 9253:
9254: strcpy(fileresplb,"PLB_");
9255: strcat(fileresplb,fileresu);
9256: if((ficresplb=fopen(fileresplb,"w"))==NULL) {
9257: printf("Problem with period (stable) back prevalence resultfile: %s\n", fileresplb);return 1;
9258: fprintf(ficlog,"Problem with period (stable) back prevalence resultfile: %s\n", fileresplb);return 1;
9259: }
9260: printf("Computing period (stable) back prevalence: result on file '%s' \n", fileresplb);
9261: fprintf(ficlog,"Computing period (stable) back prevalence: result on file '%s' \n", fileresplb);
9262: pstamp(ficresplb);
9263: fprintf(ficresplb,"# Period (stable) back prevalence. Precision given by ftolpl=%g \n", ftolpl);
9264: fprintf(ficresplb,"#Age ");
9265: for(i=1; i<=nlstate;i++) fprintf(ficresplb,"%d-%d ",i,i);
9266: fprintf(ficresplb,"\n");
9267:
1.218 brouard 9268:
9269: /* prlim=matrix(1,nlstate,1,nlstate);*/ /* back in main */
9270:
9271: agebase=ageminpar;
9272: agelim=agemaxpar;
9273:
9274:
1.227 brouard 9275: i1=pow(2,cptcoveff);
1.218 brouard 9276: if (cptcovn < 1){i1=1;}
1.227 brouard 9277:
1.238 brouard 9278: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
9279: for(k=1; k<=i1;k++){ /* For any combination of dummy covariates, fixed and varying */
9280: if(TKresult[nres]!= k)
9281: continue;
9282: //printf("cptcov=%d cptcod=%d codtab=%d\n",cptcov, cptcod,codtabm(cptcod,cptcov));
9283: fprintf(ficresplb,"#******");
9284: printf("#******");
9285: fprintf(ficlog,"#******");
9286: for(j=1;j<=cptcoveff ;j++) {/* all covariates */
9287: fprintf(ficresplb," V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9288: printf(" V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9289: fprintf(ficlog," V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9290: }
9291: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
9292: printf(" V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
9293: fprintf(ficresplb," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
9294: fprintf(ficlog," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
9295: }
9296: fprintf(ficresplb,"******\n");
9297: printf("******\n");
9298: fprintf(ficlog,"******\n");
9299: if(invalidvarcomb[k]){
9300: printf("\nCombination (%d) ignored because no cases \n",k);
9301: fprintf(ficresplb,"#Combination (%d) ignored because no cases \n",k);
9302: fprintf(ficlog,"\nCombination (%d) ignored because no cases \n",k);
9303: continue;
9304: }
1.218 brouard 9305:
1.238 brouard 9306: fprintf(ficresplb,"#Age ");
9307: for(j=1;j<=cptcoveff;j++) {
9308: fprintf(ficresplb,"V%d %d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9309: }
9310: for(i=1; i<=nlstate;i++) fprintf(ficresplb," %d-%d ",i,i);
9311: fprintf(ficresplb,"Total Years_to_converge\n");
1.218 brouard 9312:
9313:
1.238 brouard 9314: for (age=agebase; age<=agelim; age++){
9315: /* for (age=agebase; age<=agebase; age++){ */
9316: if(mobilavproj > 0){
9317: /* bprevalim(bprlim, mobaverage, nlstate, p, age, ageminpar, agemaxpar, oldm, savm, doldm, dsavm, ftolpl, ncvyearp, k); */
9318: /* bprevalim(bprlim, mobaverage, nlstate, p, age, oldm, savm, dnewm, doldm, dsavm, ftolpl, ncvyearp, k); */
1.242 brouard 9319: bprevalim(bprlim, mobaverage, nlstate, p, age, ftolpl, ncvyearp, k, nres);
1.238 brouard 9320: }else if (mobilavproj == 0){
9321: 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);
9322: 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);
9323: exit(1);
9324: }else{
9325: /* bprevalim(bprlim, probs, nlstate, p, age, oldm, savm, dnewm, doldm, dsavm, ftolpl, ncvyearp, k); */
1.242 brouard 9326: bprevalim(bprlim, probs, nlstate, p, age, ftolpl, ncvyearp, k,nres);
1.238 brouard 9327: }
9328: fprintf(ficresplb,"%.0f ",age );
9329: for(j=1;j<=cptcoveff;j++)
9330: fprintf(ficresplb,"%d %d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9331: tot=0.;
9332: for(i=1; i<=nlstate;i++){
9333: tot += bprlim[i][i];
9334: fprintf(ficresplb," %.5f", bprlim[i][i]);
9335: }
9336: fprintf(ficresplb," %.3f %d\n", tot, *ncvyearp);
9337: } /* Age */
9338: /* was end of cptcod */
9339: } /* end of any combination */
9340: } /* end of nres */
1.218 brouard 9341: /* hBijx(p, bage, fage); */
9342: /* fclose(ficrespijb); */
9343:
9344: return 0;
1.217 brouard 9345: }
1.218 brouard 9346:
1.180 brouard 9347: int hPijx(double *p, int bage, int fage){
9348: /*------------- h Pij x at various ages ------------*/
9349:
9350: int stepsize;
9351: int agelim;
9352: int hstepm;
9353: int nhstepm;
1.235 brouard 9354: int h, i, i1, j, k, k4, nres=0;
1.180 brouard 9355:
9356: double agedeb;
9357: double ***p3mat;
9358:
1.201 brouard 9359: strcpy(filerespij,"PIJ_"); strcat(filerespij,fileresu);
1.180 brouard 9360: if((ficrespij=fopen(filerespij,"w"))==NULL) {
9361: printf("Problem with Pij resultfile: %s\n", filerespij); return 1;
9362: fprintf(ficlog,"Problem with Pij resultfile: %s\n", filerespij); return 1;
9363: }
9364: printf("Computing pij: result on file '%s' \n", filerespij);
9365: fprintf(ficlog,"Computing pij: result on file '%s' \n", filerespij);
9366:
9367: stepsize=(int) (stepm+YEARM-1)/YEARM;
9368: /*if (stepm<=24) stepsize=2;*/
9369:
9370: agelim=AGESUP;
9371: hstepm=stepsize*YEARM; /* Every year of age */
9372: hstepm=hstepm/stepm; /* Typically 2 years, = 2/6 months = 4 */
1.218 brouard 9373:
1.180 brouard 9374: /* hstepm=1; aff par mois*/
9375: pstamp(ficrespij);
9376: fprintf(ficrespij,"#****** h Pij x Probability to be in state j at age x+h being in i at x ");
1.227 brouard 9377: i1= pow(2,cptcoveff);
1.218 brouard 9378: /* for(cptcov=1,k=0;cptcov<=i1;cptcov++){ */
9379: /* /\*for(cptcod=1;cptcod<=ncodemax[cptcov];cptcod++){*\/ */
9380: /* k=k+1; */
1.235 brouard 9381: for(nres=1; nres <= nresult; nres++) /* For each resultline */
9382: for(k=1; k<=i1;k++){
9383: if(TKresult[nres]!= k)
9384: continue;
1.183 brouard 9385: fprintf(ficrespij,"\n#****** ");
1.227 brouard 9386: for(j=1;j<=cptcoveff;j++)
1.198 brouard 9387: fprintf(ficrespij,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
1.235 brouard 9388: for (k4=1; k4<= nsq; k4++){ /* For each selected (single) quantitative value */
9389: printf(" V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
9390: fprintf(ficrespij," V%d=%f ",Tvqresult[nres][k4],Tqresult[nres][k4]);
9391: }
1.183 brouard 9392: fprintf(ficrespij,"******\n");
9393:
9394: for (agedeb=fage; agedeb>=bage; agedeb--){ /* If stepm=6 months */
9395: nhstepm=(int) rint((agelim-agedeb)*YEARM/stepm); /* Typically 20 years = 20*12/6=40 */
9396: nhstepm = nhstepm/hstepm; /* Typically 40/4=10 */
9397:
9398: /* nhstepm=nhstepm*YEARM; aff par mois*/
1.180 brouard 9399:
1.183 brouard 9400: p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
9401: oldm=oldms;savm=savms;
1.235 brouard 9402: hpxij(p3mat,nhstepm,agedeb,hstepm,p,nlstate,stepm,oldm,savm, k, nres);
1.183 brouard 9403: fprintf(ficrespij,"# Cov Agex agex+h hpijx with i,j=");
9404: for(i=1; i<=nlstate;i++)
9405: for(j=1; j<=nlstate+ndeath;j++)
9406: fprintf(ficrespij," %1d-%1d",i,j);
9407: fprintf(ficrespij,"\n");
9408: for (h=0; h<=nhstepm; h++){
9409: /*agedebphstep = agedeb + h*hstepm/YEARM*stepm;*/
9410: fprintf(ficrespij,"%d %3.f %3.f",k, agedeb, agedeb + h*hstepm/YEARM*stepm );
1.180 brouard 9411: for(i=1; i<=nlstate;i++)
9412: for(j=1; j<=nlstate+ndeath;j++)
1.183 brouard 9413: fprintf(ficrespij," %.5f", p3mat[i][j][h]);
1.180 brouard 9414: fprintf(ficrespij,"\n");
9415: }
1.183 brouard 9416: free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
9417: fprintf(ficrespij,"\n");
9418: }
1.180 brouard 9419: /*}*/
9420: }
1.218 brouard 9421: return 0;
1.180 brouard 9422: }
1.218 brouard 9423:
9424: int hBijx(double *p, int bage, int fage, double ***prevacurrent){
1.217 brouard 9425: /*------------- h Bij x at various ages ------------*/
9426:
9427: int stepsize;
1.218 brouard 9428: /* int agelim; */
9429: int ageminl;
1.217 brouard 9430: int hstepm;
9431: int nhstepm;
1.238 brouard 9432: int h, i, i1, j, k, nres;
1.218 brouard 9433:
1.217 brouard 9434: double agedeb;
9435: double ***p3mat;
1.218 brouard 9436:
9437: strcpy(filerespijb,"PIJB_"); strcat(filerespijb,fileresu);
9438: if((ficrespijb=fopen(filerespijb,"w"))==NULL) {
9439: printf("Problem with Pij back resultfile: %s\n", filerespijb); return 1;
9440: fprintf(ficlog,"Problem with Pij back resultfile: %s\n", filerespijb); return 1;
9441: }
9442: printf("Computing pij back: result on file '%s' \n", filerespijb);
9443: fprintf(ficlog,"Computing pij back: result on file '%s' \n", filerespijb);
9444:
9445: stepsize=(int) (stepm+YEARM-1)/YEARM;
9446: /*if (stepm<=24) stepsize=2;*/
1.217 brouard 9447:
1.218 brouard 9448: /* agelim=AGESUP; */
9449: ageminl=30;
9450: hstepm=stepsize*YEARM; /* Every year of age */
9451: hstepm=hstepm/stepm; /* Typically 2 years, = 2/6 months = 4 */
9452:
9453: /* hstepm=1; aff par mois*/
9454: pstamp(ficrespijb);
9455: fprintf(ficrespijb,"#****** h Pij x Back Probability to be in state i at age x-h being in j at x ");
1.227 brouard 9456: i1= pow(2,cptcoveff);
1.218 brouard 9457: /* for(cptcov=1,k=0;cptcov<=i1;cptcov++){ */
9458: /* /\*for(cptcod=1;cptcod<=ncodemax[cptcov];cptcod++){*\/ */
9459: /* k=k+1; */
1.238 brouard 9460: for(nres=1; nres <= nresult; nres++){ /* For each resultline */
9461: for(k=1; k<=i1;k++){ /* For any combination of dummy covariates, fixed and varying */
9462: if(TKresult[nres]!= k)
9463: continue;
9464: fprintf(ficrespijb,"\n#****** ");
9465: for(j=1;j<=cptcoveff;j++)
9466: fprintf(ficrespijb,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
9467: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
9468: fprintf(ficrespijb," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
9469: }
9470: fprintf(ficrespijb,"******\n");
9471: if(invalidvarcomb[k]){
9472: fprintf(ficrespijb,"\n#Combination (%d) ignored because no cases \n",k);
9473: continue;
9474: }
9475:
9476: /* for (agedeb=fage; agedeb>=bage; agedeb--){ /\* If stepm=6 months *\/ */
9477: for (agedeb=bage; agedeb<=fage; agedeb++){ /* If stepm=6 months and estepm=24 (2 years) */
9478: /* nhstepm=(int) rint((agelim-agedeb)*YEARM/stepm); /\* Typically 20 years = 20*12/6=40 *\/ */
9479: nhstepm=(int) rint((agedeb-ageminl)*YEARM/stepm); /* Typically 20 years = 20*12/6=40 */
9480: nhstepm = nhstepm/hstepm; /* Typically 40/4=10, because estepm=24 stepm=6 => hstepm=24/6=4 */
9481:
9482: /* nhstepm=nhstepm*YEARM; aff par mois*/
9483:
9484: p3mat=ma3x(1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
9485: /* oldm=oldms;savm=savms; */
9486: /* hbxij(p3mat,nhstepm,agedeb,hstepm,p,nlstate,stepm,oldm,savm, k); */
9487: hbxij(p3mat,nhstepm,agedeb,hstepm,p,prevacurrent,nlstate,stepm, k);
9488: /* hbxij(p3mat,nhstepm,agedeb,hstepm,p,prevacurrent,nlstate,stepm,oldm,savm, dnewm, doldm, dsavm, k); */
9489: fprintf(ficrespijb,"# Cov Agex agex-h hpijx with i,j=");
1.217 brouard 9490: for(i=1; i<=nlstate;i++)
9491: for(j=1; j<=nlstate+ndeath;j++)
1.238 brouard 9492: fprintf(ficrespijb," %1d-%1d",i,j);
1.217 brouard 9493: fprintf(ficrespijb,"\n");
1.238 brouard 9494: for (h=0; h<=nhstepm; h++){
9495: /*agedebphstep = agedeb + h*hstepm/YEARM*stepm;*/
9496: fprintf(ficrespijb,"%d %3.f %3.f",k, agedeb, agedeb - h*hstepm/YEARM*stepm );
9497: /* fprintf(ficrespijb,"%d %3.f %3.f",k, agedeb, agedeb + h*hstepm/YEARM*stepm ); */
9498: for(i=1; i<=nlstate;i++)
9499: for(j=1; j<=nlstate+ndeath;j++)
9500: fprintf(ficrespijb," %.5f", p3mat[i][j][h]);
9501: fprintf(ficrespijb,"\n");
9502: }
9503: free_ma3x(p3mat,1,nlstate+ndeath,1, nlstate+ndeath, 0,nhstepm);
9504: fprintf(ficrespijb,"\n");
9505: } /* end age deb */
9506: } /* end combination */
9507: } /* end nres */
1.218 brouard 9508: return 0;
9509: } /* hBijx */
1.217 brouard 9510:
1.180 brouard 9511:
1.136 brouard 9512: /***********************************************/
9513: /**************** Main Program *****************/
9514: /***********************************************/
9515:
9516: int main(int argc, char *argv[])
9517: {
9518: #ifdef GSL
9519: const gsl_multimin_fminimizer_type *T;
9520: size_t iteri = 0, it;
9521: int rval = GSL_CONTINUE;
9522: int status = GSL_SUCCESS;
9523: double ssval;
9524: #endif
9525: int movingaverage(double ***probs, double bage,double fage, double ***mobaverage, int mobilav);
1.164 brouard 9526: int i,j, k, n=MAXN,iter=0,m,size=100, cptcod;
1.209 brouard 9527: int ncvyear=0; /* Number of years needed for the period prevalence to converge */
1.164 brouard 9528: int jj, ll, li, lj, lk;
1.136 brouard 9529: int numlinepar=0; /* Current linenumber of parameter file */
1.197 brouard 9530: int num_filled;
1.136 brouard 9531: int itimes;
9532: int NDIM=2;
9533: int vpopbased=0;
1.235 brouard 9534: int nres=0;
1.136 brouard 9535:
1.164 brouard 9536: char ca[32], cb[32];
1.136 brouard 9537: /* FILE *fichtm; *//* Html File */
9538: /* FILE *ficgp;*/ /*Gnuplot File */
9539: struct stat info;
1.191 brouard 9540: double agedeb=0.;
1.194 brouard 9541:
9542: double ageminpar=AGEOVERFLOW,agemin=AGEOVERFLOW, agemaxpar=-AGEOVERFLOW, agemax=-AGEOVERFLOW;
1.219 brouard 9543: double ageminout=-AGEOVERFLOW,agemaxout=AGEOVERFLOW; /* Smaller Age range redefined after movingaverage */
1.136 brouard 9544:
1.165 brouard 9545: double fret;
1.191 brouard 9546: double dum=0.; /* Dummy variable */
1.136 brouard 9547: double ***p3mat;
1.218 brouard 9548: /* double ***mobaverage; */
1.164 brouard 9549:
9550: char line[MAXLINE];
1.197 brouard 9551: char path[MAXLINE],pathc[MAXLINE],pathcd[MAXLINE],pathtot[MAXLINE];
9552:
1.234 brouard 9553: char modeltemp[MAXLINE];
1.230 brouard 9554: char resultline[MAXLINE];
9555:
1.136 brouard 9556: char pathr[MAXLINE], pathimach[MAXLINE];
1.164 brouard 9557: char *tok, *val; /* pathtot */
1.136 brouard 9558: int firstobs=1, lastobs=10;
1.195 brouard 9559: int c, h , cpt, c2;
1.191 brouard 9560: int jl=0;
9561: int i1, j1, jk, stepsize=0;
1.194 brouard 9562: int count=0;
9563:
1.164 brouard 9564: int *tab;
1.136 brouard 9565: int mobilavproj=0 , prevfcast=0 ; /* moving average of prev, If prevfcast=1 prevalence projection */
1.217 brouard 9566: int backcast=0;
1.136 brouard 9567: int mobilav=0,popforecast=0;
1.191 brouard 9568: int hstepm=0, nhstepm=0;
1.136 brouard 9569: int agemortsup;
9570: float sumlpop=0.;
9571: double jprev1=1, mprev1=1,anprev1=2000,jprev2=1, mprev2=1,anprev2=2000;
9572: double jpyram=1, mpyram=1,anpyram=2000,jpyram1=1, mpyram1=1,anpyram1=2000;
9573:
1.191 brouard 9574: double bage=0, fage=110., age, agelim=0., agebase=0.;
1.136 brouard 9575: double ftolpl=FTOL;
9576: double **prlim;
1.217 brouard 9577: double **bprlim;
1.136 brouard 9578: double ***param; /* Matrix of parameters */
9579: double *p;
9580: double **matcov; /* Matrix of covariance */
1.203 brouard 9581: double **hess; /* Hessian matrix */
1.136 brouard 9582: double ***delti3; /* Scale */
9583: double *delti; /* Scale */
9584: double ***eij, ***vareij;
9585: double **varpl; /* Variances of prevalence limits by age */
9586: double *epj, vepp;
1.164 brouard 9587:
1.136 brouard 9588: double dateprev1, dateprev2,jproj1=1,mproj1=1,anproj1=2000,jproj2=1,mproj2=1,anproj2=2000;
1.217 brouard 9589: double jback1=1,mback1=1,anback1=2000,jback2=1,mback2=1,anback2=2000;
9590:
1.136 brouard 9591: double **ximort;
1.145 brouard 9592: char *alph[]={"a","a","b","c","d","e"}, str[4]="1234";
1.136 brouard 9593: int *dcwave;
9594:
1.164 brouard 9595: char z[1]="c";
1.136 brouard 9596:
9597: /*char *strt;*/
9598: char strtend[80];
1.126 brouard 9599:
1.164 brouard 9600:
1.126 brouard 9601: /* setlocale (LC_ALL, ""); */
9602: /* bindtextdomain (PACKAGE, LOCALEDIR); */
9603: /* textdomain (PACKAGE); */
9604: /* setlocale (LC_CTYPE, ""); */
9605: /* setlocale (LC_MESSAGES, ""); */
9606:
9607: /* gettimeofday(&start_time, (struct timezone*)0); */ /* at first time */
1.157 brouard 9608: rstart_time = time(NULL);
9609: /* (void) gettimeofday(&start_time,&tzp);*/
9610: start_time = *localtime(&rstart_time);
1.126 brouard 9611: curr_time=start_time;
1.157 brouard 9612: /*tml = *localtime(&start_time.tm_sec);*/
9613: /* strcpy(strstart,asctime(&tml)); */
9614: strcpy(strstart,asctime(&start_time));
1.126 brouard 9615:
9616: /* printf("Localtime (at start)=%s",strstart); */
1.157 brouard 9617: /* tp.tm_sec = tp.tm_sec +86400; */
9618: /* tm = *localtime(&start_time.tm_sec); */
1.126 brouard 9619: /* tmg.tm_year=tmg.tm_year +dsign*dyear; */
9620: /* tmg.tm_mon=tmg.tm_mon +dsign*dmonth; */
9621: /* tmg.tm_hour=tmg.tm_hour + 1; */
1.157 brouard 9622: /* tp.tm_sec = mktime(&tmg); */
1.126 brouard 9623: /* strt=asctime(&tmg); */
9624: /* printf("Time(after) =%s",strstart); */
9625: /* (void) time (&time_value);
9626: * printf("time=%d,t-=%d\n",time_value,time_value-86400);
9627: * tm = *localtime(&time_value);
9628: * strstart=asctime(&tm);
9629: * printf("tim_value=%d,asctime=%s\n",time_value,strstart);
9630: */
9631:
9632: nberr=0; /* Number of errors and warnings */
9633: nbwarn=0;
1.184 brouard 9634: #ifdef WIN32
9635: _getcwd(pathcd, size);
9636: #else
1.126 brouard 9637: getcwd(pathcd, size);
1.184 brouard 9638: #endif
1.191 brouard 9639: syscompilerinfo(0);
1.196 brouard 9640: printf("\nIMaCh version %s, %s\n%s",version, copyright, fullversion);
1.126 brouard 9641: if(argc <=1){
9642: printf("\nEnter the parameter file name: ");
1.205 brouard 9643: if(!fgets(pathr,FILENAMELENGTH,stdin)){
9644: printf("ERROR Empty parameter file name\n");
9645: goto end;
9646: }
1.126 brouard 9647: i=strlen(pathr);
9648: if(pathr[i-1]=='\n')
9649: pathr[i-1]='\0';
1.156 brouard 9650: i=strlen(pathr);
1.205 brouard 9651: if(i >= 1 && pathr[i-1]==' ') {/* This may happen when dragging on oS/X! */
1.156 brouard 9652: pathr[i-1]='\0';
1.205 brouard 9653: }
9654: i=strlen(pathr);
9655: if( i==0 ){
9656: printf("ERROR Empty parameter file name\n");
9657: goto end;
9658: }
9659: for (tok = pathr; tok != NULL; ){
1.126 brouard 9660: printf("Pathr |%s|\n",pathr);
9661: while ((val = strsep(&tok, "\"" )) != NULL && *val == '\0');
9662: printf("val= |%s| pathr=%s\n",val,pathr);
9663: strcpy (pathtot, val);
9664: if(pathr[0] == '\0') break; /* Dirty */
9665: }
9666: }
9667: else{
9668: strcpy(pathtot,argv[1]);
9669: }
9670: /*if(getcwd(pathcd, MAXLINE)!= NULL)printf ("Error pathcd\n");*/
9671: /*cygwin_split_path(pathtot,path,optionfile);
9672: printf("pathtot=%s, path=%s, optionfile=%s\n",pathtot,path,optionfile);*/
9673: /* cutv(path,optionfile,pathtot,'\\');*/
9674:
9675: /* Split argv[0], imach program to get pathimach */
9676: printf("\nargv[0]=%s argv[1]=%s, \n",argv[0],argv[1]);
9677: split(argv[0],pathimach,optionfile,optionfilext,optionfilefiname);
9678: printf("\nargv[0]=%s pathimach=%s, \noptionfile=%s \noptionfilext=%s \noptionfilefiname=%s\n",argv[0],pathimach,optionfile,optionfilext,optionfilefiname);
9679: /* strcpy(pathimach,argv[0]); */
9680: /* Split argv[1]=pathtot, parameter file name to get path, optionfile, extension and name */
9681: split(pathtot,path,optionfile,optionfilext,optionfilefiname);
9682: printf("\npathtot=%s,\npath=%s,\noptionfile=%s \noptionfilext=%s \noptionfilefiname=%s\n",pathtot,path,optionfile,optionfilext,optionfilefiname);
1.184 brouard 9683: #ifdef WIN32
9684: _chdir(path); /* Can be a relative path */
9685: if(_getcwd(pathcd,MAXLINE) > 0) /* So pathcd is the full path */
9686: #else
1.126 brouard 9687: chdir(path); /* Can be a relative path */
1.184 brouard 9688: if (getcwd(pathcd, MAXLINE) > 0) /* So pathcd is the full path */
9689: #endif
9690: printf("Current directory %s!\n",pathcd);
1.126 brouard 9691: strcpy(command,"mkdir ");
9692: strcat(command,optionfilefiname);
9693: if((outcmd=system(command)) != 0){
1.169 brouard 9694: printf("Directory already exists (or can't create it) %s%s, err=%d\n",path,optionfilefiname,outcmd);
1.126 brouard 9695: /* fprintf(ficlog,"Problem creating directory %s%s\n",path,optionfilefiname); */
9696: /* fclose(ficlog); */
9697: /* exit(1); */
9698: }
9699: /* if((imk=mkdir(optionfilefiname))<0){ */
9700: /* perror("mkdir"); */
9701: /* } */
9702:
9703: /*-------- arguments in the command line --------*/
9704:
1.186 brouard 9705: /* Main Log file */
1.126 brouard 9706: strcat(filelog, optionfilefiname);
9707: strcat(filelog,".log"); /* */
9708: if((ficlog=fopen(filelog,"w"))==NULL) {
9709: printf("Problem with logfile %s\n",filelog);
9710: goto end;
9711: }
9712: fprintf(ficlog,"Log filename:%s\n",filelog);
1.197 brouard 9713: fprintf(ficlog,"Version %s %s",version,fullversion);
1.126 brouard 9714: fprintf(ficlog,"\nEnter the parameter file name: \n");
9715: fprintf(ficlog,"pathimach=%s\npathtot=%s\n\
9716: path=%s \n\
9717: optionfile=%s\n\
9718: optionfilext=%s\n\
1.156 brouard 9719: optionfilefiname='%s'\n",pathimach,pathtot,path,optionfile,optionfilext,optionfilefiname);
1.126 brouard 9720:
1.197 brouard 9721: syscompilerinfo(1);
1.167 brouard 9722:
1.126 brouard 9723: printf("Local time (at start):%s",strstart);
9724: fprintf(ficlog,"Local time (at start): %s",strstart);
9725: fflush(ficlog);
9726: /* (void) gettimeofday(&curr_time,&tzp); */
1.157 brouard 9727: /* printf("Elapsed time %d\n", asc_diff_time(curr_time.tm_sec-start_time.tm_sec,tmpout)); */
1.126 brouard 9728:
9729: /* */
9730: strcpy(fileres,"r");
9731: strcat(fileres, optionfilefiname);
1.201 brouard 9732: strcat(fileresu, optionfilefiname); /* Without r in front */
1.126 brouard 9733: strcat(fileres,".txt"); /* Other files have txt extension */
1.201 brouard 9734: strcat(fileresu,".txt"); /* Other files have txt extension */
1.126 brouard 9735:
1.186 brouard 9736: /* Main ---------arguments file --------*/
1.126 brouard 9737:
9738: if((ficpar=fopen(optionfile,"r"))==NULL) {
1.155 brouard 9739: printf("Problem with optionfile '%s' with errno='%s'\n",optionfile,strerror(errno));
9740: fprintf(ficlog,"Problem with optionfile '%s' with errno='%s'\n",optionfile,strerror(errno));
1.126 brouard 9741: fflush(ficlog);
1.149 brouard 9742: /* goto end; */
9743: exit(70);
1.126 brouard 9744: }
9745:
9746:
9747:
9748: strcpy(filereso,"o");
1.201 brouard 9749: strcat(filereso,fileresu);
1.126 brouard 9750: if((ficparo=fopen(filereso,"w"))==NULL) { /* opened on subdirectory */
9751: printf("Problem with Output resultfile: %s\n", filereso);
9752: fprintf(ficlog,"Problem with Output resultfile: %s\n", filereso);
9753: fflush(ficlog);
9754: goto end;
9755: }
9756:
9757: /* Reads comments: lines beginning with '#' */
9758: numlinepar=0;
1.197 brouard 9759:
9760: /* First parameter line */
9761: while(fgets(line, MAXLINE, ficpar)) {
9762: /* If line starts with a # it is a comment */
9763: if (line[0] == '#') {
9764: numlinepar++;
9765: fputs(line,stdout);
9766: fputs(line,ficparo);
9767: fputs(line,ficlog);
9768: continue;
9769: }else
9770: break;
9771: }
9772: if((num_filled=sscanf(line,"title=%s datafile=%s lastobs=%d firstpass=%d lastpass=%d\n", \
9773: title, datafile, &lastobs, &firstpass,&lastpass)) !=EOF){
9774: if (num_filled != 5) {
9775: printf("Should be 5 parameters\n");
9776: }
1.126 brouard 9777: numlinepar++;
1.197 brouard 9778: printf("title=%s datafile=%s lastobs=%d firstpass=%d lastpass=%d\n", title, datafile, lastobs, firstpass,lastpass);
9779: }
9780: /* Second parameter line */
9781: while(fgets(line, MAXLINE, ficpar)) {
9782: /* If line starts with a # it is a comment */
9783: if (line[0] == '#') {
9784: numlinepar++;
9785: fputs(line,stdout);
9786: fputs(line,ficparo);
9787: fputs(line,ficlog);
9788: continue;
9789: }else
9790: break;
9791: }
1.223 brouard 9792: 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", \
9793: &ftol, &stepm, &ncovcol, &nqv, &ntv, &nqtv, &nlstate, &ndeath, &maxwav, &mle, &weightopt)) !=EOF){
9794: if (num_filled != 11) {
9795: 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 9796: printf("but line=%s\n",line);
1.197 brouard 9797: }
1.223 brouard 9798: 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 9799: }
1.203 brouard 9800: /* ftolpl=6*ftol*1.e5; /\* 6.e-3 make convergences in less than 80 loops for the prevalence limit *\/ */
1.209 brouard 9801: /*ftolpl=6.e-4; *//* 6.e-3 make convergences in less than 80 loops for the prevalence limit */
1.197 brouard 9802: /* Third parameter line */
9803: while(fgets(line, MAXLINE, ficpar)) {
9804: /* If line starts with a # it is a comment */
9805: if (line[0] == '#') {
9806: numlinepar++;
9807: fputs(line,stdout);
9808: fputs(line,ficparo);
9809: fputs(line,ficlog);
9810: continue;
9811: }else
9812: break;
9813: }
1.201 brouard 9814: if((num_filled=sscanf(line,"model=1+age%[^.\n]", model)) !=EOF){
9815: if (num_filled == 0)
9816: model[0]='\0';
9817: else if (num_filled != 1){
1.197 brouard 9818: printf("ERROR %d: Model should be at minimum 'model=1+age.' %s\n",num_filled, line);
9819: fprintf(ficlog,"ERROR %d: Model should be at minimum 'model=1+age.' %s\n",num_filled, line);
9820: model[0]='\0';
9821: goto end;
9822: }
9823: else{
9824: if (model[0]=='+'){
9825: for(i=1; i<=strlen(model);i++)
9826: modeltemp[i-1]=model[i];
1.201 brouard 9827: strcpy(model,modeltemp);
1.197 brouard 9828: }
9829: }
1.199 brouard 9830: /* printf(" model=1+age%s modeltemp= %s, model=%s\n",model, modeltemp, model);fflush(stdout); */
1.203 brouard 9831: printf("model=1+age+%s\n",model);fflush(stdout);
1.197 brouard 9832: }
9833: /* 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); */
9834: /* numlinepar=numlinepar+3; /\* In general *\/ */
9835: /* 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 9836: 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);
9837: 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 9838: fflush(ficlog);
1.190 brouard 9839: /* if(model[0]=='#'|| model[0]== '\0'){ */
9840: if(model[0]=='#'){
1.187 brouard 9841: printf("Error in 'model' line: model should start with 'model=1+age+' and end with '.' \n \
9842: 'model=1+age+.' or 'model=1+age+V1.' or 'model=1+age+age*age+V1+V1*age.' or \n \
9843: 'model=1+age+V1+V2.' or 'model=1+age+V1+V2+V1*V2.' etc. \n"); \
9844: if(mle != -1){
9845: printf("Fix the model line and run imach with mle=-1 to get a correct template of the parameter file.\n");
9846: exit(1);
9847: }
9848: }
1.126 brouard 9849: while((c=getc(ficpar))=='#' && c!= EOF){
9850: ungetc(c,ficpar);
9851: fgets(line, MAXLINE, ficpar);
9852: numlinepar++;
1.195 brouard 9853: if(line[1]=='q'){ /* This #q will quit imach (the answer is q) */
9854: z[0]=line[1];
9855: }
9856: /* printf("****line [1] = %c \n",line[1]); */
1.141 brouard 9857: fputs(line, stdout);
9858: //puts(line);
1.126 brouard 9859: fputs(line,ficparo);
9860: fputs(line,ficlog);
9861: }
9862: ungetc(c,ficpar);
9863:
9864:
1.145 brouard 9865: covar=matrix(0,NCOVMAX,1,n); /**< used in readdata */
1.225 brouard 9866: coqvar=matrix(1,nqv,1,n); /**< Fixed quantitative covariate */
1.233 brouard 9867: cotvar=ma3x(1,maxwav,1,ntv+nqtv,1,n); /**< Time varying covariate (dummy and quantitative)*/
1.225 brouard 9868: cotqvar=ma3x(1,maxwav,1,nqtv,1,n); /**< Time varying quantitative covariate */
1.136 brouard 9869: cptcovn=0; /*Number of covariates, i.e. number of '+' in model statement plus one, indepently of n in Vn*/
9870: /* v1+v2+v3+v2*v4+v5*age makes cptcovn = 5
9871: v1+v2*age+v2*v3 makes cptcovn = 3
9872: */
9873: if (strlen(model)>1)
1.187 brouard 9874: 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 9875: else
1.187 brouard 9876: ncovmodel=2; /* Constant and age */
1.133 brouard 9877: nforce= (nlstate+ndeath-1)*nlstate; /* Number of forces ij from state i to j */
9878: npar= nforce*ncovmodel; /* Number of parameters like aij*/
1.131 brouard 9879: if(npar >MAXPARM || nlstate >NLSTATEMAX || ndeath >NDEATHMAX || ncovmodel>NCOVMAX){
9880: 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);
9881: 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);
9882: fflush(stdout);
9883: fclose (ficlog);
9884: goto end;
9885: }
1.126 brouard 9886: delti3= ma3x(1,nlstate,1,nlstate+ndeath-1,1,ncovmodel);
9887: delti=delti3[1][1];
9888: /*delti=vector(1,npar); *//* Scale of each paramater (output from hesscov)*/
9889: if(mle==-1){ /* Print a wizard for help writing covariance matrix */
1.247 brouard 9890: /* We could also provide initial parameters values giving by simple logistic regression
9891: * only one way, that is without matrix product. We will have nlstate maximizations */
9892: /* for(i=1;i<nlstate;i++){ */
9893: /* /\*reducing xi for 1 to npar to 1 to ncovmodel; *\/ */
9894: /* mlikeli(ficres,p, ncovmodel, ncovmodel, nlstate, ftol, funcnoprod); */
9895: /* } */
1.126 brouard 9896: prwizard(ncovmodel, nlstate, ndeath, model, ficparo);
1.191 brouard 9897: printf(" You chose mle=-1, look at file %s for a template of covariance matrix \n",filereso);
9898: fprintf(ficlog," You chose mle=-1, look at file %s for a template of covariance matrix \n",filereso);
1.126 brouard 9899: free_ma3x(delti3,1,nlstate,1, nlstate+ndeath-1,1,ncovmodel);
9900: fclose (ficparo);
9901: fclose (ficlog);
9902: goto end;
9903: exit(0);
1.248 ! brouard 9904: } else if(mle==-2) { /* Guessing from means */
! 9905: prwizard(ncovmodel, nlstate, ndeath, model, ficparo);
! 9906: printf(" You chose mle=-2, look at file %s for a template of covariance matrix \n",filereso);
! 9907: fprintf(ficlog," You chose mle=-2, look at file %s for a template of covariance matrix \n",filereso);
! 9908:
1.220 brouard 9909: } else if(mle==-5) { /* Main Wizard */
1.126 brouard 9910: prwizard(ncovmodel, nlstate, ndeath, model, ficparo);
1.192 brouard 9911: printf(" You chose mle=-3, look at file %s for a template of covariance matrix \n",filereso);
9912: fprintf(ficlog," You chose mle=-3, look at file %s for a template of covariance matrix \n",filereso);
1.126 brouard 9913: param= ma3x(1,nlstate,1,nlstate+ndeath-1,1,ncovmodel);
9914: matcov=matrix(1,npar,1,npar);
1.203 brouard 9915: hess=matrix(1,npar,1,npar);
1.220 brouard 9916: } else{ /* Begin of mle != -1 or -5 */
1.145 brouard 9917: /* Read guessed parameters */
1.126 brouard 9918: /* Reads comments: lines beginning with '#' */
9919: while((c=getc(ficpar))=='#' && c!= EOF){
9920: ungetc(c,ficpar);
9921: fgets(line, MAXLINE, ficpar);
9922: numlinepar++;
1.141 brouard 9923: fputs(line,stdout);
1.126 brouard 9924: fputs(line,ficparo);
9925: fputs(line,ficlog);
9926: }
9927: ungetc(c,ficpar);
9928:
9929: param= ma3x(1,nlstate,1,nlstate+ndeath-1,1,ncovmodel);
9930: for(i=1; i <=nlstate; i++){
1.234 brouard 9931: j=0;
1.126 brouard 9932: for(jj=1; jj <=nlstate+ndeath; jj++){
1.234 brouard 9933: if(jj==i) continue;
9934: j++;
9935: fscanf(ficpar,"%1d%1d",&i1,&j1);
9936: if ((i1 != i) || (j1 != jj)){
9937: printf("Error in line parameters number %d, %1d%1d instead of %1d%1d \n \
1.126 brouard 9938: It might be a problem of design; if ncovcol and the model are correct\n \
9939: run imach with mle=-1 to get a correct template of the parameter file.\n",numlinepar, i,j, i1, j1);
1.234 brouard 9940: exit(1);
9941: }
9942: fprintf(ficparo,"%1d%1d",i1,j1);
9943: if(mle==1)
9944: printf("%1d%1d",i,jj);
9945: fprintf(ficlog,"%1d%1d",i,jj);
9946: for(k=1; k<=ncovmodel;k++){
9947: fscanf(ficpar," %lf",¶m[i][j][k]);
9948: if(mle==1){
9949: printf(" %lf",param[i][j][k]);
9950: fprintf(ficlog," %lf",param[i][j][k]);
9951: }
9952: else
9953: fprintf(ficlog," %lf",param[i][j][k]);
9954: fprintf(ficparo," %lf",param[i][j][k]);
9955: }
9956: fscanf(ficpar,"\n");
9957: numlinepar++;
9958: if(mle==1)
9959: printf("\n");
9960: fprintf(ficlog,"\n");
9961: fprintf(ficparo,"\n");
1.126 brouard 9962: }
9963: }
9964: fflush(ficlog);
1.234 brouard 9965:
1.145 brouard 9966: /* Reads scales values */
1.126 brouard 9967: p=param[1][1];
9968:
9969: /* Reads comments: lines beginning with '#' */
9970: while((c=getc(ficpar))=='#' && c!= EOF){
9971: ungetc(c,ficpar);
9972: fgets(line, MAXLINE, ficpar);
9973: numlinepar++;
1.141 brouard 9974: fputs(line,stdout);
1.126 brouard 9975: fputs(line,ficparo);
9976: fputs(line,ficlog);
9977: }
9978: ungetc(c,ficpar);
9979:
9980: for(i=1; i <=nlstate; i++){
9981: for(j=1; j <=nlstate+ndeath-1; j++){
1.234 brouard 9982: fscanf(ficpar,"%1d%1d",&i1,&j1);
9983: if ( (i1-i) * (j1-j) != 0){
9984: printf("Error in line parameters number %d, %1d%1d instead of %1d%1d \n",numlinepar, i,j, i1, j1);
9985: exit(1);
9986: }
9987: printf("%1d%1d",i,j);
9988: fprintf(ficparo,"%1d%1d",i1,j1);
9989: fprintf(ficlog,"%1d%1d",i1,j1);
9990: for(k=1; k<=ncovmodel;k++){
9991: fscanf(ficpar,"%le",&delti3[i][j][k]);
9992: printf(" %le",delti3[i][j][k]);
9993: fprintf(ficparo," %le",delti3[i][j][k]);
9994: fprintf(ficlog," %le",delti3[i][j][k]);
9995: }
9996: fscanf(ficpar,"\n");
9997: numlinepar++;
9998: printf("\n");
9999: fprintf(ficparo,"\n");
10000: fprintf(ficlog,"\n");
1.126 brouard 10001: }
10002: }
10003: fflush(ficlog);
1.234 brouard 10004:
1.145 brouard 10005: /* Reads covariance matrix */
1.126 brouard 10006: delti=delti3[1][1];
1.220 brouard 10007:
10008:
1.126 brouard 10009: /* 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 10010:
1.126 brouard 10011: /* Reads comments: lines beginning with '#' */
10012: while((c=getc(ficpar))=='#' && c!= EOF){
10013: ungetc(c,ficpar);
10014: fgets(line, MAXLINE, ficpar);
10015: numlinepar++;
1.141 brouard 10016: fputs(line,stdout);
1.126 brouard 10017: fputs(line,ficparo);
10018: fputs(line,ficlog);
10019: }
10020: ungetc(c,ficpar);
1.220 brouard 10021:
1.126 brouard 10022: matcov=matrix(1,npar,1,npar);
1.203 brouard 10023: hess=matrix(1,npar,1,npar);
1.131 brouard 10024: for(i=1; i <=npar; i++)
10025: for(j=1; j <=npar; j++) matcov[i][j]=0.;
1.220 brouard 10026:
1.194 brouard 10027: /* Scans npar lines */
1.126 brouard 10028: for(i=1; i <=npar; i++){
1.226 brouard 10029: count=fscanf(ficpar,"%1d%1d%d",&i1,&j1,&jk);
1.194 brouard 10030: if(count != 3){
1.226 brouard 10031: printf("Error! Error in parameter file %s at line %d after line starting with %1d%1d%1d\n\
1.194 brouard 10032: This is probably because your covariance matrix doesn't \n contain exactly %d lines corresponding to your model line '1+age+%s'.\n\
10033: Please run with mle=-1 to get a correct covariance matrix.\n",optionfile,numlinepar, i1,j1,jk, npar, model);
1.226 brouard 10034: fprintf(ficlog,"Error! Error in parameter file %s at line %d after line starting with %1d%1d%1d\n\
1.194 brouard 10035: This is probably because your covariance matrix doesn't \n contain exactly %d lines corresponding to your model line '1+age+%s'.\n\
10036: Please run with mle=-1 to get a correct covariance matrix.\n",optionfile,numlinepar, i1,j1,jk, npar, model);
1.226 brouard 10037: exit(1);
1.220 brouard 10038: }else{
1.226 brouard 10039: if(mle==1)
10040: printf("%1d%1d%d",i1,j1,jk);
10041: }
10042: fprintf(ficlog,"%1d%1d%d",i1,j1,jk);
10043: fprintf(ficparo,"%1d%1d%d",i1,j1,jk);
1.126 brouard 10044: for(j=1; j <=i; j++){
1.226 brouard 10045: fscanf(ficpar," %le",&matcov[i][j]);
10046: if(mle==1){
10047: printf(" %.5le",matcov[i][j]);
10048: }
10049: fprintf(ficlog," %.5le",matcov[i][j]);
10050: fprintf(ficparo," %.5le",matcov[i][j]);
1.126 brouard 10051: }
10052: fscanf(ficpar,"\n");
10053: numlinepar++;
10054: if(mle==1)
1.220 brouard 10055: printf("\n");
1.126 brouard 10056: fprintf(ficlog,"\n");
10057: fprintf(ficparo,"\n");
10058: }
1.194 brouard 10059: /* End of read covariance matrix npar lines */
1.126 brouard 10060: for(i=1; i <=npar; i++)
10061: for(j=i+1;j<=npar;j++)
1.226 brouard 10062: matcov[i][j]=matcov[j][i];
1.126 brouard 10063:
10064: if(mle==1)
10065: printf("\n");
10066: fprintf(ficlog,"\n");
10067:
10068: fflush(ficlog);
10069:
10070: /*-------- Rewriting parameter file ----------*/
10071: strcpy(rfileres,"r"); /* "Rparameterfile */
10072: strcat(rfileres,optionfilefiname); /* Parameter file first name*/
10073: strcat(rfileres,"."); /* */
10074: strcat(rfileres,optionfilext); /* Other files have txt extension */
10075: if((ficres =fopen(rfileres,"w"))==NULL) {
1.201 brouard 10076: printf("Problem writing new parameter file: %s\n", rfileres);goto end;
10077: fprintf(ficlog,"Problem writing new parameter file: %s\n", rfileres);goto end;
1.126 brouard 10078: }
10079: fprintf(ficres,"#%s\n",version);
10080: } /* End of mle != -3 */
1.218 brouard 10081:
1.186 brouard 10082: /* Main data
10083: */
1.126 brouard 10084: n= lastobs;
10085: num=lvector(1,n);
10086: moisnais=vector(1,n);
10087: annais=vector(1,n);
10088: moisdc=vector(1,n);
10089: andc=vector(1,n);
1.220 brouard 10090: weight=vector(1,n);
1.126 brouard 10091: agedc=vector(1,n);
10092: cod=ivector(1,n);
1.220 brouard 10093: for(i=1;i<=n;i++){
1.234 brouard 10094: num[i]=0;
10095: moisnais[i]=0;
10096: annais[i]=0;
10097: moisdc[i]=0;
10098: andc[i]=0;
10099: agedc[i]=0;
10100: cod[i]=0;
10101: weight[i]=1.0; /* Equal weights, 1 by default */
10102: }
1.126 brouard 10103: mint=matrix(1,maxwav,1,n);
10104: anint=matrix(1,maxwav,1,n);
1.131 brouard 10105: s=imatrix(1,maxwav+1,1,n); /* s[i][j] health state for wave i and individual j */
1.126 brouard 10106: tab=ivector(1,NCOVMAX);
1.144 brouard 10107: ncodemax=ivector(1,NCOVMAX); /* Number of code per covariate; if O and 1 only, 2**ncov; V1+V2+V3+V4=>16 */
1.192 brouard 10108: 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 10109:
1.136 brouard 10110: /* Reads data from file datafile */
10111: if (readdata(datafile, firstobs, lastobs, &imx)==1)
10112: goto end;
10113:
10114: /* Calculation of the number of parameters from char model */
1.234 brouard 10115: /* modelsav=V2+V1+V4+age*V3 strb=age*V3 stra=V2+V1+V4
1.137 brouard 10116: k=4 (age*V3) Tvar[k=4]= 3 (from V3) Tag[cptcovage=1]=4
10117: k=3 V4 Tvar[k=3]= 4 (from V4)
10118: k=2 V1 Tvar[k=2]= 1 (from V1)
10119: k=1 Tvar[1]=2 (from V2)
1.234 brouard 10120: */
10121:
10122: Tvar=ivector(1,NCOVMAX); /* Was 15 changed to NCOVMAX. */
10123: TvarsDind=ivector(1,NCOVMAX); /* */
10124: TvarsD=ivector(1,NCOVMAX); /* */
10125: TvarsQind=ivector(1,NCOVMAX); /* */
10126: TvarsQ=ivector(1,NCOVMAX); /* */
1.232 brouard 10127: TvarF=ivector(1,NCOVMAX); /* */
10128: TvarFind=ivector(1,NCOVMAX); /* */
10129: TvarV=ivector(1,NCOVMAX); /* */
10130: TvarVind=ivector(1,NCOVMAX); /* */
10131: TvarA=ivector(1,NCOVMAX); /* */
10132: TvarAind=ivector(1,NCOVMAX); /* */
1.231 brouard 10133: TvarFD=ivector(1,NCOVMAX); /* */
10134: TvarFDind=ivector(1,NCOVMAX); /* */
10135: TvarFQ=ivector(1,NCOVMAX); /* */
10136: TvarFQind=ivector(1,NCOVMAX); /* */
10137: TvarVD=ivector(1,NCOVMAX); /* */
10138: TvarVDind=ivector(1,NCOVMAX); /* */
10139: TvarVQ=ivector(1,NCOVMAX); /* */
10140: TvarVQind=ivector(1,NCOVMAX); /* */
10141:
1.230 brouard 10142: Tvalsel=vector(1,NCOVMAX); /* */
1.233 brouard 10143: Tvarsel=ivector(1,NCOVMAX); /* */
1.226 brouard 10144: Typevar=ivector(-1,NCOVMAX); /* -1 to 2 */
10145: Fixed=ivector(-1,NCOVMAX); /* -1 to 3 */
10146: Dummy=ivector(-1,NCOVMAX); /* -1 to 3 */
1.137 brouard 10147: /* V2+V1+V4+age*V3 is a model with 4 covariates (3 plus signs).
10148: For each model-covariate stores the data-covariate id. Tvar[1]=2, Tvar[2]=1, Tvar[3]=4,
10149: Tvar[4=age*V3] is 3 and 'age' is recorded in Tage.
10150: */
10151: /* For model-covariate k tells which data-covariate to use but
10152: because this model-covariate is a construction we invent a new column
10153: ncovcol + k1
10154: If already ncovcol=4 and model=V2+V1+V1*V4+age*V3
10155: Tvar[3=V1*V4]=4+1 etc */
1.227 brouard 10156: Tprod=ivector(1,NCOVMAX); /* Gives the k position of the k1 product */
10157: Tposprod=ivector(1,NCOVMAX); /* Gives the k1 product from the k position */
1.137 brouard 10158: /* Tprod[k1=1]=3(=V1*V4) for V2+V1+V1*V4+age*V3
10159: if V2+V1+V1*V4+age*V3+V3*V2 TProd[k1=2]=5 (V3*V2)
1.227 brouard 10160: Tposprod[k]=k1 , Tposprod[3]=1, Tposprod[5]=2
1.137 brouard 10161: */
1.145 brouard 10162: Tvaraff=ivector(1,NCOVMAX); /* Unclear */
10163: 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 10164: * For V3*V2 (in V2+V1+V1*V4+age*V3+V3*V2), V3*V2 position is 2nd.
10165: * Tvard[k1=2][1]=3 (V3) Tvard[k1=2][2]=2(V2) */
1.145 brouard 10166: Tage=ivector(1,NCOVMAX); /* Gives the covariate id of covariates associated with age: V2 + V1 + age*V4 + V3*age
1.137 brouard 10167: 4 covariates (3 plus signs)
10168: Tage[1=V3*age]= 4; Tage[2=age*V4] = 3
10169: */
1.230 brouard 10170: Tmodelind=ivector(1,NCOVMAX);/** gives the k model position of an
1.227 brouard 10171: * individual dummy, fixed or varying:
10172: * Tmodelind[Tvaraff[3]]=9,Tvaraff[1]@9={4,
10173: * 3, 1, 0, 0, 0, 0, 0, 0},
1.230 brouard 10174: * model=V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1 ,
10175: * V1 df, V2 qf, V3 & V4 dv, V5 qv
10176: * Tmodelind[1]@9={9,0,3,2,}*/
10177: TmodelInvind=ivector(1,NCOVMAX); /* TmodelInvind=Tvar[k]- ncovcol-nqv={5-2-1=2,*/
10178: TmodelInvQind=ivector(1,NCOVMAX);/** gives the k model position of an
1.228 brouard 10179: * individual quantitative, fixed or varying:
10180: * Tmodelqind[1]=1,Tvaraff[1]@9={4,
10181: * 3, 1, 0, 0, 0, 0, 0, 0},
10182: * model=V5+V4+V3+V4*V3+V5*age+V2+V1*V2+V1*age+V1*/
1.186 brouard 10183: /* Main decodemodel */
10184:
1.187 brouard 10185:
1.223 brouard 10186: if(decodemodel(model, lastobs) == 1) /* In order to get Tvar[k] V4+V3+V5 p Tvar[1]@3 = {4, 3, 5}*/
1.136 brouard 10187: goto end;
10188:
1.137 brouard 10189: if((double)(lastobs-imx)/(double)imx > 1.10){
10190: nbwarn++;
10191: 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);
10192: 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);
10193: }
1.136 brouard 10194: /* if(mle==1){*/
1.137 brouard 10195: if (weightopt != 1) { /* Maximisation without weights. We can have weights different from 1 but want no weight*/
10196: for(i=1;i<=imx;i++) weight[i]=1.0; /* changed to imx */
1.136 brouard 10197: }
10198:
10199: /*-calculation of age at interview from date of interview and age at death -*/
10200: agev=matrix(1,maxwav,1,imx);
10201:
10202: if(calandcheckages(imx, maxwav, &agemin, &agemax, &nberr, &nbwarn) == 1)
10203: goto end;
10204:
1.126 brouard 10205:
1.136 brouard 10206: agegomp=(int)agemin;
10207: free_vector(moisnais,1,n);
10208: free_vector(annais,1,n);
1.126 brouard 10209: /* free_matrix(mint,1,maxwav,1,n);
10210: free_matrix(anint,1,maxwav,1,n);*/
1.215 brouard 10211: /* free_vector(moisdc,1,n); */
10212: /* free_vector(andc,1,n); */
1.145 brouard 10213: /* */
10214:
1.126 brouard 10215: wav=ivector(1,imx);
1.214 brouard 10216: /* dh=imatrix(1,lastpass-firstpass+1,1,imx); */
10217: /* bh=imatrix(1,lastpass-firstpass+1,1,imx); */
10218: /* mw=imatrix(1,lastpass-firstpass+1,1,imx); */
10219: 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.*/
10220: bh=imatrix(1,lastpass-firstpass+2,1,imx);
10221: mw=imatrix(1,lastpass-firstpass+2,1,imx);
1.126 brouard 10222:
10223: /* Concatenates waves */
1.214 brouard 10224: /* Concatenates waves: wav[i] is the number of effective (useful waves) of individual i.
10225: Death is a valid wave (if date is known).
10226: mw[mi][i] is the number of (mi=1 to wav[i]) effective wave out of mi of individual i
10227: dh[m][i] or dh[mw[mi][i]][i] is the delay between two effective waves m=mw[mi][i]
10228: and mw[mi+1][i]. dh depends on stepm.
10229: */
10230:
1.126 brouard 10231: concatwav(wav, dh, bh, mw, s, agedc, agev, firstpass, lastpass, imx, nlstate, stepm);
1.248 ! brouard 10232: /* Concatenates waves */
1.145 brouard 10233:
1.215 brouard 10234: free_vector(moisdc,1,n);
10235: free_vector(andc,1,n);
10236:
1.126 brouard 10237: /* Routine tricode is to calculate cptcoveff (real number of unique covariates) and to associate covariable number and modality */
10238: nbcode=imatrix(0,NCOVMAX,0,NCOVMAX);
10239: ncodemax[1]=1;
1.145 brouard 10240: Ndum =ivector(-1,NCOVMAX);
1.225 brouard 10241: cptcoveff=0;
1.220 brouard 10242: if (ncovmodel-nagesqr > 2 ){ /* That is if covariate other than cst, age and age*age */
10243: tricode(&cptcoveff,Tvar,nbcode,imx, Ndum); /**< Fills nbcode[Tvar[j]][l]; */
1.227 brouard 10244: }
10245:
10246: ncovcombmax=pow(2,cptcoveff);
10247: invalidvarcomb=ivector(1, ncovcombmax);
10248: for(i=1;i<ncovcombmax;i++)
10249: invalidvarcomb[i]=0;
10250:
1.211 brouard 10251: /* Nbcode gives the value of the lth modality (currently 1 to 2) of jth covariate, in
1.186 brouard 10252: V2+V1*age, there are 3 covariates Tvar[2]=1 (V1).*/
1.211 brouard 10253: /* 1 to ncodemax[j] which is the maximum value of this jth covariate */
1.227 brouard 10254:
1.200 brouard 10255: /* codtab=imatrix(1,100,1,10);*/ /* codtab[h,k]=( (h-1) - mod(k-1,2**(k-1) )/2**(k-1) */
1.198 brouard 10256: /*printf(" codtab[1,1],codtab[100,10]=%d,%d\n", codtab[1][1],codtabm(100,10));*/
1.186 brouard 10257: /* codtab gives the value 1 or 2 of the hth combination of k covariates (1 or 2).*/
1.211 brouard 10258: /* nbcode[Tvaraff[j]][codtabm(h,j)]) : if there are only 2 modalities for a covariate j,
10259: * codtabm(h,j) gives its value classified at position h and nbcode gives how it is coded
10260: * (currently 0 or 1) in the data.
10261: * In a loop on h=1 to 2**k, and a loop on j (=1 to k), we get the value of
10262: * corresponding modality (h,j).
10263: */
10264:
1.145 brouard 10265: h=0;
10266: /*if (cptcovn > 0) */
1.126 brouard 10267: m=pow(2,cptcoveff);
10268:
1.144 brouard 10269: /**< codtab(h,k) k = codtab[h,k]=( (h-1) - mod(k-1,2**(k-1) )/2**(k-1) + 1
1.211 brouard 10270: * For k=4 covariates, h goes from 1 to m=2**k
10271: * codtabm(h,k)= (1 & (h-1) >> (k-1)) + 1;
10272: * #define codtabm(h,k) (1 & (h-1) >> (k-1))+1
1.186 brouard 10273: * h\k 1 2 3 4
1.143 brouard 10274: *______________________________
10275: * 1 i=1 1 i=1 1 i=1 1 i=1 1
10276: * 2 2 1 1 1
10277: * 3 i=2 1 2 1 1
10278: * 4 2 2 1 1
10279: * 5 i=3 1 i=2 1 2 1
10280: * 6 2 1 2 1
10281: * 7 i=4 1 2 2 1
10282: * 8 2 2 2 1
1.197 brouard 10283: * 9 i=5 1 i=3 1 i=2 1 2
10284: * 10 2 1 1 2
10285: * 11 i=6 1 2 1 2
10286: * 12 2 2 1 2
10287: * 13 i=7 1 i=4 1 2 2
10288: * 14 2 1 2 2
10289: * 15 i=8 1 2 2 2
10290: * 16 2 2 2 2
1.143 brouard 10291: */
1.212 brouard 10292: /* How to do the opposite? From combination h (=1 to 2**k) how to get the value on the covariates? */
1.211 brouard 10293: /* from h=5 and m, we get then number of covariates k=log(m)/log(2)=4
10294: * and the value of each covariate?
10295: * V1=1, V2=1, V3=2, V4=1 ?
10296: * h-1=4 and 4 is 0100 or reverse 0010, and +1 is 1121 ok.
10297: * h=6, 6-1=5, 5 is 0101, 1010, 2121, V1=2nd, V2=1st, V3=2nd, V4=1st.
10298: * In order to get the real value in the data, we use nbcode
10299: * nbcode[Tvar[3][2nd]]=1 and nbcode[Tvar[4][1]]=0
10300: * We are keeping this crazy system in order to be able (in the future?)
10301: * to have more than 2 values (0 or 1) for a covariate.
10302: * #define codtabm(h,k) (1 & (h-1) >> (k-1))+1
10303: * h=6, k=2? h-1=5=0101, reverse 1010, +1=2121, k=2nd position: value is 1: codtabm(6,2)=1
10304: * bbbbbbbb
10305: * 76543210
10306: * h-1 00000101 (6-1=5)
1.219 brouard 10307: *(h-1)>>(k-1)= 00000010 >> (2-1) = 1 right shift
1.211 brouard 10308: * &
10309: * 1 00000001 (1)
1.219 brouard 10310: * 00000000 = 1 & ((h-1) >> (k-1))
10311: * +1= 00000001 =1
1.211 brouard 10312: *
10313: * h=14, k=3 => h'=h-1=13, k'=k-1=2
10314: * h' 1101 =2^3+2^2+0x2^1+2^0
10315: * >>k' 11
10316: * & 00000001
10317: * = 00000001
10318: * +1 = 00000010=2 = codtabm(14,3)
10319: * Reverse h=6 and m=16?
10320: * cptcoveff=log(16)/log(2)=4 covariate: 6-1=5=0101 reversed=1010 +1=2121 =>V1=2, V2=1, V3=2, V4=1.
10321: * for (j=1 to cptcoveff) Vj=decodtabm(j,h,cptcoveff)
10322: * decodtabm(h,j,cptcoveff)= (((h-1) >> (j-1)) & 1) +1
10323: * decodtabm(h,j,cptcoveff)= (h <= (1<<cptcoveff)?(((h-1) >> (j-1)) & 1) +1 : -1)
10324: * V3=decodtabm(14,3,2**4)=2
10325: * h'=13 1101 =2^3+2^2+0x2^1+2^0
10326: *(h-1) >> (j-1) 0011 =13 >> 2
10327: * &1 000000001
10328: * = 000000001
10329: * +1= 000000010 =2
10330: * 2211
10331: * V1=1+1, V2=0+1, V3=1+1, V4=1+1
10332: * V3=2
1.220 brouard 10333: * codtabm and decodtabm are identical
1.211 brouard 10334: */
10335:
1.145 brouard 10336:
10337: free_ivector(Ndum,-1,NCOVMAX);
10338:
10339:
1.126 brouard 10340:
1.186 brouard 10341: /* Initialisation of ----------- gnuplot -------------*/
1.126 brouard 10342: strcpy(optionfilegnuplot,optionfilefiname);
10343: if(mle==-3)
1.201 brouard 10344: strcat(optionfilegnuplot,"-MORT_");
1.126 brouard 10345: strcat(optionfilegnuplot,".gp");
10346:
10347: if((ficgp=fopen(optionfilegnuplot,"w"))==NULL) {
10348: printf("Problem with file %s",optionfilegnuplot);
10349: }
10350: else{
1.204 brouard 10351: fprintf(ficgp,"\n# IMaCh-%s\n", version);
1.126 brouard 10352: fprintf(ficgp,"# %s\n", optionfilegnuplot);
1.141 brouard 10353: //fprintf(ficgp,"set missing 'NaNq'\n");
10354: fprintf(ficgp,"set datafile missing 'NaNq'\n");
1.126 brouard 10355: }
10356: /* fclose(ficgp);*/
1.186 brouard 10357:
10358:
10359: /* Initialisation of --------- index.htm --------*/
1.126 brouard 10360:
10361: strcpy(optionfilehtm,optionfilefiname); /* Main html file */
10362: if(mle==-3)
1.201 brouard 10363: strcat(optionfilehtm,"-MORT_");
1.126 brouard 10364: strcat(optionfilehtm,".htm");
10365: if((fichtm=fopen(optionfilehtm,"w"))==NULL) {
1.131 brouard 10366: printf("Problem with %s \n",optionfilehtm);
10367: exit(0);
1.126 brouard 10368: }
10369:
10370: strcpy(optionfilehtmcov,optionfilefiname); /* Only for matrix of covariance */
10371: strcat(optionfilehtmcov,"-cov.htm");
10372: if((fichtmcov=fopen(optionfilehtmcov,"w"))==NULL) {
10373: printf("Problem with %s \n",optionfilehtmcov), exit(0);
10374: }
10375: else{
10376: fprintf(fichtmcov,"<html><head>\n<title>IMaCh Cov %s</title></head>\n <body><font size=\"2\">%s <br> %s</font> \
10377: <hr size=\"2\" color=\"#EC5E5E\"> \n\
1.204 brouard 10378: Title=%s <br>Datafile=%s Firstpass=%d Lastpass=%d Stepm=%d Weight=%d Model=1+age+%s<br>\n",\
1.126 brouard 10379: optionfilehtmcov,version,fullversion,title,datafile,firstpass,lastpass,stepm, weightopt, model);
10380: }
10381:
1.213 brouard 10382: 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 10383: <hr size=\"2\" color=\"#EC5E5E\"> \n\
10384: <font size=\"2\">IMaCh-%s <br> %s</font> \
1.126 brouard 10385: <hr size=\"2\" color=\"#EC5E5E\"> \n\
1.204 brouard 10386: Title=%s <br>Datafile=%s Firstpass=%d Lastpass=%d Stepm=%d Weight=%d Model=1+age+%s<br>\n\
1.126 brouard 10387: \n\
10388: <hr size=\"2\" color=\"#EC5E5E\">\
10389: <ul><li><h4>Parameter files</h4>\n\
10390: - Parameter file: <a href=\"%s.%s\">%s.%s</a><br>\n\
10391: - Copy of the parameter file: <a href=\"o%s\">o%s</a><br>\n\
10392: - Log file of the run: <a href=\"%s\">%s</a><br>\n\
10393: - Gnuplot file name: <a href=\"%s\">%s</a><br>\n\
10394: - Date and time at start: %s</ul>\n",\
10395: optionfilehtm,version,fullversion,title,datafile,firstpass,lastpass,stepm, weightopt, model,\
10396: optionfilefiname,optionfilext,optionfilefiname,optionfilext,\
10397: fileres,fileres,\
10398: filelog,filelog,optionfilegnuplot,optionfilegnuplot,strstart);
10399: fflush(fichtm);
10400:
10401: strcpy(pathr,path);
10402: strcat(pathr,optionfilefiname);
1.184 brouard 10403: #ifdef WIN32
10404: _chdir(optionfilefiname); /* Move to directory named optionfile */
10405: #else
1.126 brouard 10406: chdir(optionfilefiname); /* Move to directory named optionfile */
1.184 brouard 10407: #endif
10408:
1.126 brouard 10409:
1.220 brouard 10410: /* Calculates basic frequencies. Computes observed prevalence at single age
10411: and for any valid combination of covariates
1.126 brouard 10412: and prints on file fileres'p'. */
1.227 brouard 10413: freqsummary(fileres, agemin, agemax, s, agev, nlstate, imx, Tvaraff, invalidvarcomb, nbcode, ncodemax,mint,anint,strstart, \
10414: firstpass, lastpass, stepm, weightopt, model);
1.126 brouard 10415:
10416: fprintf(fichtm,"\n");
10417: fprintf(fichtm,"<br>Total number of observations=%d <br>\n\
10418: Youngest age at first (selected) pass %.2f, oldest age %.2f<br>\n\
10419: Interval (in months) between two waves: Min=%d Max=%d Mean=%.2lf<br>\n",\
10420: imx,agemin,agemax,jmin,jmax,jmean);
10421: pmmij= matrix(1,nlstate+ndeath,1,nlstate+ndeath); /* creation */
1.220 brouard 10422: oldms= matrix(1,nlstate+ndeath,1,nlstate+ndeath); /* creation */
10423: newms= matrix(1,nlstate+ndeath,1,nlstate+ndeath); /* creation */
10424: savms= matrix(1,nlstate+ndeath,1,nlstate+ndeath); /* creation */
10425: oldm=oldms; newm=newms; savm=savms; /* Keeps fixed addresses to free */
1.218 brouard 10426:
1.126 brouard 10427: /* For Powell, parameters are in a vector p[] starting at p[1]
10428: so we point p on param[1][1] so that p[1] maps on param[1][1][1] */
10429: p=param[1][1]; /* *(*(*(param +1)+1)+0) */
10430:
10431: globpr=0; /* To get the number ipmx of contributions and the sum of weights*/
1.186 brouard 10432: /* For mortality only */
1.126 brouard 10433: if (mle==-3){
1.136 brouard 10434: ximort=matrix(1,NDIM,1,NDIM);
1.248 ! brouard 10435: for(i=1;i<=NDIM;i++)
! 10436: for(j=1;j<=NDIM;j++)
! 10437: ximort[i][j]=0.;
1.186 brouard 10438: /* ximort=gsl_matrix_alloc(1,NDIM,1,NDIM); */
1.126 brouard 10439: cens=ivector(1,n);
10440: ageexmed=vector(1,n);
10441: agecens=vector(1,n);
10442: dcwave=ivector(1,n);
1.223 brouard 10443:
1.126 brouard 10444: for (i=1; i<=imx; i++){
10445: dcwave[i]=-1;
10446: for (m=firstpass; m<=lastpass; m++)
1.226 brouard 10447: if (s[m][i]>nlstate) {
10448: dcwave[i]=m;
10449: /* printf("i=%d j=%d s=%d dcwave=%d\n",i,j, s[j][i],dcwave[i]);*/
10450: break;
10451: }
1.126 brouard 10452: }
1.226 brouard 10453:
1.126 brouard 10454: for (i=1; i<=imx; i++) {
10455: if (wav[i]>0){
1.226 brouard 10456: ageexmed[i]=agev[mw[1][i]][i];
10457: j=wav[i];
10458: agecens[i]=1.;
10459:
10460: if (ageexmed[i]> 1 && wav[i] > 0){
10461: agecens[i]=agev[mw[j][i]][i];
10462: cens[i]= 1;
10463: }else if (ageexmed[i]< 1)
10464: cens[i]= -1;
10465: if (agedc[i]< AGESUP && agedc[i]>1 && dcwave[i]>firstpass && dcwave[i]<=lastpass)
10466: cens[i]=0 ;
1.126 brouard 10467: }
10468: else cens[i]=-1;
10469: }
10470:
10471: for (i=1;i<=NDIM;i++) {
10472: for (j=1;j<=NDIM;j++)
1.226 brouard 10473: ximort[i][j]=(i == j ? 1.0 : 0.0);
1.126 brouard 10474: }
10475:
1.145 brouard 10476: /*p[1]=0.0268; p[NDIM]=0.083;*/
1.126 brouard 10477: /*printf("%lf %lf", p[1], p[2]);*/
10478:
10479:
1.136 brouard 10480: #ifdef GSL
10481: printf("GSL optimization\n"); fprintf(ficlog,"Powell\n");
1.162 brouard 10482: #else
1.126 brouard 10483: printf("Powell\n"); fprintf(ficlog,"Powell\n");
1.136 brouard 10484: #endif
1.201 brouard 10485: strcpy(filerespow,"POW-MORT_");
10486: strcat(filerespow,fileresu);
1.126 brouard 10487: if((ficrespow=fopen(filerespow,"w"))==NULL) {
10488: printf("Problem with resultfile: %s\n", filerespow);
10489: fprintf(ficlog,"Problem with resultfile: %s\n", filerespow);
10490: }
1.136 brouard 10491: #ifdef GSL
10492: fprintf(ficrespow,"# GSL optimization\n# iter -2*LL");
1.162 brouard 10493: #else
1.126 brouard 10494: fprintf(ficrespow,"# Powell\n# iter -2*LL");
1.136 brouard 10495: #endif
1.126 brouard 10496: /* for (i=1;i<=nlstate;i++)
10497: for(j=1;j<=nlstate+ndeath;j++)
10498: if(j!=i)fprintf(ficrespow," p%1d%1d",i,j);
10499: */
10500: fprintf(ficrespow,"\n");
1.136 brouard 10501: #ifdef GSL
10502: /* gsl starts here */
10503: T = gsl_multimin_fminimizer_nmsimplex;
10504: gsl_multimin_fminimizer *sfm = NULL;
10505: gsl_vector *ss, *x;
10506: gsl_multimin_function minex_func;
10507:
10508: /* Initial vertex size vector */
10509: ss = gsl_vector_alloc (NDIM);
10510:
10511: if (ss == NULL){
10512: GSL_ERROR_VAL ("failed to allocate space for ss", GSL_ENOMEM, 0);
10513: }
10514: /* Set all step sizes to 1 */
10515: gsl_vector_set_all (ss, 0.001);
10516:
10517: /* Starting point */
1.126 brouard 10518:
1.136 brouard 10519: x = gsl_vector_alloc (NDIM);
10520:
10521: if (x == NULL){
10522: gsl_vector_free(ss);
10523: GSL_ERROR_VAL ("failed to allocate space for x", GSL_ENOMEM, 0);
10524: }
10525:
10526: /* Initialize method and iterate */
10527: /* p[1]=0.0268; p[NDIM]=0.083; */
1.186 brouard 10528: /* gsl_vector_set(x, 0, 0.0268); */
10529: /* gsl_vector_set(x, 1, 0.083); */
1.136 brouard 10530: gsl_vector_set(x, 0, p[1]);
10531: gsl_vector_set(x, 1, p[2]);
10532:
10533: minex_func.f = &gompertz_f;
10534: minex_func.n = NDIM;
10535: minex_func.params = (void *)&p; /* ??? */
10536:
10537: sfm = gsl_multimin_fminimizer_alloc (T, NDIM);
10538: gsl_multimin_fminimizer_set (sfm, &minex_func, x, ss);
10539:
10540: printf("Iterations beginning .....\n\n");
10541: printf("Iter. # Intercept Slope -Log Likelihood Simplex size\n");
10542:
10543: iteri=0;
10544: while (rval == GSL_CONTINUE){
10545: iteri++;
10546: status = gsl_multimin_fminimizer_iterate(sfm);
10547:
10548: if (status) printf("error: %s\n", gsl_strerror (status));
10549: fflush(0);
10550:
10551: if (status)
10552: break;
10553:
10554: rval = gsl_multimin_test_size (gsl_multimin_fminimizer_size (sfm), 1e-6);
10555: ssval = gsl_multimin_fminimizer_size (sfm);
10556:
10557: if (rval == GSL_SUCCESS)
10558: printf ("converged to a local maximum at\n");
10559:
10560: printf("%5d ", iteri);
10561: for (it = 0; it < NDIM; it++){
10562: printf ("%10.5f ", gsl_vector_get (sfm->x, it));
10563: }
10564: printf("f() = %-10.5f ssize = %.7f\n", sfm->fval, ssval);
10565: }
10566:
10567: printf("\n\n Please note: Program should be run many times with varying starting points to detemine global maximum\n\n");
10568:
10569: gsl_vector_free(x); /* initial values */
10570: gsl_vector_free(ss); /* inital step size */
10571: for (it=0; it<NDIM; it++){
10572: p[it+1]=gsl_vector_get(sfm->x,it);
10573: fprintf(ficrespow," %.12lf", p[it]);
10574: }
10575: gsl_multimin_fminimizer_free (sfm); /* p *(sfm.x.data) et p *(sfm.x.data+1) */
10576: #endif
10577: #ifdef POWELL
10578: powell(p,ximort,NDIM,ftol,&iter,&fret,gompertz);
10579: #endif
1.126 brouard 10580: fclose(ficrespow);
10581:
1.203 brouard 10582: hesscov(matcov, hess, p, NDIM, delti, 1e-4, gompertz);
1.126 brouard 10583:
10584: for(i=1; i <=NDIM; i++)
10585: for(j=i+1;j<=NDIM;j++)
1.220 brouard 10586: matcov[i][j]=matcov[j][i];
1.126 brouard 10587:
10588: printf("\nCovariance matrix\n ");
1.203 brouard 10589: fprintf(ficlog,"\nCovariance matrix\n ");
1.126 brouard 10590: for(i=1; i <=NDIM; i++) {
10591: for(j=1;j<=NDIM;j++){
1.220 brouard 10592: printf("%f ",matcov[i][j]);
10593: fprintf(ficlog,"%f ",matcov[i][j]);
1.126 brouard 10594: }
1.203 brouard 10595: printf("\n "); fprintf(ficlog,"\n ");
1.126 brouard 10596: }
10597:
10598: printf("iter=%d MLE=%f Eq=%lf*exp(%lf*(age-%d))\n",iter,-gompertz(p),p[1],p[2],agegomp);
1.193 brouard 10599: for (i=1;i<=NDIM;i++) {
1.126 brouard 10600: printf("%f [%f ; %f]\n",p[i],p[i]-2*sqrt(matcov[i][i]),p[i]+2*sqrt(matcov[i][i]));
1.193 brouard 10601: fprintf(ficlog,"%f [%f ; %f]\n",p[i],p[i]-2*sqrt(matcov[i][i]),p[i]+2*sqrt(matcov[i][i]));
10602: }
1.126 brouard 10603: lsurv=vector(1,AGESUP);
10604: lpop=vector(1,AGESUP);
10605: tpop=vector(1,AGESUP);
10606: lsurv[agegomp]=100000;
10607:
10608: for (k=agegomp;k<=AGESUP;k++) {
10609: agemortsup=k;
10610: if (p[1]*exp(p[2]*(k-agegomp))>1) break;
10611: }
10612:
10613: for (k=agegomp;k<agemortsup;k++)
10614: lsurv[k+1]=lsurv[k]-lsurv[k]*(p[1]*exp(p[2]*(k-agegomp)));
10615:
10616: for (k=agegomp;k<agemortsup;k++){
10617: lpop[k]=(lsurv[k]+lsurv[k+1])/2.;
10618: sumlpop=sumlpop+lpop[k];
10619: }
10620:
10621: tpop[agegomp]=sumlpop;
10622: for (k=agegomp;k<(agemortsup-3);k++){
10623: /* tpop[k+1]=2;*/
10624: tpop[k+1]=tpop[k]-lpop[k];
10625: }
10626:
10627:
10628: printf("\nAge lx qx dx Lx Tx e(x)\n");
10629: for (k=agegomp;k<(agemortsup-2);k++)
10630: 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]);
10631:
10632:
10633: replace_back_to_slash(pathc,pathcd); /* Even gnuplot wants a / */
1.220 brouard 10634: ageminpar=50;
10635: agemaxpar=100;
1.194 brouard 10636: if(ageminpar == AGEOVERFLOW ||agemaxpar == AGEOVERFLOW){
10637: printf("Warning! Error in gnuplot file with ageminpar %f or agemaxpar %f overflow\n\
10638: This is probably because your parameter file doesn't \n contain the exact number of lines (or columns) corresponding to your model line.\n\
10639: Please run with mle=-1 to get a correct covariance matrix.\n",ageminpar,agemaxpar);
10640: fprintf(ficlog,"Warning! Error in gnuplot file with ageminpar %f or agemaxpar %f overflow\n\
10641: This is probably because your parameter file doesn't \n contain the exact number of lines (or columns) corresponding to your model line.\n\
10642: Please run with mle=-1 to get a correct covariance matrix.\n",ageminpar,agemaxpar);
1.220 brouard 10643: }else{
10644: printf("Warning! ageminpar %f and agemaxpar %f have been fixed because for simplification until it is fixed...\n\n",ageminpar,agemaxpar);
10645: 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 10646: printinggnuplotmort(fileresu, optionfilefiname,ageminpar,agemaxpar,fage, pathc,p);
1.220 brouard 10647: }
1.201 brouard 10648: printinghtmlmort(fileresu,title,datafile, firstpass, lastpass, \
1.126 brouard 10649: stepm, weightopt,\
10650: model,imx,p,matcov,agemortsup);
10651:
10652: free_vector(lsurv,1,AGESUP);
10653: free_vector(lpop,1,AGESUP);
10654: free_vector(tpop,1,AGESUP);
1.220 brouard 10655: free_matrix(ximort,1,NDIM,1,NDIM);
1.136 brouard 10656: free_ivector(cens,1,n);
10657: free_vector(agecens,1,n);
10658: free_ivector(dcwave,1,n);
1.220 brouard 10659: #ifdef GSL
1.136 brouard 10660: #endif
1.186 brouard 10661: } /* Endof if mle==-3 mortality only */
1.205 brouard 10662: /* Standard */
10663: else{ /* For mle !=- 3, could be 0 or 1 or 4 etc. */
10664: globpr=0;/* Computes sum of likelihood for globpr=1 and funcone */
10665: /* Computes likelihood for initial parameters, uses funcone to compute gpimx and gsw */
1.132 brouard 10666: likelione(ficres, p, npar, nlstate, &globpr, &ipmx, &sw, &fretone, funcone); /* Prints the contributions to the likelihood */
1.126 brouard 10667: printf("First Likeli=%12.6f ipmx=%ld sw=%12.6f",fretone,ipmx,sw);
10668: for (k=1; k<=npar;k++)
10669: printf(" %d %8.5f",k,p[k]);
10670: printf("\n");
1.205 brouard 10671: if(mle>=1){ /* Could be 1 or 2, Real Maximization */
10672: /* mlikeli uses func not funcone */
1.247 brouard 10673: /* for(i=1;i<nlstate;i++){ */
10674: /* /\*reducing xi for 1 to npar to 1 to ncovmodel; *\/ */
10675: /* mlikeli(ficres,p, ncovmodel, ncovmodel, nlstate, ftol, funcnoprod); */
10676: /* } */
1.205 brouard 10677: mlikeli(ficres,p, npar, ncovmodel, nlstate, ftol, func);
10678: }
10679: if(mle==0) {/* No optimization, will print the likelihoods for the datafile */
10680: globpr=0;/* Computes sum of likelihood for globpr=1 and funcone */
10681: /* Computes likelihood for initial parameters, uses funcone to compute gpimx and gsw */
10682: likelione(ficres, p, npar, nlstate, &globpr, &ipmx, &sw, &fretone, funcone); /* Prints the contributions to the likelihood */
10683: }
10684: globpr=1; /* again, to print the individual contributions using computed gpimx and gsw */
1.126 brouard 10685: likelione(ficres, p, npar, nlstate, &globpr, &ipmx, &sw, &fretone, funcone); /* Prints the contributions to the likelihood */
10686: printf("Second Likeli=%12.6f ipmx=%ld sw=%12.6f",fretone,ipmx,sw);
10687: for (k=1; k<=npar;k++)
10688: printf(" %d %8.5f",k,p[k]);
10689: printf("\n");
10690:
10691: /*--------- results files --------------*/
1.224 brouard 10692: 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 10693:
10694:
10695: fprintf(ficres,"# Parameters nlstate*nlstate*ncov a12*1 + b12 * age + ...\n");
10696: printf("# Parameters nlstate*nlstate*ncov a12*1 + b12 * age + ...\n");
10697: fprintf(ficlog,"# Parameters nlstate*nlstate*ncov a12*1 + b12 * age + ...\n");
10698: for(i=1,jk=1; i <=nlstate; i++){
10699: for(k=1; k <=(nlstate+ndeath); k++){
1.225 brouard 10700: if (k != i) {
10701: printf("%d%d ",i,k);
10702: fprintf(ficlog,"%d%d ",i,k);
10703: fprintf(ficres,"%1d%1d ",i,k);
10704: for(j=1; j <=ncovmodel; j++){
10705: printf("%12.7f ",p[jk]);
10706: fprintf(ficlog,"%12.7f ",p[jk]);
10707: fprintf(ficres,"%12.7f ",p[jk]);
10708: jk++;
10709: }
10710: printf("\n");
10711: fprintf(ficlog,"\n");
10712: fprintf(ficres,"\n");
10713: }
1.126 brouard 10714: }
10715: }
1.203 brouard 10716: if(mle != 0){
10717: /* Computing hessian and covariance matrix only at a peak of the Likelihood, that is after optimization */
1.126 brouard 10718: ftolhess=ftol; /* Usually correct */
1.203 brouard 10719: hesscov(matcov, hess, p, npar, delti, ftolhess, func);
10720: 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");
10721: 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");
10722: for(i=1,jk=1; i <=nlstate; i++){
1.225 brouard 10723: for(k=1; k <=(nlstate+ndeath); k++){
10724: if (k != i) {
10725: printf("%d%d ",i,k);
10726: fprintf(ficlog,"%d%d ",i,k);
10727: for(j=1; j <=ncovmodel; j++){
10728: 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]));
10729: 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]));
10730: jk++;
10731: }
10732: printf("\n");
10733: fprintf(ficlog,"\n");
10734: }
10735: }
1.193 brouard 10736: }
1.203 brouard 10737: } /* end of hesscov and Wald tests */
1.225 brouard 10738:
1.203 brouard 10739: /* */
1.126 brouard 10740: fprintf(ficres,"# Scales (for hessian or gradient estimation)\n");
10741: printf("# Scales (for hessian or gradient estimation)\n");
10742: fprintf(ficlog,"# Scales (for hessian or gradient estimation)\n");
10743: for(i=1,jk=1; i <=nlstate; i++){
10744: for(j=1; j <=nlstate+ndeath; j++){
1.225 brouard 10745: if (j!=i) {
10746: fprintf(ficres,"%1d%1d",i,j);
10747: printf("%1d%1d",i,j);
10748: fprintf(ficlog,"%1d%1d",i,j);
10749: for(k=1; k<=ncovmodel;k++){
10750: printf(" %.5e",delti[jk]);
10751: fprintf(ficlog," %.5e",delti[jk]);
10752: fprintf(ficres," %.5e",delti[jk]);
10753: jk++;
10754: }
10755: printf("\n");
10756: fprintf(ficlog,"\n");
10757: fprintf(ficres,"\n");
10758: }
1.126 brouard 10759: }
10760: }
10761:
10762: 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 10763: if(mle >= 1) /* To big for the screen */
1.126 brouard 10764: 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");
10765: 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");
10766: /* # 121 Var(a12)\n\ */
10767: /* # 122 Cov(b12,a12) Var(b12)\n\ */
10768: /* # 131 Cov(a13,a12) Cov(a13,b12, Var(a13)\n\ */
10769: /* # 132 Cov(b13,a12) Cov(b13,b12, Cov(b13,a13) Var(b13)\n\ */
10770: /* # 212 Cov(a21,a12) Cov(a21,b12, Cov(a21,a13) Cov(a21,b13) Var(a21)\n\ */
10771: /* # 212 Cov(b21,a12) Cov(b21,b12, Cov(b21,a13) Cov(b21,b13) Cov(b21,a21) Var(b21)\n\ */
10772: /* # 232 Cov(a23,a12) Cov(a23,b12, Cov(a23,a13) Cov(a23,b13) Cov(a23,a21) Cov(a23,b21) Var(a23)\n\ */
10773: /* # 232 Cov(b23,a12) Cov(b23,b12) ... Var (b23)\n" */
10774:
10775:
10776: /* Just to have a covariance matrix which will be more understandable
10777: even is we still don't want to manage dictionary of variables
10778: */
10779: for(itimes=1;itimes<=2;itimes++){
10780: jj=0;
10781: for(i=1; i <=nlstate; i++){
1.225 brouard 10782: for(j=1; j <=nlstate+ndeath; j++){
10783: if(j==i) continue;
10784: for(k=1; k<=ncovmodel;k++){
10785: jj++;
10786: ca[0]= k+'a'-1;ca[1]='\0';
10787: if(itimes==1){
10788: if(mle>=1)
10789: printf("#%1d%1d%d",i,j,k);
10790: fprintf(ficlog,"#%1d%1d%d",i,j,k);
10791: fprintf(ficres,"#%1d%1d%d",i,j,k);
10792: }else{
10793: if(mle>=1)
10794: printf("%1d%1d%d",i,j,k);
10795: fprintf(ficlog,"%1d%1d%d",i,j,k);
10796: fprintf(ficres,"%1d%1d%d",i,j,k);
10797: }
10798: ll=0;
10799: for(li=1;li <=nlstate; li++){
10800: for(lj=1;lj <=nlstate+ndeath; lj++){
10801: if(lj==li) continue;
10802: for(lk=1;lk<=ncovmodel;lk++){
10803: ll++;
10804: if(ll<=jj){
10805: cb[0]= lk +'a'-1;cb[1]='\0';
10806: if(ll<jj){
10807: if(itimes==1){
10808: if(mle>=1)
10809: printf(" Cov(%s%1d%1d,%s%1d%1d)",ca,i,j,cb, li,lj);
10810: fprintf(ficlog," Cov(%s%1d%1d,%s%1d%1d)",ca,i,j,cb, li,lj);
10811: fprintf(ficres," Cov(%s%1d%1d,%s%1d%1d)",ca,i,j,cb, li,lj);
10812: }else{
10813: if(mle>=1)
10814: printf(" %.5e",matcov[jj][ll]);
10815: fprintf(ficlog," %.5e",matcov[jj][ll]);
10816: fprintf(ficres," %.5e",matcov[jj][ll]);
10817: }
10818: }else{
10819: if(itimes==1){
10820: if(mle>=1)
10821: printf(" Var(%s%1d%1d)",ca,i,j);
10822: fprintf(ficlog," Var(%s%1d%1d)",ca,i,j);
10823: fprintf(ficres," Var(%s%1d%1d)",ca,i,j);
10824: }else{
10825: if(mle>=1)
10826: printf(" %.7e",matcov[jj][ll]);
10827: fprintf(ficlog," %.7e",matcov[jj][ll]);
10828: fprintf(ficres," %.7e",matcov[jj][ll]);
10829: }
10830: }
10831: }
10832: } /* end lk */
10833: } /* end lj */
10834: } /* end li */
10835: if(mle>=1)
10836: printf("\n");
10837: fprintf(ficlog,"\n");
10838: fprintf(ficres,"\n");
10839: numlinepar++;
10840: } /* end k*/
10841: } /*end j */
1.126 brouard 10842: } /* end i */
10843: } /* end itimes */
10844:
10845: fflush(ficlog);
10846: fflush(ficres);
1.225 brouard 10847: while(fgets(line, MAXLINE, ficpar)) {
10848: /* If line starts with a # it is a comment */
10849: if (line[0] == '#') {
10850: numlinepar++;
10851: fputs(line,stdout);
10852: fputs(line,ficparo);
10853: fputs(line,ficlog);
10854: continue;
10855: }else
10856: break;
10857: }
10858:
1.209 brouard 10859: /* while((c=getc(ficpar))=='#' && c!= EOF){ */
10860: /* ungetc(c,ficpar); */
10861: /* fgets(line, MAXLINE, ficpar); */
10862: /* fputs(line,stdout); */
10863: /* fputs(line,ficparo); */
10864: /* } */
10865: /* ungetc(c,ficpar); */
1.126 brouard 10866:
10867: estepm=0;
1.209 brouard 10868: 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 10869:
10870: if (num_filled != 6) {
10871: 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);
10872: 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);
10873: goto end;
10874: }
10875: printf("agemin=%lf agemax=%lf bage=%lf fage=%lf estepm=%d ftolpl=%lf\n",ageminpar,agemaxpar, bage, fage, estepm, ftolpl);
10876: }
10877: /* ftolpl=6*ftol*1.e5; /\* 6.e-3 make convergences in less than 80 loops for the prevalence limit *\/ */
10878: /*ftolpl=6.e-4;*/ /* 6.e-3 make convergences in less than 80 loops for the prevalence limit */
10879:
1.209 brouard 10880: /* fscanf(ficpar,"agemin=%lf agemax=%lf bage=%lf fage=%lf estepm=%d ftolpl=%\n",&ageminpar,&agemaxpar, &bage, &fage, &estepm); */
1.126 brouard 10881: if (estepm==0 || estepm < stepm) estepm=stepm;
10882: if (fage <= 2) {
10883: bage = ageminpar;
10884: fage = agemaxpar;
10885: }
10886:
10887: fprintf(ficres,"# agemin agemax for life expectancy, bage fage (if mle==0 ie no data nor Max likelihood).\n");
1.211 brouard 10888: fprintf(ficres,"agemin=%.0f agemax=%.0f bage=%.0f fage=%.0f estepm=%d ftolpl=%e\n",ageminpar,agemaxpar,bage,fage, estepm, ftolpl);
10889: fprintf(ficparo,"agemin=%.0f agemax=%.0f bage=%.0f fage=%.0f estepm=%d, ftolpl=%e\n",ageminpar,agemaxpar,bage,fage, estepm, ftolpl);
1.220 brouard 10890:
1.186 brouard 10891: /* Other stuffs, more or less useful */
1.126 brouard 10892: while((c=getc(ficpar))=='#' && c!= EOF){
10893: ungetc(c,ficpar);
10894: fgets(line, MAXLINE, ficpar);
1.141 brouard 10895: fputs(line,stdout);
1.126 brouard 10896: fputs(line,ficparo);
10897: }
10898: ungetc(c,ficpar);
10899:
10900: fscanf(ficpar,"begin-prev-date=%lf/%lf/%lf end-prev-date=%lf/%lf/%lf mov_average=%d\n",&jprev1, &mprev1,&anprev1,&jprev2, &mprev2,&anprev2,&mobilav);
10901: 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);
10902: 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);
10903: printf("begin-prev-date=%.lf/%.lf/%.lf end-prev-date=%.lf/%.lf/%.lf mov_average=%d\n",jprev1, mprev1,anprev1,jprev2, mprev2,anprev2,mobilav);
10904: 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);
10905:
10906: while((c=getc(ficpar))=='#' && c!= EOF){
10907: ungetc(c,ficpar);
10908: fgets(line, MAXLINE, ficpar);
1.141 brouard 10909: fputs(line,stdout);
1.126 brouard 10910: fputs(line,ficparo);
10911: }
10912: ungetc(c,ficpar);
10913:
10914:
10915: dateprev1=anprev1+(mprev1-1)/12.+(jprev1-1)/365.;
10916: dateprev2=anprev2+(mprev2-1)/12.+(jprev2-1)/365.;
10917:
10918: fscanf(ficpar,"pop_based=%d\n",&popbased);
1.193 brouard 10919: fprintf(ficlog,"pop_based=%d\n",popbased);
1.126 brouard 10920: fprintf(ficparo,"pop_based=%d\n",popbased);
10921: fprintf(ficres,"pop_based=%d\n",popbased);
10922:
10923: while((c=getc(ficpar))=='#' && c!= EOF){
10924: ungetc(c,ficpar);
10925: fgets(line, MAXLINE, ficpar);
1.141 brouard 10926: fputs(line,stdout);
1.238 brouard 10927: fputs(line,ficres);
1.126 brouard 10928: fputs(line,ficparo);
10929: }
10930: ungetc(c,ficpar);
10931:
10932: fscanf(ficpar,"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);
10933: 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);
10934: 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);
10935: 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);
10936: 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);
10937: /* day and month of proj2 are not used but only year anproj2.*/
10938:
1.217 brouard 10939: while((c=getc(ficpar))=='#' && c!= EOF){
10940: ungetc(c,ficpar);
10941: fgets(line, MAXLINE, ficpar);
10942: fputs(line,stdout);
10943: fputs(line,ficparo);
1.238 brouard 10944: fputs(line,ficres);
1.217 brouard 10945: }
10946: ungetc(c,ficpar);
10947:
10948: 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);
1.223 brouard 10949: 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);
10950: 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);
10951: 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);
1.217 brouard 10952: /* day and month of proj2 are not used but only year anproj2.*/
1.126 brouard 10953:
1.230 brouard 10954: /* Results */
1.235 brouard 10955: nresult=0;
1.230 brouard 10956: while(fgets(line, MAXLINE, ficpar)) {
10957: /* If line starts with a # it is a comment */
10958: if (line[0] == '#') {
10959: numlinepar++;
10960: fputs(line,stdout);
10961: fputs(line,ficparo);
10962: fputs(line,ficlog);
1.238 brouard 10963: fputs(line,ficres);
1.230 brouard 10964: continue;
10965: }else
10966: break;
10967: }
1.240 brouard 10968: if (!feof(ficpar))
1.230 brouard 10969: while((num_filled=sscanf(line,"result:%[^\n]\n",resultline)) !=EOF){
1.240 brouard 10970: if (num_filled == 0){
1.230 brouard 10971: resultline[0]='\0';
1.240 brouard 10972: break;
10973: } else if (num_filled != 1){
1.230 brouard 10974: printf("ERROR %d: result line should be at minimum 'result=' %s\n",num_filled, line);
10975: }
1.235 brouard 10976: nresult++; /* Sum of resultlines */
10977: printf("Result %d: result=%s\n",nresult, resultline);
10978: if(nresult > MAXRESULTLINES){
10979: printf("ERROR: Current version of IMaCh limits the number of resultlines to %d, you used %d\n",MAXRESULTLINES,nresult);
10980: fprintf(ficlog,"ERROR: Current version of IMaCh limits the number of resultlines to %d, you used %d\n",MAXRESULTLINES,nresult);
10981: goto end;
10982: }
10983: decoderesult(resultline, nresult); /* Fills TKresult[nresult] combination and Tresult[nresult][k4+1] combination values */
1.238 brouard 10984: fprintf(ficparo,"result: %s\n",resultline);
10985: fprintf(ficres,"result: %s\n",resultline);
10986: fprintf(ficlog,"result: %s\n",resultline);
1.230 brouard 10987: while(fgets(line, MAXLINE, ficpar)) {
10988: /* If line starts with a # it is a comment */
10989: if (line[0] == '#') {
10990: numlinepar++;
10991: fputs(line,stdout);
10992: fputs(line,ficparo);
1.238 brouard 10993: fputs(line,ficres);
1.230 brouard 10994: fputs(line,ficlog);
10995: continue;
10996: }else
10997: break;
10998: }
10999: if (feof(ficpar))
11000: break;
11001: else{ /* Processess output results for this combination of covariate values */
11002: }
1.240 brouard 11003: } /* end while */
1.230 brouard 11004:
11005:
1.126 brouard 11006:
1.230 brouard 11007: /* freqsummary(fileres, agemin, agemax, s, agev, nlstate, imx,Tvaraff,nbcode, ncodemax,mint,anint); */
1.145 brouard 11008: /* ,dateprev1,dateprev2,jprev1, mprev1,anprev1,jprev2, mprev2,anprev2); */
1.126 brouard 11009:
11010: replace_back_to_slash(pathc,pathcd); /* Even gnuplot wants a / */
1.194 brouard 11011: if(ageminpar == AGEOVERFLOW ||agemaxpar == -AGEOVERFLOW){
1.230 brouard 11012: printf("Warning! Error in gnuplot file with ageminpar %f or agemaxpar %f overflow\n\
1.194 brouard 11013: This is probably because your parameter file doesn't \n contain the exact number of lines (or columns) corresponding to your model line.\n\
11014: Please run with mle=-1 to get a correct covariance matrix.\n",ageminpar,agemaxpar);
1.230 brouard 11015: fprintf(ficlog,"Warning! Error in gnuplot file with ageminpar %f or agemaxpar %f overflow\n\
1.194 brouard 11016: This is probably because your parameter file doesn't \n contain the exact number of lines (or columns) corresponding to your model line.\n\
11017: Please run with mle=-1 to get a correct covariance matrix.\n",ageminpar,agemaxpar);
1.220 brouard 11018: }else{
1.218 brouard 11019: printinggnuplot(fileresu, optionfilefiname,ageminpar,agemaxpar,fage, prevfcast, backcast, pathc,p);
1.220 brouard 11020: }
11021: printinghtml(fileresu,title,datafile, firstpass, lastpass, stepm, weightopt, \
1.225 brouard 11022: model,imx,jmin,jmax,jmean,rfileres,popforecast,prevfcast,backcast, estepm, \
11023: jprev1,mprev1,anprev1,dateprev1,jprev2,mprev2,anprev2,dateprev2);
1.220 brouard 11024:
1.225 brouard 11025: /*------------ free_vector -------------*/
11026: /* chdir(path); */
1.220 brouard 11027:
1.215 brouard 11028: /* free_ivector(wav,1,imx); */ /* Moved after last prevalence call */
11029: /* free_imatrix(dh,1,lastpass-firstpass+2,1,imx); */
11030: /* free_imatrix(bh,1,lastpass-firstpass+2,1,imx); */
11031: /* free_imatrix(mw,1,lastpass-firstpass+2,1,imx); */
1.126 brouard 11032: free_lvector(num,1,n);
11033: free_vector(agedc,1,n);
11034: /*free_matrix(covar,0,NCOVMAX,1,n);*/
11035: /*free_matrix(covar,1,NCOVMAX,1,n);*/
11036: fclose(ficparo);
11037: fclose(ficres);
1.220 brouard 11038:
11039:
1.186 brouard 11040: /* Other results (useful)*/
1.220 brouard 11041:
11042:
1.126 brouard 11043: /*--------------- Prevalence limit (period or stable prevalence) --------------*/
1.180 brouard 11044: /*#include "prevlim.h"*/ /* Use ficrespl, ficlog */
11045: prlim=matrix(1,nlstate,1,nlstate);
1.209 brouard 11046: prevalence_limit(p, prlim, ageminpar, agemaxpar, ftolpl, &ncvyear);
1.126 brouard 11047: fclose(ficrespl);
11048:
11049: /*------------- h Pij x at various ages ------------*/
1.180 brouard 11050: /*#include "hpijx.h"*/
11051: hPijx(p, bage, fage);
1.145 brouard 11052: fclose(ficrespij);
1.227 brouard 11053:
1.220 brouard 11054: /* ncovcombmax= pow(2,cptcoveff); */
1.219 brouard 11055: /*-------------- Variance of one-step probabilities---*/
1.145 brouard 11056: k=1;
1.126 brouard 11057: varprob(optionfilefiname, matcov, p, delti, nlstate, bage, fage,k,Tvar,nbcode, ncodemax,strstart);
1.227 brouard 11058:
1.219 brouard 11059: /* Prevalence for each covariates in probs[age][status][cov] */
1.218 brouard 11060: probs= ma3x(1,AGESUP,1,nlstate+ndeath, 1,ncovcombmax);
1.126 brouard 11061: for(i=1;i<=AGESUP;i++)
1.219 brouard 11062: for(j=1;j<=nlstate+ndeath;j++) /* ndeath is useless but a necessity to be compared with mobaverages */
1.225 brouard 11063: for(k=1;k<=ncovcombmax;k++)
11064: probs[i][j][k]=0.;
1.219 brouard 11065: prevalence(probs, ageminpar, agemaxpar, s, agev, nlstate, imx, Tvar, nbcode, ncodemax, mint, anint, dateprev1, dateprev2, firstpass, lastpass);
11066: if (mobilav!=0 ||mobilavproj !=0 ) {
11067: mobaverages= ma3x(1, AGESUP,1,nlstate+ndeath, 1,ncovcombmax);
1.227 brouard 11068: for(i=1;i<=AGESUP;i++)
11069: for(j=1;j<=nlstate;j++)
11070: for(k=1;k<=ncovcombmax;k++)
11071: mobaverages[i][j][k]=0.;
1.219 brouard 11072: mobaverage=mobaverages;
11073: if (mobilav!=0) {
1.235 brouard 11074: printf("Movingaveraging observed prevalence\n");
1.227 brouard 11075: if (movingaverage(probs, ageminpar, agemaxpar, mobaverage, mobilav)!=0){
11076: fprintf(ficlog," Error in movingaverage mobilav=%d\n",mobilav);
11077: printf(" Error in movingaverage mobilav=%d\n",mobilav);
11078: }
1.219 brouard 11079: }
11080: /* /\* Prevalence for each covariates in probs[age][status][cov] *\/ */
11081: /* prevalence(probs, ageminpar, agemaxpar, s, agev, nlstate, imx, Tvar, nbcode, ncodemax, mint, anint, dateprev1, dateprev2, firstpass, lastpass); */
11082: else if (mobilavproj !=0) {
1.235 brouard 11083: printf("Movingaveraging projected observed prevalence\n");
1.227 brouard 11084: if (movingaverage(probs, ageminpar, agemaxpar, mobaverage, mobilavproj)!=0){
11085: fprintf(ficlog," Error in movingaverage mobilavproj=%d\n",mobilavproj);
11086: printf(" Error in movingaverage mobilavproj=%d\n",mobilavproj);
11087: }
1.219 brouard 11088: }
11089: }/* end if moving average */
1.227 brouard 11090:
1.126 brouard 11091: /*---------- Forecasting ------------------*/
11092: /*if((stepm == 1) && (strcmp(model,".")==0)){*/
11093: if(prevfcast==1){
11094: /* if(stepm ==1){*/
1.225 brouard 11095: prevforecast(fileresu, anproj1, mproj1, jproj1, agemin, agemax, dateprev1, dateprev2, mobilavproj, bage, fage, firstpass, lastpass, anproj2, p, cptcoveff);
1.126 brouard 11096: }
1.217 brouard 11097: if(backcast==1){
1.219 brouard 11098: ddnewms=matrix(1,nlstate+ndeath,1,nlstate+ndeath);
11099: ddoldms=matrix(1,nlstate+ndeath,1,nlstate+ndeath);
11100: ddsavms=matrix(1,nlstate+ndeath,1,nlstate+ndeath);
11101:
11102: /*--------------- Back Prevalence limit (period or stable prevalence) --------------*/
11103:
11104: bprlim=matrix(1,nlstate,1,nlstate);
11105: back_prevalence_limit(p, bprlim, ageminpar, agemaxpar, ftolpl, &ncvyear, dateprev1, dateprev2, firstpass, lastpass, mobilavproj);
11106: fclose(ficresplb);
11107:
1.222 brouard 11108: hBijx(p, bage, fage, mobaverage);
11109: fclose(ficrespijb);
1.219 brouard 11110: free_matrix(bprlim,1,nlstate,1,nlstate); /*here or after loop ? */
11111:
11112: /* prevbackforecast(fileresu, anback1, mback1, jback1, agemin, agemax, dateprev1, dateprev2, mobilavproj,
1.225 brouard 11113: bage, fage, firstpass, lastpass, anback2, p, cptcoveff); */
1.219 brouard 11114: free_matrix(ddnewms, 1, nlstate+ndeath, 1, nlstate+ndeath);
11115: free_matrix(ddsavms, 1, nlstate+ndeath, 1, nlstate+ndeath);
11116: free_matrix(ddoldms, 1, nlstate+ndeath, 1, nlstate+ndeath);
11117: }
1.217 brouard 11118:
1.186 brouard 11119:
11120: /* ------ Other prevalence ratios------------ */
1.126 brouard 11121:
1.215 brouard 11122: free_ivector(wav,1,imx);
11123: free_imatrix(dh,1,lastpass-firstpass+2,1,imx);
11124: free_imatrix(bh,1,lastpass-firstpass+2,1,imx);
11125: free_imatrix(mw,1,lastpass-firstpass+2,1,imx);
1.218 brouard 11126:
11127:
1.127 brouard 11128: /*---------- Health expectancies, no variances ------------*/
1.218 brouard 11129:
1.201 brouard 11130: strcpy(filerese,"E_");
11131: strcat(filerese,fileresu);
1.126 brouard 11132: if((ficreseij=fopen(filerese,"w"))==NULL) {
11133: printf("Problem with Health Exp. resultfile: %s\n", filerese); exit(0);
11134: fprintf(ficlog,"Problem with Health Exp. resultfile: %s\n", filerese); exit(0);
11135: }
1.208 brouard 11136: printf("Computing Health Expectancies: result on file '%s' ...", filerese);fflush(stdout);
11137: fprintf(ficlog,"Computing Health Expectancies: result on file '%s' ...", filerese);fflush(ficlog);
1.238 brouard 11138:
11139: pstamp(ficreseij);
1.219 brouard 11140:
1.235 brouard 11141: i1=pow(2,cptcoveff); /* Number of combination of dummy covariates */
11142: if (cptcovn < 1){i1=1;}
11143:
11144: for(nres=1; nres <= nresult; nres++) /* For each resultline */
11145: for(k=1; k<=i1;k++){ /* For any combination of dummy covariates, fixed and varying */
11146: if(TKresult[nres]!= k)
11147: continue;
1.219 brouard 11148: fprintf(ficreseij,"\n#****** ");
1.235 brouard 11149: printf("\n#****** ");
1.225 brouard 11150: for(j=1;j<=cptcoveff;j++) {
1.227 brouard 11151: fprintf(ficreseij,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
1.235 brouard 11152: printf("V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
11153: }
11154: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
11155: printf(" V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11156: fprintf(ficreseij," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
1.219 brouard 11157: }
11158: fprintf(ficreseij,"******\n");
1.235 brouard 11159: printf("******\n");
1.219 brouard 11160:
11161: eij=ma3x(1,nlstate,1,nlstate,(int) bage, (int) fage);
11162: oldm=oldms;savm=savms;
1.235 brouard 11163: evsij(eij, p, nlstate, stepm, (int) bage, (int)fage, oldm, savm, k, estepm, strstart, nres);
1.127 brouard 11164:
1.219 brouard 11165: free_ma3x(eij,1,nlstate,1,nlstate,(int) bage, (int)fage);
1.127 brouard 11166: }
11167: fclose(ficreseij);
1.208 brouard 11168: printf("done evsij\n");fflush(stdout);
11169: fprintf(ficlog,"done evsij\n");fflush(ficlog);
1.218 brouard 11170:
1.227 brouard 11171: /*---------- State-specific expectancies and variances ------------*/
1.218 brouard 11172:
11173:
1.201 brouard 11174: strcpy(filerest,"T_");
11175: strcat(filerest,fileresu);
1.127 brouard 11176: if((ficrest=fopen(filerest,"w"))==NULL) {
11177: printf("Problem with total LE resultfile: %s\n", filerest);goto end;
11178: fprintf(ficlog,"Problem with total LE resultfile: %s\n", filerest);goto end;
11179: }
1.208 brouard 11180: printf("Computing Total Life expectancies with their standard errors: file '%s' ...\n", filerest); fflush(stdout);
11181: fprintf(ficlog,"Computing Total Life expectancies with their standard errors: file '%s' ...\n", filerest); fflush(ficlog);
1.218 brouard 11182:
1.126 brouard 11183:
1.201 brouard 11184: strcpy(fileresstde,"STDE_");
11185: strcat(fileresstde,fileresu);
1.126 brouard 11186: if((ficresstdeij=fopen(fileresstde,"w"))==NULL) {
1.227 brouard 11187: printf("Problem with State specific Exp. and std errors resultfile: %s\n", fileresstde); exit(0);
11188: fprintf(ficlog,"Problem with State specific Exp. and std errors resultfile: %s\n", fileresstde); exit(0);
1.126 brouard 11189: }
1.227 brouard 11190: printf(" Computing State-specific Expectancies and standard errors: result on file '%s' \n", fileresstde);
11191: fprintf(ficlog," Computing State-specific Expectancies and standard errors: result on file '%s' \n", fileresstde);
1.126 brouard 11192:
1.201 brouard 11193: strcpy(filerescve,"CVE_");
11194: strcat(filerescve,fileresu);
1.126 brouard 11195: if((ficrescveij=fopen(filerescve,"w"))==NULL) {
1.227 brouard 11196: printf("Problem with Covar. State-specific Exp. resultfile: %s\n", filerescve); exit(0);
11197: fprintf(ficlog,"Problem with Covar. State-specific Exp. resultfile: %s\n", filerescve); exit(0);
1.126 brouard 11198: }
1.227 brouard 11199: printf(" Computing Covar. of State-specific Expectancies: result on file '%s' \n", filerescve);
11200: fprintf(ficlog," Computing Covar. of State-specific Expectancies: result on file '%s' \n", filerescve);
1.126 brouard 11201:
1.201 brouard 11202: strcpy(fileresv,"V_");
11203: strcat(fileresv,fileresu);
1.126 brouard 11204: if((ficresvij=fopen(fileresv,"w"))==NULL) {
11205: printf("Problem with variance resultfile: %s\n", fileresv);exit(0);
11206: fprintf(ficlog,"Problem with variance resultfile: %s\n", fileresv);exit(0);
11207: }
1.227 brouard 11208: printf(" Computing Variance-covariance of State-specific Expectancies: file '%s' ... ", fileresv);fflush(stdout);
11209: fprintf(ficlog," Computing Variance-covariance of State-specific Expectancies: file '%s' ... ", fileresv);fflush(ficlog);
1.126 brouard 11210:
1.145 brouard 11211: /*for(cptcov=1,k=0;cptcov<=i1;cptcov++){
11212: for(cptcod=1;cptcod<=ncodemax[cptcov];cptcod++){*/
11213:
1.235 brouard 11214: i1=pow(2,cptcoveff); /* Number of combination of dummy covariates */
11215: if (cptcovn < 1){i1=1;}
11216:
11217: for(nres=1; nres <= nresult; nres++) /* For each resultline */
11218: for(k=1; k<=i1;k++){ /* For any combination of dummy covariates, fixed and varying */
11219: if(TKresult[nres]!= k)
11220: continue;
1.242 brouard 11221: printf("\n#****** Result for:");
11222: fprintf(ficrest,"\n#****** Result for:");
11223: fprintf(ficlog,"\n#****** Result for:");
1.227 brouard 11224: for(j=1;j<=cptcoveff;j++){
11225: printf("V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
11226: fprintf(ficrest,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
11227: fprintf(ficlog,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
11228: }
1.235 brouard 11229: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
11230: printf(" V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11231: fprintf(ficrest," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11232: fprintf(ficlog," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11233: }
1.208 brouard 11234: fprintf(ficrest,"******\n");
1.227 brouard 11235: fprintf(ficlog,"******\n");
11236: printf("******\n");
1.208 brouard 11237:
11238: fprintf(ficresstdeij,"\n#****** ");
11239: fprintf(ficrescveij,"\n#****** ");
1.225 brouard 11240: for(j=1;j<=cptcoveff;j++) {
1.227 brouard 11241: fprintf(ficresstdeij,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
11242: fprintf(ficrescveij,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
1.208 brouard 11243: }
1.235 brouard 11244: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
11245: fprintf(ficresstdeij," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11246: fprintf(ficrescveij," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11247: }
1.208 brouard 11248: fprintf(ficresstdeij,"******\n");
11249: fprintf(ficrescveij,"******\n");
11250:
11251: fprintf(ficresvij,"\n#****** ");
1.238 brouard 11252: /* pstamp(ficresvij); */
1.225 brouard 11253: for(j=1;j<=cptcoveff;j++)
1.227 brouard 11254: fprintf(ficresvij,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
1.235 brouard 11255: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
11256: fprintf(ficresvij," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11257: }
1.208 brouard 11258: fprintf(ficresvij,"******\n");
11259:
11260: eij=ma3x(1,nlstate,1,nlstate,(int) bage, (int) fage);
11261: oldm=oldms;savm=savms;
1.235 brouard 11262: printf(" cvevsij ");
11263: fprintf(ficlog, " cvevsij ");
11264: cvevsij(eij, p, nlstate, stepm, (int) bage, (int)fage, oldm, savm, k, estepm, delti, matcov, strstart, nres);
1.208 brouard 11265: printf(" end cvevsij \n ");
11266: fprintf(ficlog, " end cvevsij \n ");
11267:
11268: /*
11269: */
11270: /* goto endfree; */
11271:
11272: vareij=ma3x(1,nlstate,1,nlstate,(int) bage, (int) fage);
11273: pstamp(ficrest);
11274:
11275:
11276: for(vpopbased=0; vpopbased <= popbased; vpopbased++){ /* Done for vpopbased=0 and vpopbased=1 if popbased==1*/
1.227 brouard 11277: oldm=oldms;savm=savms; /* ZZ Segmentation fault */
11278: cptcod= 0; /* To be deleted */
11279: printf("varevsij vpopbased=%d \n",vpopbased);
11280: fprintf(ficlog, "varevsij vpopbased=%d \n",vpopbased);
1.235 brouard 11281: 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 11282: 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 ");
11283: if(vpopbased==1)
11284: 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);
11285: else
11286: fprintf(ficrest,"the age specific period (stable) prevalences in each health state \n");
11287: fprintf(ficrest,"# Age popbased mobilav e.. (std) ");
11288: for (i=1;i<=nlstate;i++) fprintf(ficrest,"e.%d (std) ",i);
11289: fprintf(ficrest,"\n");
11290: /* printf("Which p?\n"); for(i=1;i<=npar;i++)printf("p[i=%d]=%lf,",i,p[i]);printf("\n"); */
11291: epj=vector(1,nlstate+1);
11292: printf("Computing age specific period (stable) prevalences in each health state \n");
11293: fprintf(ficlog,"Computing age specific period (stable) prevalences in each health state \n");
11294: for(age=bage; age <=fage ;age++){
1.235 brouard 11295: prevalim(prlim, nlstate, p, age, oldm, savm, ftolpl, &ncvyear, k, nres); /*ZZ Is it the correct prevalim */
1.227 brouard 11296: if (vpopbased==1) {
11297: if(mobilav ==0){
11298: for(i=1; i<=nlstate;i++)
11299: prlim[i][i]=probs[(int)age][i][k];
11300: }else{ /* mobilav */
11301: for(i=1; i<=nlstate;i++)
11302: prlim[i][i]=mobaverage[(int)age][i][k];
11303: }
11304: }
1.219 brouard 11305:
1.227 brouard 11306: fprintf(ficrest," %4.0f %d %d",age, vpopbased, mobilav);
11307: /* fprintf(ficrest," %4.0f %d %d %d %d",age, vpopbased, mobilav,Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]); */ /* to be done */
11308: /* printf(" age %4.0f ",age); */
11309: for(j=1, epj[nlstate+1]=0.;j <=nlstate;j++){
11310: for(i=1, epj[j]=0.;i <=nlstate;i++) {
11311: epj[j] += prlim[i][i]*eij[i][j][(int)age];
11312: /*ZZZ printf("%lf %lf ", prlim[i][i] ,eij[i][j][(int)age]);*/
11313: /* printf("%lf %lf ", prlim[i][i] ,eij[i][j][(int)age]); */
11314: }
11315: epj[nlstate+1] +=epj[j];
11316: }
11317: /* printf(" age %4.0f \n",age); */
1.219 brouard 11318:
1.227 brouard 11319: for(i=1, vepp=0.;i <=nlstate;i++)
11320: for(j=1;j <=nlstate;j++)
11321: vepp += vareij[i][j][(int)age];
11322: fprintf(ficrest," %7.3f (%7.3f)", epj[nlstate+1],sqrt(vepp));
11323: for(j=1;j <=nlstate;j++){
11324: fprintf(ficrest," %7.3f (%7.3f)", epj[j],sqrt(vareij[j][j][(int)age]));
11325: }
11326: fprintf(ficrest,"\n");
11327: }
1.208 brouard 11328: } /* End vpopbased */
11329: free_ma3x(eij,1,nlstate,1,nlstate,(int) bage, (int)fage);
11330: free_ma3x(vareij,1,nlstate,1,nlstate,(int) bage, (int)fage);
11331: free_vector(epj,1,nlstate+1);
1.235 brouard 11332: printf("done selection\n");fflush(stdout);
11333: fprintf(ficlog,"done selection\n");fflush(ficlog);
1.208 brouard 11334:
1.145 brouard 11335: /*}*/
1.235 brouard 11336: } /* End k selection */
1.227 brouard 11337:
11338: printf("done State-specific expectancies\n");fflush(stdout);
11339: fprintf(ficlog,"done State-specific expectancies\n");fflush(ficlog);
11340:
1.126 brouard 11341: /*------- Variance of period (stable) prevalence------*/
1.227 brouard 11342:
1.201 brouard 11343: strcpy(fileresvpl,"VPL_");
11344: strcat(fileresvpl,fileresu);
1.126 brouard 11345: if((ficresvpl=fopen(fileresvpl,"w"))==NULL) {
11346: printf("Problem with variance of period (stable) prevalence resultfile: %s\n", fileresvpl);
11347: exit(0);
11348: }
1.208 brouard 11349: printf("Computing Variance-covariance of period (stable) prevalence: file '%s' ...", fileresvpl);fflush(stdout);
11350: fprintf(ficlog, "Computing Variance-covariance of period (stable) prevalence: file '%s' ...", fileresvpl);fflush(ficlog);
1.227 brouard 11351:
1.145 brouard 11352: /*for(cptcov=1,k=0;cptcov<=i1;cptcov++){
11353: for(cptcod=1;cptcod<=ncodemax[cptcov];cptcod++){*/
1.227 brouard 11354:
1.235 brouard 11355: i1=pow(2,cptcoveff);
11356: if (cptcovn < 1){i1=1;}
11357:
11358: for(nres=1; nres <= nresult; nres++) /* For each resultline */
11359: for(k=1; k<=i1;k++){
11360: if(TKresult[nres]!= k)
11361: continue;
1.227 brouard 11362: fprintf(ficresvpl,"\n#****** ");
11363: printf("\n#****** ");
11364: fprintf(ficlog,"\n#****** ");
11365: for(j=1;j<=cptcoveff;j++) {
11366: fprintf(ficresvpl,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
11367: fprintf(ficlog,"V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
11368: printf("V%d=%d ",Tvaraff[j],nbcode[Tvaraff[j]][codtabm(k,j)]);
11369: }
1.235 brouard 11370: for (j=1; j<= nsq; j++){ /* For each selected (single) quantitative value */
11371: printf(" V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11372: fprintf(ficresvpl," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11373: fprintf(ficlog," V%d=%f ",Tvqresult[nres][j],Tqresult[nres][j]);
11374: }
1.227 brouard 11375: fprintf(ficresvpl,"******\n");
11376: printf("******\n");
11377: fprintf(ficlog,"******\n");
11378:
11379: varpl=matrix(1,nlstate,(int) bage, (int) fage);
11380: oldm=oldms;savm=savms;
1.235 brouard 11381: varprevlim(fileres, varpl, matcov, p, delti, nlstate, stepm, (int) bage, (int) fage, oldm, savm, prlim, ftolpl, &ncvyear, k, strstart, nres);
1.227 brouard 11382: free_matrix(varpl,1,nlstate,(int) bage, (int)fage);
1.145 brouard 11383: /*}*/
1.126 brouard 11384: }
1.227 brouard 11385:
1.126 brouard 11386: fclose(ficresvpl);
1.208 brouard 11387: printf("done variance-covariance of period prevalence\n");fflush(stdout);
11388: fprintf(ficlog,"done variance-covariance of period prevalence\n");fflush(ficlog);
1.227 brouard 11389:
11390: free_vector(weight,1,n);
11391: free_imatrix(Tvard,1,NCOVMAX,1,2);
11392: free_imatrix(s,1,maxwav+1,1,n);
11393: free_matrix(anint,1,maxwav,1,n);
11394: free_matrix(mint,1,maxwav,1,n);
11395: free_ivector(cod,1,n);
11396: free_ivector(tab,1,NCOVMAX);
11397: fclose(ficresstdeij);
11398: fclose(ficrescveij);
11399: fclose(ficresvij);
11400: fclose(ficrest);
11401: fclose(ficpar);
11402:
11403:
1.126 brouard 11404: /*---------- End : free ----------------*/
1.219 brouard 11405: if (mobilav!=0 ||mobilavproj !=0)
11406: 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 11407: free_ma3x(probs,1,AGESUP,1,nlstate+ndeath, 1,ncovcombmax);
1.220 brouard 11408: free_matrix(prlim,1,nlstate,1,nlstate); /*here or after loop ? */
11409: free_matrix(pmmij,1,nlstate+ndeath,1,nlstate+ndeath);
1.126 brouard 11410: } /* mle==-3 arrives here for freeing */
1.227 brouard 11411: /* endfree:*/
11412: free_matrix(oldms, 1,nlstate+ndeath,1,nlstate+ndeath);
11413: free_matrix(newms, 1,nlstate+ndeath,1,nlstate+ndeath);
11414: free_matrix(savms, 1,nlstate+ndeath,1,nlstate+ndeath);
11415: free_ma3x(cotqvar,1,maxwav,1,nqtv,1,n);
1.233 brouard 11416: free_ma3x(cotvar,1,maxwav,1,ntv+nqtv,1,n);
1.227 brouard 11417: free_matrix(coqvar,1,maxwav,1,n);
11418: free_matrix(covar,0,NCOVMAX,1,n);
11419: free_matrix(matcov,1,npar,1,npar);
11420: free_matrix(hess,1,npar,1,npar);
11421: /*free_vector(delti,1,npar);*/
11422: free_ma3x(delti3,1,nlstate,1, nlstate+ndeath-1,1,ncovmodel);
11423: free_matrix(agev,1,maxwav,1,imx);
11424: free_ma3x(param,1,nlstate,1, nlstate+ndeath-1,1,ncovmodel);
11425:
11426: free_ivector(ncodemax,1,NCOVMAX);
11427: free_ivector(ncodemaxwundef,1,NCOVMAX);
11428: free_ivector(Dummy,-1,NCOVMAX);
11429: free_ivector(Fixed,-1,NCOVMAX);
1.238 brouard 11430: free_ivector(DummyV,1,NCOVMAX);
11431: free_ivector(FixedV,1,NCOVMAX);
1.227 brouard 11432: free_ivector(Typevar,-1,NCOVMAX);
11433: free_ivector(Tvar,1,NCOVMAX);
1.234 brouard 11434: free_ivector(TvarsQ,1,NCOVMAX);
11435: free_ivector(TvarsQind,1,NCOVMAX);
11436: free_ivector(TvarsD,1,NCOVMAX);
11437: free_ivector(TvarsDind,1,NCOVMAX);
1.231 brouard 11438: free_ivector(TvarFD,1,NCOVMAX);
11439: free_ivector(TvarFDind,1,NCOVMAX);
1.232 brouard 11440: free_ivector(TvarF,1,NCOVMAX);
11441: free_ivector(TvarFind,1,NCOVMAX);
11442: free_ivector(TvarV,1,NCOVMAX);
11443: free_ivector(TvarVind,1,NCOVMAX);
11444: free_ivector(TvarA,1,NCOVMAX);
11445: free_ivector(TvarAind,1,NCOVMAX);
1.231 brouard 11446: free_ivector(TvarFQ,1,NCOVMAX);
11447: free_ivector(TvarFQind,1,NCOVMAX);
11448: free_ivector(TvarVD,1,NCOVMAX);
11449: free_ivector(TvarVDind,1,NCOVMAX);
11450: free_ivector(TvarVQ,1,NCOVMAX);
11451: free_ivector(TvarVQind,1,NCOVMAX);
1.230 brouard 11452: free_ivector(Tvarsel,1,NCOVMAX);
11453: free_vector(Tvalsel,1,NCOVMAX);
1.227 brouard 11454: free_ivector(Tposprod,1,NCOVMAX);
11455: free_ivector(Tprod,1,NCOVMAX);
11456: free_ivector(Tvaraff,1,NCOVMAX);
11457: free_ivector(invalidvarcomb,1,ncovcombmax);
11458: free_ivector(Tage,1,NCOVMAX);
11459: free_ivector(Tmodelind,1,NCOVMAX);
1.228 brouard 11460: free_ivector(TmodelInvind,1,NCOVMAX);
11461: free_ivector(TmodelInvQind,1,NCOVMAX);
1.227 brouard 11462:
11463: free_imatrix(nbcode,0,NCOVMAX,0,NCOVMAX);
11464: /* free_imatrix(codtab,1,100,1,10); */
1.126 brouard 11465: fflush(fichtm);
11466: fflush(ficgp);
11467:
1.227 brouard 11468:
1.126 brouard 11469: if((nberr >0) || (nbwarn>0)){
1.216 brouard 11470: printf("End of Imach with %d errors and/or %d warnings. Please look at the log file for details.\n",nberr,nbwarn);
11471: 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 11472: }else{
11473: printf("End of Imach\n");
11474: fprintf(ficlog,"End of Imach\n");
11475: }
11476: printf("See log file on %s\n",filelog);
11477: /* gettimeofday(&end_time, (struct timezone*)0);*/ /* after time */
1.157 brouard 11478: /*(void) gettimeofday(&end_time,&tzp);*/
11479: rend_time = time(NULL);
11480: end_time = *localtime(&rend_time);
11481: /* tml = *localtime(&end_time.tm_sec); */
11482: strcpy(strtend,asctime(&end_time));
1.126 brouard 11483: printf("Local time at start %s\nLocal time at end %s",strstart, strtend);
11484: fprintf(ficlog,"Local time at start %s\nLocal time at end %s\n",strstart, strtend);
1.157 brouard 11485: printf("Total time used %s\n", asc_diff_time(rend_time -rstart_time,tmpout));
1.227 brouard 11486:
1.157 brouard 11487: printf("Total time was %.0lf Sec.\n", difftime(rend_time,rstart_time));
11488: fprintf(ficlog,"Total time used %s\n", asc_diff_time(rend_time -rstart_time,tmpout));
11489: fprintf(ficlog,"Total time was %.0lf Sec.\n", difftime(rend_time,rstart_time));
1.126 brouard 11490: /* printf("Total time was %d uSec.\n", total_usecs);*/
11491: /* if(fileappend(fichtm,optionfilehtm)){ */
11492: fprintf(fichtm,"<br>Local time at start %s<br>Local time at end %s<br>\n</body></html>",strstart, strtend);
11493: fclose(fichtm);
11494: fprintf(fichtmcov,"<br>Local time at start %s<br>Local time at end %s<br>\n</body></html>",strstart, strtend);
11495: fclose(fichtmcov);
11496: fclose(ficgp);
11497: fclose(ficlog);
11498: /*------ End -----------*/
1.227 brouard 11499:
11500:
11501: printf("Before Current directory %s!\n",pathcd);
1.184 brouard 11502: #ifdef WIN32
1.227 brouard 11503: if (_chdir(pathcd) != 0)
11504: printf("Can't move to directory %s!\n",path);
11505: if(_getcwd(pathcd,MAXLINE) > 0)
1.184 brouard 11506: #else
1.227 brouard 11507: if(chdir(pathcd) != 0)
11508: printf("Can't move to directory %s!\n", path);
11509: if (getcwd(pathcd, MAXLINE) > 0)
1.184 brouard 11510: #endif
1.126 brouard 11511: printf("Current directory %s!\n",pathcd);
11512: /*strcat(plotcmd,CHARSEPARATOR);*/
11513: sprintf(plotcmd,"gnuplot");
1.157 brouard 11514: #ifdef _WIN32
1.126 brouard 11515: sprintf(plotcmd,"\"%sgnuplot.exe\"",pathimach);
11516: #endif
11517: if(!stat(plotcmd,&info)){
1.158 brouard 11518: printf("Error or gnuplot program not found: '%s'\n",plotcmd);fflush(stdout);
1.126 brouard 11519: if(!stat(getenv("GNUPLOTBIN"),&info)){
1.158 brouard 11520: printf("Error or gnuplot program not found: '%s' Environment GNUPLOTBIN not set.\n",plotcmd);fflush(stdout);
1.126 brouard 11521: }else
11522: strcpy(pplotcmd,plotcmd);
1.157 brouard 11523: #ifdef __unix
1.126 brouard 11524: strcpy(plotcmd,GNUPLOTPROGRAM);
11525: if(!stat(plotcmd,&info)){
1.158 brouard 11526: printf("Error gnuplot program not found: '%s'\n",plotcmd);fflush(stdout);
1.126 brouard 11527: }else
11528: strcpy(pplotcmd,plotcmd);
11529: #endif
11530: }else
11531: strcpy(pplotcmd,plotcmd);
11532:
11533: sprintf(plotcmd,"%s %s",pplotcmd, optionfilegnuplot);
1.158 brouard 11534: printf("Starting graphs with: '%s'\n",plotcmd);fflush(stdout);
1.227 brouard 11535:
1.126 brouard 11536: if((outcmd=system(plotcmd)) != 0){
1.158 brouard 11537: printf("gnuplot command might not be in your path: '%s', err=%d\n", plotcmd, outcmd);
1.154 brouard 11538: printf("\n Trying if gnuplot resides on the same directory that IMaCh\n");
1.152 brouard 11539: sprintf(plotcmd,"%sgnuplot %s", pathimach, optionfilegnuplot);
1.150 brouard 11540: if((outcmd=system(plotcmd)) != 0)
1.153 brouard 11541: printf("\n Still a problem with gnuplot command %s, err=%d\n", plotcmd, outcmd);
1.126 brouard 11542: }
1.158 brouard 11543: printf(" Successful, please wait...");
1.126 brouard 11544: while (z[0] != 'q') {
11545: /* chdir(path); */
1.154 brouard 11546: printf("\nType e to edit results with your browser, g to graph again and q for exit: ");
1.126 brouard 11547: scanf("%s",z);
11548: /* if (z[0] == 'c') system("./imach"); */
11549: if (z[0] == 'e') {
1.158 brouard 11550: #ifdef __APPLE__
1.152 brouard 11551: sprintf(pplotcmd, "open %s", optionfilehtm);
1.157 brouard 11552: #elif __linux
11553: sprintf(pplotcmd, "xdg-open %s", optionfilehtm);
1.153 brouard 11554: #else
1.152 brouard 11555: sprintf(pplotcmd, "%s", optionfilehtm);
1.153 brouard 11556: #endif
11557: printf("Starting browser with: %s",pplotcmd);fflush(stdout);
11558: system(pplotcmd);
1.126 brouard 11559: }
11560: else if (z[0] == 'g') system(plotcmd);
11561: else if (z[0] == 'q') exit(0);
11562: }
1.227 brouard 11563: end:
1.126 brouard 11564: while (z[0] != 'q') {
1.195 brouard 11565: printf("\nType q for exiting: "); fflush(stdout);
1.126 brouard 11566: scanf("%s",z);
11567: }
11568: }
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