imach/src/ChangeLog - view

File: [Local Repository] / imach / src / ChangeLog
Revision 1.68: download - view: text, annotated - select for diffs
Wed Apr 24 20:53:06 2024 UTC (6 months, 1 week ago) by brouard
Branches: MAIN
CVS tags: HEAD

Summary: First IMaCh version using praxis method and adapted C code from Gegenfürtner

1: 2024-04-24 Nicolas Brouard <brouard@ined.fr> 2: 3: * (Module): This version comes late after having tested 4: successfully the praxis C version of Buckhardt. But Buckardt's 5: version was difficult to read and Gegenfurtner's version had a few 6: typos which made its results less reliable than Buckhardt's 7: results. The most important work consisted in retyping the Brent 8: original PRAXIS program written in Algol W (published with errors, 9: ommitting the transposition of matrix V before its QR reduction 10: from Golub. I used the recent "awe" compiler from Gkynn Webster. 11: The awe library had errors, for example in arc tangent function 12: which have been fixed. 13: 14: The main objective was to get identical results with the three 15: versions: (1) Algol W, (2) Buckhardt'C version as well (3) 16: Gegenfürtner C versions on the various test functions published by 17: Brent in 1973 in Algol W. 18: 19: Also, in order to compare them, the random function had to produce 20: the same sequence for the 3 softwares. The random function used in 21: imach corresponds to original Brent's random function written in 22: Algol W. Other point, in Algol W, the arrays of dimension n are 23: 'normal' mathematical arrays starting from 1 to n. But this is a 24: real issue in C where, by default, arrays are starting from 0 to 25: n-1. In Buckhardt, as well as in Gegenfürtner C code, it can be 26: seen that authors while trying to mimick original Brent Algol W 27: code are hesitating by changing either a loop originally from 1 to 28: n in a loop from 0 to n-1, or keeping Brent's loop from 1 to n and 29: shifting the index from original X(I) in Algol W to x[i-1] in C. 30: But as IMaCh is using, since the beginning, the Numerical Recipes 31: functions vector or matrix, I changed Geggenfürtner code to mimick 32: the original Algol W arrays. Thus the X(I) is translated in C as 33: x[i] which minimizes the errors. The Golub QR algorithm was 34: published in Algol with overflow errors which were reproduced in 35: Brent's Algol W code. Buckhardt code fixed these errors which are 36: much more problematic in C than in Algol W. Thus Buckhardt code 37: seems very safe, but i haven't chosen it for IMaCh because the C 38: style is horrible and almost unreadable compared to Gegenfürtner 39: CO code which is very close to Brent's original. Also what makes 40: Buckhardt code more difficult to read is, instead of passing the 41: minimum of parameters in the functions calls, as it is in Algol 42: Brent's code or Gegenfürtner's code, the list of parameters is 43: high. For example, the flin function LONG REAL PROCEDURE FLIN 44: (LONG REAL VALUE L) has only one parameter in Algol W, the 45: Gegenfürtner flin function had two parameters: static double 46: flin(l, j) double l; { int i; double tflin[N];} but Buckhardt 47: function has 14 parameters which makes the code unreadable and 48: useless. Gegenfürtner used a lot of static variables or functions 49: which I tried to minimize. Also in Gegefürtner, array dimensions 50: were fixed to N. In my adaptation the flin is static double 51: flin(double l, int j) and the parameter used are global variables. 52: 53: 2023-06-14 Nicolas Brouard <brouard@ined.fr> 54: 55: * imach.c (Module): Testing if conjugate gradient could be quicker 56: when lot of variables POWELLORIGINCONJUGATE 57: 58: 2023-05-23 Nicolas Brouard <brouard@ined.fr> 59: 60: * imach.c (Module): Fixed PROB_r 61: 62: 2023-05-22 Nicolas Brouard <brouard@ined.fr> 63: 64: * imach.c (Module): In the ILK....txt file, the number of columns 65: before the covariates values is dependent of the number of states (16+nlstate): 0.99r46 66: 67: 2023-05-08 Nicolas Brouard <brouard@ined.fr> 68: 69: * (Module): Error V0 when result:. and model 1+age+V1: fixed 70: 71: 2023-04-29 Nicolas Brouard <brouard@ined.fr> 72: 73: * imach.c (Module): Inverting the model equation and fixingg bugs 74: in the drawings 75: "1+age+V7*V4*age+V6*V4*age+V7*V3*age+V6*V3*age+V6*V2*age+V7*age+V6*age+V4*age+V3*age+V2*age+V7*V4+V6*V4+V7*V3+V6*V3+V7*V2+V6*V2+V7+V6+V4+V3+V2\000\0003+V2" 76: 77: 2023-04-24 Nicolas Brouard <brouard@ined.fr> 78: 79: * imach.c (Module): some bugs in printinggnuplot 80: 81: 2022-09-19 Nicolas Brouard <brouard@ined.fr> 82: 83: * imach.c (Module): Version 0.99r42 needed a newer version of 84: Gnuplot. But newer version 0.99r43 should run with the Gnuplot 85: version 5.0 or 5.1 distributed with IMaCh. 86: 87: 2022-09-18 Nicolas Brouard <brouard@ined.fr> 88: 89: * imach.c (Module): Fixing when a state in the data is 0 or higher 90: than lstate+ndeath. Fixing the plot of contribution to Likelihood. 91: 92: 2022-09-16 Nicolas Brouard <brouard@ined.fr> 93: 94: * imach.c (Module): 0.99r41 Was an error when product of timevarying and fixed. Using FixedV[of name] now. Thank you Feinuo 95: 96: 2022-09-14 Nicolas Brouard <brouard@ined.fr> 97: 98: * imach.c (Module): Fixing names of variables in T_ (thanks to Feinuo) 99: 100: 2022-09-14 Nicolas Brouard <brouard@ined.fr> 101: 102: * imach.c (Module): Version 0.99r39 with colored dummy covariates 103: (fixed or time varying), using new last columns of 104: ILK_parameter.txt file. 105: 106: 2022-09-11 Nicolas Brouard <brouard@ined.fr> 107: 108: * imach.c (Module): Adding timevarying products of any kinds, 109: should work before shifting cotvar from ncovcol+nqv columns in 110: order to have a correspondance between the column of cotvar and 111: the id of column. 112: (Module): Some cleaning and adding covariates in ILK.txt 113: 114: 2022-09-09 Nicolas Brouard <brouard@ined.fr> 115: 116: * imach.c (Module): Many improvements for fixing products of fixed 117: timevarying as well as fixed * fixed, and test with quantitative 118: covariate. 119: 120: 2022-09-04 Nicolas Brouard <brouard@ined.fr> 121: 122: * imach.c (Module): Now the easy runs i.e. without result or 123: model=1+age only did not work. The defautl combination should be 1 124: and not 0 because everything hasn't been tranformed yet. 125: 126: 2022-09-02 Nicolas Brouard <brouard@ined.fr> 127: 128: * imach.c: Version 0.99r35 because it outputs same results with 129: 1+age+V1+V1*age for females and 1+age for females only 130: (education=1 noweight) 131: 132: 2022-08-31 Nicolas Brouard <brouard@ined.fr> 133: 134: * imach.c (Module): Some improvments in fichtm and many verifications 0.99r34 135: 136: 2022-08-21 Nicolas Brouard <brouard@ined.fr> 137: 138: * imach.c (Module): Version 0.99r33 A lot of changes in 139: reassigning covariates: my first idea was that people will always 140: use the first covariate V1 into the model but in fact they are 141: producing data with many covariates and can use an equation model 142: with some of the covariate; it means that in a model V2+V3 instead 143: of codtabm(k,Tvaraff[j]) which calculates for combination k, for 144: three covariates (V1, V2, V3) the value of Tvaraff[j], but in fact 145: the equation model is restricted to two variables only (V2, V3) 146: and the combination for V2 should be codtabm(k,1) instead of 147: (codtabm(k,2), and the code should be 148: codtabm(k,TnsdVar[Tvaraff[j]]. Many many changes have been 149: made. All of these should be simplified once a day like we did in 150: hpxij() for example by using precov[nres] which is computed in 151: decoderesult for each nres of each resultline. Loop should be done 152: on the equation model globally by distinguishing only product with 153: age (which are changing with age) and no more on type of 154: covariates, single dummies, single covariates. 155: 156: 2022-08-06 Nicolas Brouard <brouard@ined.fr> 157: 158: * imach.c (Module): Version of imach using partly decoderesult to rebuild xpxij function 159: 160: 2022-08-03 Nicolas Brouard <brouard@ined.fr> 161: 162: * imach.c (Module): Many errors in graphs fixed with Vn*age covariates. 163: 164: 2022-07-25 Brouard Nicolas <brouard@brouard.name> 165: 166: * imach.c (Module): Error cptcovn instead of nsd in bmij (was 167: coredumped, revealed by Feiuno, thank you. 168: 169: 2022-07-23 Nicolas Brouard <brouard@ined.fr> 170: 171: * r29 W and not sqrt(Wald) 172: 173: 2022-07-22 Nicolas Brouard <brouard@ined.fr> 174: 175: * imach.c (Module): Output of Wald test in the htm file and not only in the log. 176: 177: 2022-06-02 Brouard Nicolas <brouard@brouard.name> 178: 179: * imach.c (Module): Adding the Wald tests from the log to the main 180: htm for better display of the maximum likelihood estimators. 181: 182: 2022-05-30 Brouard Nicolas <brouard@brouard.name> 183: 184: * imach.c: With products of covariates (age or dummies or 185: quantitatives), estimates of parameters were good but when 186: estimating variances the positions of newly created covariates 187: were wrongly assigned and results were wrong. Thank to Feinua_Sun! 188: 189: 2022-05-24 Nicolas Brouard <brouard@ined.fr> 190: 191: * imach.c (Module): Some attempts to find a bug of wrong estimates 192: of confidencce intervals with product in the equation modelC 193: 194: 2022-05-15 Nicolas Brouard <brouard@ined.fr> 195: 196: * imach.c (Module): Some minor improvements 197: 198: 2022-04-13 Brouard Nicolas <brouard@brouard.name> 199: 200: * imach.c (Module): Adding link to text data files 201: 202: 2022-04-11 Nicolas Brouard <brouard@ined.fr> 203: 204: * imach.c (Module): Error in rewriting the 'r' file with yearsfproj or yearsbproj fixed 205: 206: 2022-04-05 Brouard Nicolas <brouard@ined.fr> 207: 208: * imach.c (Module): Fixed covariates (dummy or quantitative) 209: with missing values have never been allowed but are ERRORS and 210: program quits. Standard deviations of fixed covariates were 211: wrongly computed. Mean and standard deviations of time varying 212: covariates are still not computed. 213: 214: 2022-03-17 Brouard Nicolas <brouard@ined.fr> 215: 216: * 99r25 Improvements in fixing discrepancies between covariates 217: missing in result line but in model. 218: 219: *Wed Mar 31 2021 Nicolas Brouard <brouard@ined.fr> 220: * imach.c (Module): Still bugs in the result loop. Thank to Holly Benett 221: 222: *2021-03-08 Nicolas Brouard <brouard@ined.fr> 223: 224: * ./ (Module): Fixed bug on result: 225: 226: *2021-02-20 Nicolas Brouard <brouard@ined.fr> 227: 228: * imach.c (Module): Fix bug on quitting after result lines! 229: (Module): Version 0.99r21 230: 231: 2021-02-12 Nicolas Brouard <brouard@ined.fr> 232: 233: * imach.c (Module): The use of a Windows BOM (huge) file is now an error 234: 235: 236: 2021-02-11 Nicolas Brouard <brouard@ined.fr> 237: 238: * (Module): imach.c Someone entered 'results:' instead of 'result:'. Now it is an error which is printed. 239: 240: 2020-02-22 Brouard Nicolas <brouard@ined.fr> 241: 242: * (Module): imach.c Update mle=-3 (for computing Life expectancy 243: and life table from the data without any state) 244: 245: 2019-05-20 Brouard Nicolas <brouard@ined.fr> 246: 247: * version.h: Summary: version 0.99.r19 248: 249: * imach.c: Summary: Projection syntax simplified 250: We can now start projections, forward or backward, from the mean date 251: of inteviews up to or down to a number of years of projection: 252: prevforecast=1 yearsfproj=15.3 mobil_average=0 253: or 254: prevforecast=1 starting-proj-date=1/1/2007 final-proj-date=12/31/2017 mobil_average=0 255: or 256: prevbackcast=1 yearsbproj=12.3 mobil_average=1 257: or 258: prevbackcast=1 starting-back-date=1/10/1999 final-back-date=1/1/1985 mobil_average=1 259: 260: 2019-05-18 Brouard Nicolas <brouard@ined.fr> 261: 262: * imach.c: Summary: doxygen tex bug 263: 264: 2019-05-16 Brouard Nicolas <brouard@ined.fr> 265: 266: * imach.c: Summary: There was some wrong lines added 267: 268: 2019-05-09 Brouard Nicolas <brouard@ined.fr> 269: 270: * Makefile, imach.c: *** empty log message *** 271: 272: * ChangeLog: Summary: C= 273: 274: * imach.c: Summary: Some updates 275: 276: * imach.c: Summary: Before ncovmax 277: 278: * version.h: *** empty log message *** 279: 280: * imach.c: Summary: 0.99r18 unlimited number of individuals 281: The number n which was limited to 20,000 cases is now unlimited, from firstobs to lastobs. If the number is too for the virtual memory, probably an error will occur. 282: 283: 2018-12-13 Brouard Nicolas <brouard@ined.fr> 284: 285: * imach.c: Summary: Bug for young ages (<-30) will be in r17 286: 287: 2018-12-05 Brouard Nicolas <brouard@ined.fr> 288: 289: * Makefile: Summary: 0.99r16 290: 291: 2018-05-02 Brouard Nicolas <brouard@ined.fr> 292: 293: * imach.c: Summary: Some bugs fixed 294: 295: 2018-05-02 Brouard Nicolas <brouard@ined.fr> 296: 297: * imach.c: Summary: Bug for young ages (<-30) will be in r17 298: 299: 2018-05-02 Brouard Nicolas <brouard@ined.fr> 300: 301: * imach.c: Summary: Some bugs fixed 302: 303: 2018-05-01 Brouard Nicolas <brouard@ined.fr> 304: 305: * imach.c: 306: Summary: Bug fixed by providing frequencies only for non missing covariates 307: 308: 2018-04-27 Brouard Nicolas <brouard@ined.fr> 309: 310: * imach.c: Summary: some minor bugs 311: 312: 2018-04-21 Brouard Nicolas <brouard@ined.fr> 313: 314: * README.txt: *** empty log message *** 315: 316: * imach.c: Summary: Some bugs fixed, valgrind tested 317: 318: 2018-04-20 Brouard Nicolas <brouard@ined.fr> 319: 320: * Makefile, README.txt, setup.iss: Summary: imach 0.99r16 321: 322: * imach.c: 323: Summary: Computing mean and stdeviation of fixed quantitative variables 324: 325: 2018-04-19 Brouard Nicolas <brouard@ined.fr> 326: 327: * imach.c: Summary: Some minor bugs fixed 328: 329: * version.h: *** empty log message *** 330: 331: 2018-02-27 Brouard Nicolas <brouard@ined.fr> 332: 333: * imach.c: *** empty log message *** 334: 335: * imach.c: Summary: Adding second argument for quitting 336: 337: 2018-02-21 Brouard Nicolas <brouard@ined.fr> 338: 339: * Makefile, README.txt, imach.c: Summary: 0.99r15 340: New Makefile with recent VirtualBox 5.26. Bug in sqrt negatve in imach.c 341: 342: 2017-07-20 Brouard Nicolas <brouard@ined.fr> 343: 344: * imach.c: Summary: temporary working 345: 346: 2017-07-19 Brouard Nicolas <brouard@ined.fr> 347: 348: * imach.c: Summary: Bug for mobil_average=0 and prevforecast fixed(?) 349: 350: 2017-07-17 Brouard Nicolas <brouard@ined.fr> 351: 352: * imach.c: Summary: BOM files can be read now 353: 354: 2017-06-30 Brouard Nicolas <brouard@ined.fr> 355: 356: * imach.c: Summary: Graphs improvements 357: 358: * imach.c: Summary: Saito's color 359: 360: 2017-06-29 Brouard Nicolas <brouard@ined.fr> 361: 362: * imach.c: Summary: Version 0.99r14 363: 364: 2017-06-27 Brouard Nicolas <brouard@ined.fr> 365: 366: * imach.c: Summary: More documentation on projections 367: 368: * imach.c: Summary: Color of backprojection changed from 6 to 5(yellow) 369: 370: * imach.c: Summary: Some bug with rint 371: 372: 2017-05-24 Brouard Nicolas <brouard@ined.fr> 373: 374: * imach.c: *** empty log message *** 375: 376: 2017-05-23 Brouard Nicolas <brouard@ined.fr> 377: 378: * imach.c: Summary: Code into subroutine, cleanings 379: 380: 2017-05-18 Brouard Nicolas <brouard@ined.fr> 381: 382: * imach.c: 383: Summary: backprojection and confidence intervals of backprevalence 384: 385: 2017-05-13 Brouard Nicolas <brouard@ined.fr> 386: 387: * imach.c: Summary: temporary save for backprojection 388: 389: * imach.c: Summary: Version 0.99r13 (improvements and bugs fixed) 390: 391: 2017-04-26 Brouard Nicolas <brouard@ined.fr> 392: 393: * imach.c: Summary: imach 0.99r13 Some bugs fixed 394: 395: * imach.c: Summary: Labels in graphs 396: 397: 2017-04-24 Brouard Nicolas <brouard@ined.fr> 398: 399: * imach.c: Summary: to save 400: 401: 2017-04-18 Brouard Nicolas <brouard@ined.fr> 402: 403: * imach.c: *** empty log message *** 404: 405: 2017-04-05 Brouard Nicolas <brouard@ined.fr> 406: 407: * imach.c: Summary: Bug in E_ as well as in T_ fixed nres-1 vs k1-1 408: 409: 2017-04-04 Brouard Nicolas <brouard@ined.fr> 410: 411: * imach.c: Summary: Gnuplot indexations fixed (humm) 412: 413: * imach.c: 414: Summary: Some errors to warnings only if date of death is unknown but status is death we could set to pi3 415: 416: 2017-04-03 Brouard Nicolas <brouard@ined.fr> 417: 418: * imach.c: Summary: Version 0.99r12 419: Some cleanings, conformed with updated documentation. 420: 421: 2017-03-30 Brouard Nicolas <brouard@ined.fr> 422: 423: * version.h: *** empty log message *** 424: 425: * version.h: Summary:0.99r11 426: 427: 2017-03-29 Brouard Nicolas <brouard@ined.fr> 428: 429: * imach.c: Summary: Temp 430: 431: 2017-03-27 Brouard Nicolas <brouard@ined.fr> 432: 433: * imach.c: Summary: Temporary 434: 435: 2017-03-08 Brouard Nicolas <brouard@ined.fr> 436: 437: * imach.c: Summary: IMaCh version 0.99r10 bugs in gnuplot fixed 438: 439: * imach.c: Summary: Fixing data parameter line 440: 441: 2016-12-15 Brouard Nicolas <brouard@ined.fr> 442: 443: * imach.c: Summary: 0.99 in progress 444: 445: 2016-09-15 Brouard Nicolas <brouard@ined.fr> 446: 447: * imach.c: *** empty log message *** 448: 449: * imach.c: Summary: not working 450: 451: 2016-09-08 Brouard Nicolas <brouard@ined.fr> 452: 453: * imach.c: Summary: continue 454: 455: 2016-09-07 Brouard Nicolas <brouard@ined.fr> 456: 457: * imach.c: Summary: Starting values from frequencies 458: 459: * imach.c: *** empty log message *** 460: 461: 2016-09-02 Brouard Nicolas <brouard@ined.fr> 462: 463: * imach.c: *** empty log message *** 464: 465: 2016-08-30 Brouard Nicolas <brouard@ined.fr> 466: 467: * imach.c: Summary: Fixing a lots 468: 469: 2016-08-29 Brouard Nicolas <brouard@ined.fr> 470: 471: * imach.c: Summary: gnuplot problem in Back projection to fix 472: 473: * imach.c: Summary: Better 474: 475: 2016-08-26 Brouard Nicolas <brouard@ined.fr> 476: 477: * imach.c: 478: Summary: Improvement in Powell output in order to copy and paste 479: 480: * imach.c: Summary: Starting tests of 0.99 481: 482: * imach.c: Summary: to valgrind 483: 484: 2016-08-25 Brouard Nicolas <brouard@ined.fr> 485: 486: * imach.c: *** empty log message *** 487: 488: 2016-08-23 Brouard Nicolas <brouard@ined.fr> 489: 490: * imach.c: *** empty log message *** 491: 492: * imach.c: Summary: not working 493: 494: 2016-08-22 Brouard Nicolas <brouard@ined.fr> 495: 496: * imach.c: Summary: not working 497: 498: * imach.c: Summary: Not working 499: 500: 2016-07-23 Brouard Nicolas <brouard@ined.fr> 501: 502: * imach.c: Summary: Completing for func too 503: 504: 2016-07-22 Brouard Nicolas <brouard@ined.fr> 505: 506: * imach.c: Summary: Fixing some arrays, still debugging 507: 508: 2016-07-21 Brouard Nicolas <brouard@ined.fr> 509: 510: * imach.c: 511: Summary: 0.99 working (more or less) for Asian Workshop on multitate methods 512: 513: 2016-07-12 Brouard Nicolas <brouard@ined.fr> 514: 515: * imach.c: Summary: temp 516: 517: * imach.c: Summary: saving but not running 518: 519: 2016-07-01 Brouard Nicolas <brouard@ined.fr> 520: 521: * imach.c: Summary: Fixes 522: 523: 2016-02-19 Brouard Nicolas <brouard@ined.fr> 524: 525: * imach-0.98r.c: Summary: Kind of 0.98r? series, starting with r7 526: 527: * setup.iss, Makefile, version.h: *** empty log message *** 528: 529: * imach.c: Summary: temporary 530: 531: 2016-02-17 Brouard Nicolas <brouard@ined.fr> 532: 533: * imach.c: Summary: Probably last 0.98 stable version 0.98r6 534: 535: 2016-02-16 Brouard Nicolas <brouard@ined.fr> 536: 537: * imach.c: Summary: minor bug 538: 539: * imach.c: Summary: 0.99r2 540: 541: 2016-02-15 Brouard Nicolas <brouard@ined.fr> 542: 543: * imach.c: *** empty log message *** 544: 545: 2016-02-12 Brouard Nicolas <brouard@ined.fr> 546: 547: * imach.c: Summary: 0.99 Back projections 548: 549: 2015-12-23 Brouard Nicolas <brouard@ined.fr> 550: 551: * imach.c: Summary: Experimental backcast 552: 553: 2015-12-18 Brouard Nicolas <brouard@ined.fr> 554: 555: * imach.c: Summary: 0.98r4 Warning and status=-2 556: 557: Version 0.98r4 is now: 558: - displaying an error when status is -1, date of interview unknown and date of death known; 559: - permitting a status -2 when the vital status is unknown at a known date of right truncation. 560: Older changes concerning s=-2, dating from 2005 have been supersed. 561: 562: * Makefile: *** empty log message *** 563: 564: 2015-12-16 Brouard Nicolas <brouard@ined.fr> 565: 566: * Makefile: *** empty log message *** 567: 568: * imach.c: Summary: 0.98r4 working 569: 570: * Makefile: Summary: linuxrpm added 571: 572: * Makefile: *** empty log message *** 573: 574: * imach.c: Summary: temporary not working 575: 576: 2015-12-11 Brouard Nicolas <brouard@ined.fr> 577: 578: * imach.c: Summary: 0.98r4 579: 580: 2015-12-08 Brouard Nicolas <brouard@ined.fr> 581: 582: * Makefile: *** empty log message *** 583: 584: 2015-11-21 Brouard Nicolas <brouard@ined.fr> 585: 586: * Makefile: 587: Summary: VBoxManage change from execute --image to run --exe and --wait-exit doesn't exist any more 588: 589: * imach.c: Summary: minor typo 590: 591: * imach.c: Summary: 0.98r3 with some graph of projected cross-sectional 592: 593: Author: Nicolas Brouard 594: 595: 2015-11-18 Brouard Nicolas <brouard@ined.fr> 596: 597: * imach.c: Summary: Start working on projected prevalences 598: 599: 2015-11-17 Brouard Nicolas <brouard@ined.fr> 600: 601: * imach.c: Summary: Adding ftolpl parameter 602: Author: N Brouard 603: 604: We had difficulties to get smoothed confidence intervals. It was due 605: to the period prevalence which wasn't computed accurately. The inner 606: parameter ftolpl is now an outer parameter of the .imach parameter 607: file after estepm. If ftolpl is small 1.e-4 and estepm too, 608: computation are long. 609: 610: * version.h: *** empty log message *** 611: 612: * imach.c: Summary: temporary 613: 614: 2015-10-27 Brouard Nicolas <brouard@ined.fr> 615: 616: * Makefile: *** empty log message *** 617: 618: * setup.iss.in: Summary: some new subdirs 619: 620: * imach.c: *** empty log message *** 621: 622: * setup.iss.in: 623: Summary: Adding gnuplot 5.1 with bin and etc for pango fonts 624: 625: 2015-10-24 Brouard Nicolas <brouard@ined.fr> 626: 627: * imach.c: *** empty log message *** 628: 629: 2015-10-23 Brouard Nicolas <brouard@ined.fr> 630: 631: * imach.c: 632: Summary: 0.98r3 some clarification for graphs on likelihood contributions 633: 634: 2015-10-01 Brouard Nicolas <brouard@ined.fr> 635: 636: * imach.c: Summary: Some new graphs of contribution to likelihood 637: 638: 2015-09-30 Brouard Nicolas <brouard@ined.fr> 639: 640: * Makefile: 641: Summary: Version read from version.h, itself read with cmake 642: 643: * imach.c: Summary: looking at better estimation of the hessian 644: 645: Also a better criteria for convergence to the period prevalence And 646: therefore adding the number of years needed to converge. (The 647: prevalence in any alive state shold sum to one 648: 649: * Makefile: *** empty log message *** 650: 651: 2015-09-22 Brouard Nicolas <brouard@ined.fr> 652: 653: * imach.c: 654: Summary: Adding some overall graph on contribution to likelihood. Might change 655: 656: 2015-09-15 Brouard Nicolas <brouard@ined.fr> 657: 658: * Makefile: *** empty log message *** 659: 660: * imach.c: Summary: 0.98r0 661: 662: - Some new graphs like survival functions 663: - Some bugs fixed like model=1+age+V2. 664: 665: * ChangeLog, Makefile: Summary: 0.98r0 666: 667: * CMakeLists.txt: Summary: Some tests for setup.iss 668: 669: 2015-09-15 Nicolas J Brouard <nbrouard@tugault.ined.fr> 670: 671: * (Module): 0.98r0 Some new graphs, some bugs fixed: model=1+age+V2. etc 672: 673: 2015-08-18 Nicolas J Brouard <nbrouard@tugault.local> 674: 675: * (Module): Adding error when the covariance matrix doesn't contain the exact number of lines required by the model line. 676: 677: 2015-08-03 Nicolas J Brouard <nbrouard@tugault.ined.fr> 678: 679: * (Module): Changing Variance of one-step probabilities into Standard deviation (thanks to Yao-Chi Shih) 680: 681: 2014-12-16 Nicolas J Brouard <nbrouard@tugault.local> 682: 683: * imach.c (Module): Merging 1.61 to 1.162 684: 685: 2014-09-02 Nicolas J Brouard <nbrouard@njbrouard.ined.fr> 686: 687: * prevlim.h (Module): 688: 689: 2014-01-26 Nicolas J Brouard <nbrouard@ledermann-2.local> 690: 691: * imach.c (Module): Trying to merge old staffs together while being at Tokyo. Not tested... 692: (Module): Version 0.98nR Running ok, but output format still only works for three covariates. 693: 694: 2010-04-29 brouard <brouard@ined.fr> 695: 696: * imach.c (Module): Checking covariates for more complex models 697: than V1+V2. A lot of change to be done. Unstable. 698: 699: 2010-04-26 brouard <brouard@ined.fr> 700: 701: * imach.c (Module): merging some libgsl code. Fixing computation 702: of likelione (using inter/intrapolation if mle = 0) in order to 703: get same likelihood as if mle=1. 704: Some cleaning of code and comments added. 705: 706: 2009-10-29 brouard <brouard@ined.fr> 707: 708: * imach.c (Module): Now imach stops if date of birth, at least year of birth, is not given. Some cleaning of the code. 709: 710: 2006-06-30 Brouard Nicolas <brouard@localhost> 711: 712: * imach.c (Module): Clarifications on computing e.j 713: 714: 2006-04-28 Brouard Nicolas <brouard@localhost> 715: 716: * imach.c (Module): Yes the sum of survivors was wrong since 717: imach-114 because nhstepm was no more computed in the age 718: loop. Now we define nhstepma in the age loop. 719: (Module): In order to speed up (in case of numerous covariates) we 720: compute health expectancies (without variances) in a first step 721: and then all the health expectancies with variances or standard 722: deviation (needs data from the Hessian matrices) which slows the 723: computation. 724: In the future we should be able to stop the program is only health 725: expectancies and graph are needed without standard deviations. 726: 727: 2006-04-04 LIEVRE Agnes <lievre@PC_00697.ined.fr> 728: 729: * imach.iss (Module): 730: 731: * imach.c (Module): bugs in evsi + population file added in htm output file 732: 733: 2006-03-22 LIEVRE Agnes <lievre@PC_00697.ined.fr> 734: 735: * imach.c (Module): 736: 737: 2006-03-20 Brouard Nicolas <brouard@localhost> 738: 739: * imach.c (Module): <title> changed, corresponds to .htm file 740: name. <head> headers where missing. 741: 742: * Makefile (Module): 0.98g 743: 744: * imach.c (Module): Weights can have a decimal point as for 745: English (a comma might work with a correct LC_NUMERIC environment, 746: otherwise the weight is truncated). 747: Modification of warning when the covariates values are not 0 or 748: 1. 749: Version 0.98g 750: 751: 2006-03-16 LIEVRE Agnes <lievre@PC_00697.ined.fr> 752: 753: * imach.c (Module): Comments concerning covariates added 754: 755: * noreg-setup.iss (Module): Version 0.98f 756: 757: * Makefile (Module): Version 0.98f make _windows=1 imachcyg-setup 758: 759: * imach.c (Module): refinements in the computation of lli if 760: status=-2 in order to have more reliable computation if stepm is 761: not 1 month. Version 0.98f 762: 763: 2006-03-15 Brouard Nicolas <brouard@localhost> 764: 765: * imach.c (Module): Bug if status = -2, the loglikelihood was 766: computed as likelihood omitting the logarithm. Version O.98e 767: 768: 2006-03-14 Brouard Nicolas <brouard@localhost> 769: 770: * imach.c (Module): varevsij Comments added explaining the second 771: table of variances if popbased=1 . 772: (Module): Covariances of eij, ekl added, graphs fixed, new html link. 773: (Module): Function pstamp added 774: (Module): Version 0.98d 775: 776: 2006-03-06 Brouard Nicolas <brouard@localhost> 777: 778: * imach.c (Module): Variance-covariance wrong links and 779: varian-covariance of ej. is needed (Saito). 780: 781: 2006-02-27 Brouard Nicolas <brouard@localhost> 782: 783: * imach.c (Module): One freematrix added in mlikeli! 784: 785: 2006-02-26 Brouard Nicolas <brouard@localhost> 786: 787: * imach.c (Module): Some improvements in processing parameter 788: filename with strsep. 789: 790: 2006-02-24 Brouard Nicolas <brouard@localhost> 791: 792: * imach.c (Module): Memory leaks checks with valgrind and: 793: datafile was not closed, some imatrix were not freed and on matrix 794: allocation too. 795: 796: 2006-02-09 LIEVRE Agnes <lievre@PC_00697.ined.fr> 797: 798: * strsep.c (Module): Inclusion because not in mingw32!!! 799: 800: * setup.iss.in (Module): IMACHSETUPVERSION added 801: 802: * noreg-setup.iss (Module): added to cvs tree 803: 804: * Makefile (Module): IMACHSETUPVERSION added 805: 806: 2006-01-30 BROUARD Nicolas <brouard@localhost> 807: 808: * Makefile (Module, Module): Latest Makefile 809: 810: * imach.c (Module): Back to gnuplot.exe instead of wgnuplot.exe 811: 812: 2006-01-27 BROUARD Nicolas <brouard@localhost> 813: 814: * noreg-setup.iss (Module): ADDED 815: 816: 2006-01-24 BROUARD Nicolas <brouard@localhost> 817: 818: * imach.c (Module): Comments (lines starting with a #) are allowed in data. 819: 820: 2005-10-25 Nicolas Brouard <brouard@dhcp171.recherche.ined.fr> 821: 822: * Makefile (Module): Added Imach98.pmsp and howto crete a dmg 823: 824: * createdmg.sh (Module): Creates a dmg disk for MaC OS/X 825: 826: 2005-09-30 <lievre@PC_00595> 827: 828: * imach.c (Module): sump fixed, loop imx fixed, and simplifications. 829: (Module): If the status is missing at the last wave but we know 830: that the person is alive, then we can code his/her status as -2 831: (instead of missing=-1 in earlier versions) and his/her 832: contributions to the likelihood is 1 - Prob of dying from last 833: health status (= 1-p13= p11+p12 in the easiest case of somebody in 834: the healthy state at last known wave). Version is 0.98 835: 836: 2004-05-20 Brouard Nicolas <brouard@localhost> 837: 838: * imach.c (Repository): 839: Agnes added a direct estimation of mortality (without the need of 840: computing period prevalence and differential mortality). Thus here 841: is version 0.97a which has been distributed to some people at 842: REVES 16 in Brugge using an Inno setup.exe for PCs. Estimates of 843: mortality using covariates is not done today. Estimating direct 844: mortality is a very different process because it doesn't need 845: interpolation because it is easy to get the lx from the force of 846: the mortality mux in the simplest case as for a Gompertz (log mux 847: = a + b*x . But we have been able to incorporate the new code 848: within former imach program (0.96d) without deteriorating too much 849: the understanding of the program. 850: 851: 2003-06-25 <brouard@BROUARD> 852: 853: * imach.c (Module): On windows (cygwin) function asctime_r doesn't 854: exist so I changed back to asctime which exists. 855: (Module): Version 0.96b 856: 857: 2003-06-24 <brouard@BROUARD> 858: 859: * imach.c (Module): Some bugs corrected for windows. Also, when 860: mle=-1 a template is output in file "or"mypar.txt with the design 861: of the covariance matrix to be input. 862: 863: 2003-06-25 Brouard Nicolas <brouard@homebrou> 864: 865: * imach.c (Repository): Duplicated warning errors corrected. 866: (Repository): Elapsed time after each iteration is now output. It 867: helps to forecast when convergence will be reached. Elapsed time 868: is stamped in powell. We created a new html file for the graphs 869: concerning matrix of covariance. It has extension -cov.htm. 870: 871: 2003-06-23 Brouard Nicolas <brouard@homebrou> 872: 873: * 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. 874: 875: 2003-06-18 Brouard Nicolas <brouard@homebrou> 876: 877: * imach.c (Repository): Suppress abusive calls to append to file html (solution was to put fichtm as a global variable. Version 0.96 878: 879: 2003-06-17 Brouard Nicolas <brouard@homebrou> 880: 881: 882: * imach.c (Repository): Check when date of death was earlier that 883: current date of interview. It may happen when the death was just 884: prior to the death. In this case, dh was negative and likelihood 885: was wrong (infinity). We still send an "Error" but patch by 886: assuming that the date of death was just one stepm after the 887: interview. 888: (Repository): Because some people have very long ID (first column) 889: we changed int to long in num[] and we added a new lvector for 890: memory allocation. But we also truncated to 8 characters (left 891: truncation) 892: 893: (Repository): No more line truncation errors. 894: 895: 2003-06-14 Brouard Nicolas <brouard@homebrou> 896: 897: * imach.c (Repository): Add of a routine likelione (likelihood 898: only once) which prints on a text file 899: (ilk) the contributions to the likelihood for each 900: individual/wave. 901: 902: 2003-06-13 Brouard Nicolas <brouard@homebrou> 903: * imach.c (Repository): Replace "freqsummary" at a correct 904: place. It differs from routine "prevalence" which may be called 905: many times. Probs is memory consuming and must be used with 906: parcimony. 907: 908: 2003-06-17 <brouard@BROUARD> 909: 910: * timeval.h (Module): Added included file to make use of 911: gettimeofday working on win32 with cygwin. 912: 913: 2003-05-16 Brouard Nicolas <brouard@brouard> 914: * imach.c (Module): Insert of a warning if the delay between two 915: waves is negative. 916: 917: * imach.c (Module): 918: 919: 2003-05-03 <brouard@BROUARD> 920: 921: * imach.c: Some cleaning 922: 923: 2003-03-28 <brouard@BROUARD> 924: * imach.c (Module): In version up to 0.92 likelihood was computed 925: as if date of death was unknown. Death was treated as any other 926: health state: the date of the interview describes the actual state 927: and not the date of a change in health state. The former idea was 928: to consider that at each interview the state was recorded 929: (healthy, disable or death) and IMaCh was corrected; but when we 930: introduced the exact date of death then we should have modified 931: the contribution of an exact death to the likelihood. This new 932: contribution is smaller and very dependent of the step unit 933: stepm. It is no more the probability to die between last interview 934: and month of death but the probability to survive from last 935: interview up to one month before death multiplied by the 936: probability to die within a month. Thanks to Chris 937: Jackson for correcting this bug. Former versions increased 938: mortality artificially. The bad side is that we add another loop 939: which slows down the processing. The difference can be up to 10% 940: lower mortality. 941: 942: 2002-05-30 <brouard@BROUARD> 943: 944: * imach.c (Module): Add correlation matrix of one-step 945: probabilities and covariance matrix 946: