imach/src/ChangeLog - view

File: [Local Repository] / imach / src / ChangeLog
Revision 1.55: download - view: text, annotated - select for diffs
Wed Jul 27 14:50:01 2022 UTC (22 months, 3 weeks ago) by brouard
Branches: MAIN
CVS tags: HEAD

*** empty log message ***

1: 2022-07-25 Brouard Nicolas <brouard@brouard.name> 2: 3: * imach.c (Module): Error cptcovn instead of nsd in bmij (was 4: coredumped, revealed by Feiuno, thank you. 5: 6: 2022-06-02 Brouard Nicolas <brouard@brouard.name> 7: 8: * imach.c (Module): Adding the Wald tests from the log to the main 9: htm for better display of the maximum likelihood estimators. 10: 11: 2022-05-30 Brouard Nicolas <brouard@brouard.name> 12: 13: * imach.c: With products of covariates (age or dummies or 14: quantitatives), estimates of parameters were good but when 15: estimating variances the positions of newly created covariates 16: were wrongly assigned and results were wrong. Thank to Feinua_Sun! 17: 18: 2022-05-24 Nicolas Brouard <brouard@ined.fr> 19: 20: * imach.c (Module): Some attempts to find a bug of wrong estimates 21: of confidencce intervals with product in the equation modelC 22: 23: 2022-05-15 Nicolas Brouard <brouard@ined.fr> 24: 25: * imach.c (Module): Some minor improvements 26: 27: 2022-04-13 Brouard Nicolas <brouard@brouard.name> 28: 29: * imach.c (Module): Adding link to text data files 30: 31: 2022-04-11 Nicolas Brouard <brouard@ined.fr> 32: 33: * imach.c (Module): Error in rewriting the 'r' file with yearsfproj or yearsbproj fixed 34: 35: 2022-04-05 Brouard Nicolas <brouard@ined.fr> 36: 37: * imach.c (Module): Fixed covariates (dummy or quantitative) 38: with missing values have never been allowed but are ERRORS and 39: program quits. Standard deviations of fixed covariates were 40: wrongly computed. Mean and standard deviations of time varying 41: covariates are still not computed. 42: 43: 2022-03-17 Brouard Nicolas <brouard@ined.fr> 44: 45: * 99r25 Improvements in fixing discrepancies between covariates 46: missing in result line but in model. 47: 48: *Wed Mar 31 2021 Nicolas Brouard <brouard@ined.fr> 49: * imach.c (Module): Still bugs in the result loop. Thank to Holly Benett 50: 51: *2021-03-08 Nicolas Brouard <brouard@ined.fr> 52: 53: * ./ (Module): Fixed bug on result: 54: 55: *2021-02-20 Nicolas Brouard <brouard@ined.fr> 56: 57: * imach.c (Module): Fix bug on quitting after result lines! 58: (Module): Version 0.99r21 59: 60: 2021-02-12 Nicolas Brouard <brouard@ined.fr> 61: 62: * imach.c (Module): The use of a Windows BOM (huge) file is now an error 63: 64: 65: 2021-02-11 Nicolas Brouard <brouard@ined.fr> 66: 67: * (Module): imach.c Someone entered 'results:' instead of 'result:'. Now it is an error which is printed. 68: 69: 2020-02-22 Brouard Nicolas <brouard@ined.fr> 70: 71: * (Module): imach.c Update mle=-3 (for computing Life expectancy 72: and life table from the data without any state) 73: 74: 2019-05-20 Brouard Nicolas <brouard@ined.fr> 75: 76: * version.h: Summary: version 0.99.r19 77: 78: * imach.c: Summary: Projection syntax simplified 79: We can now start projections, forward or backward, from the mean date 80: of inteviews up to or down to a number of years of projection: 81: prevforecast=1 yearsfproj=15.3 mobil_average=0 82: or 83: prevforecast=1 starting-proj-date=1/1/2007 final-proj-date=12/31/2017 mobil_average=0 84: or 85: prevbackcast=1 yearsbproj=12.3 mobil_average=1 86: or 87: prevbackcast=1 starting-back-date=1/10/1999 final-back-date=1/1/1985 mobil_average=1 88: 89: 2019-05-18 Brouard Nicolas <brouard@ined.fr> 90: 91: * imach.c: Summary: doxygen tex bug 92: 93: 2019-05-16 Brouard Nicolas <brouard@ined.fr> 94: 95: * imach.c: Summary: There was some wrong lines added 96: 97: 2019-05-09 Brouard Nicolas <brouard@ined.fr> 98: 99: * Makefile, imach.c: *** empty log message *** 100: 101: * ChangeLog: Summary: C= 102: 103: * imach.c: Summary: Some updates 104: 105: * imach.c: Summary: Before ncovmax 106: 107: * version.h: *** empty log message *** 108: 109: * imach.c: Summary: 0.99r18 unlimited number of individuals 110: 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. 111: 112: 2018-12-13 Brouard Nicolas <brouard@ined.fr> 113: 114: * imach.c: Summary: Bug for young ages (<-30) will be in r17 115: 116: 2018-12-05 Brouard Nicolas <brouard@ined.fr> 117: 118: * Makefile: Summary: 0.99r16 119: 120: 2018-05-02 Brouard Nicolas <brouard@ined.fr> 121: 122: * imach.c: Summary: Some bugs fixed 123: 124: 2018-05-02 Brouard Nicolas <brouard@ined.fr> 125: 126: * imach.c: Summary: Bug for young ages (<-30) will be in r17 127: 128: 2018-05-02 Brouard Nicolas <brouard@ined.fr> 129: 130: * imach.c: Summary: Some bugs fixed 131: 132: 2018-05-01 Brouard Nicolas <brouard@ined.fr> 133: 134: * imach.c: 135: Summary: Bug fixed by providing frequencies only for non missing covariates 136: 137: 2018-04-27 Brouard Nicolas <brouard@ined.fr> 138: 139: * imach.c: Summary: some minor bugs 140: 141: 2018-04-21 Brouard Nicolas <brouard@ined.fr> 142: 143: * README.txt: *** empty log message *** 144: 145: * imach.c: Summary: Some bugs fixed, valgrind tested 146: 147: 2018-04-20 Brouard Nicolas <brouard@ined.fr> 148: 149: * Makefile, README.txt, setup.iss: Summary: imach 0.99r16 150: 151: * imach.c: 152: Summary: Computing mean and stdeviation of fixed quantitative variables 153: 154: 2018-04-19 Brouard Nicolas <brouard@ined.fr> 155: 156: * imach.c: Summary: Some minor bugs fixed 157: 158: * version.h: *** empty log message *** 159: 160: 2018-02-27 Brouard Nicolas <brouard@ined.fr> 161: 162: * imach.c: *** empty log message *** 163: 164: * imach.c: Summary: Adding second argument for quitting 165: 166: 2018-02-21 Brouard Nicolas <brouard@ined.fr> 167: 168: * Makefile, README.txt, imach.c: Summary: 0.99r15 169: New Makefile with recent VirtualBox 5.26. Bug in sqrt negatve in imach.c 170: 171: 2017-07-20 Brouard Nicolas <brouard@ined.fr> 172: 173: * imach.c: Summary: temporary working 174: 175: 2017-07-19 Brouard Nicolas <brouard@ined.fr> 176: 177: * imach.c: Summary: Bug for mobil_average=0 and prevforecast fixed(?) 178: 179: 2017-07-17 Brouard Nicolas <brouard@ined.fr> 180: 181: * imach.c: Summary: BOM files can be read now 182: 183: 2017-06-30 Brouard Nicolas <brouard@ined.fr> 184: 185: * imach.c: Summary: Graphs improvements 186: 187: * imach.c: Summary: Saito's color 188: 189: 2017-06-29 Brouard Nicolas <brouard@ined.fr> 190: 191: * imach.c: Summary: Version 0.99r14 192: 193: 2017-06-27 Brouard Nicolas <brouard@ined.fr> 194: 195: * imach.c: Summary: More documentation on projections 196: 197: * imach.c: Summary: Color of backprojection changed from 6 to 5(yellow) 198: 199: * imach.c: Summary: Some bug with rint 200: 201: 2017-05-24 Brouard Nicolas <brouard@ined.fr> 202: 203: * imach.c: *** empty log message *** 204: 205: 2017-05-23 Brouard Nicolas <brouard@ined.fr> 206: 207: * imach.c: Summary: Code into subroutine, cleanings 208: 209: 2017-05-18 Brouard Nicolas <brouard@ined.fr> 210: 211: * imach.c: 212: Summary: backprojection and confidence intervals of backprevalence 213: 214: 2017-05-13 Brouard Nicolas <brouard@ined.fr> 215: 216: * imach.c: Summary: temporary save for backprojection 217: 218: * imach.c: Summary: Version 0.99r13 (improvements and bugs fixed) 219: 220: 2017-04-26 Brouard Nicolas <brouard@ined.fr> 221: 222: * imach.c: Summary: imach 0.99r13 Some bugs fixed 223: 224: * imach.c: Summary: Labels in graphs 225: 226: 2017-04-24 Brouard Nicolas <brouard@ined.fr> 227: 228: * imach.c: Summary: to save 229: 230: 2017-04-18 Brouard Nicolas <brouard@ined.fr> 231: 232: * imach.c: *** empty log message *** 233: 234: 2017-04-05 Brouard Nicolas <brouard@ined.fr> 235: 236: * imach.c: Summary: Bug in E_ as well as in T_ fixed nres-1 vs k1-1 237: 238: 2017-04-04 Brouard Nicolas <brouard@ined.fr> 239: 240: * imach.c: Summary: Gnuplot indexations fixed (humm) 241: 242: * imach.c: 243: Summary: Some errors to warnings only if date of death is unknown but status is death we could set to pi3 244: 245: 2017-04-03 Brouard Nicolas <brouard@ined.fr> 246: 247: * imach.c: Summary: Version 0.99r12 248: Some cleanings, conformed with updated documentation. 249: 250: 2017-03-30 Brouard Nicolas <brouard@ined.fr> 251: 252: * version.h: *** empty log message *** 253: 254: * version.h: Summary:0.99r11 255: 256: 2017-03-29 Brouard Nicolas <brouard@ined.fr> 257: 258: * imach.c: Summary: Temp 259: 260: 2017-03-27 Brouard Nicolas <brouard@ined.fr> 261: 262: * imach.c: Summary: Temporary 263: 264: 2017-03-08 Brouard Nicolas <brouard@ined.fr> 265: 266: * imach.c: Summary: IMaCh version 0.99r10 bugs in gnuplot fixed 267: 268: * imach.c: Summary: Fixing data parameter line 269: 270: 2016-12-15 Brouard Nicolas <brouard@ined.fr> 271: 272: * imach.c: Summary: 0.99 in progress 273: 274: 2016-09-15 Brouard Nicolas <brouard@ined.fr> 275: 276: * imach.c: *** empty log message *** 277: 278: * imach.c: Summary: not working 279: 280: 2016-09-08 Brouard Nicolas <brouard@ined.fr> 281: 282: * imach.c: Summary: continue 283: 284: 2016-09-07 Brouard Nicolas <brouard@ined.fr> 285: 286: * imach.c: Summary: Starting values from frequencies 287: 288: * imach.c: *** empty log message *** 289: 290: 2016-09-02 Brouard Nicolas <brouard@ined.fr> 291: 292: * imach.c: *** empty log message *** 293: 294: 2016-08-30 Brouard Nicolas <brouard@ined.fr> 295: 296: * imach.c: Summary: Fixing a lots 297: 298: 2016-08-29 Brouard Nicolas <brouard@ined.fr> 299: 300: * imach.c: Summary: gnuplot problem in Back projection to fix 301: 302: * imach.c: Summary: Better 303: 304: 2016-08-26 Brouard Nicolas <brouard@ined.fr> 305: 306: * imach.c: 307: Summary: Improvement in Powell output in order to copy and paste 308: 309: * imach.c: Summary: Starting tests of 0.99 310: 311: * imach.c: Summary: to valgrind 312: 313: 2016-08-25 Brouard Nicolas <brouard@ined.fr> 314: 315: * imach.c: *** empty log message *** 316: 317: 2016-08-23 Brouard Nicolas <brouard@ined.fr> 318: 319: * imach.c: *** empty log message *** 320: 321: * imach.c: Summary: not working 322: 323: 2016-08-22 Brouard Nicolas <brouard@ined.fr> 324: 325: * imach.c: Summary: not working 326: 327: * imach.c: Summary: Not working 328: 329: 2016-07-23 Brouard Nicolas <brouard@ined.fr> 330: 331: * imach.c: Summary: Completing for func too 332: 333: 2016-07-22 Brouard Nicolas <brouard@ined.fr> 334: 335: * imach.c: Summary: Fixing some arrays, still debugging 336: 337: 2016-07-21 Brouard Nicolas <brouard@ined.fr> 338: 339: * imach.c: 340: Summary: 0.99 working (more or less) for Asian Workshop on multitate methods 341: 342: 2016-07-12 Brouard Nicolas <brouard@ined.fr> 343: 344: * imach.c: Summary: temp 345: 346: * imach.c: Summary: saving but not running 347: 348: 2016-07-01 Brouard Nicolas <brouard@ined.fr> 349: 350: * imach.c: Summary: Fixes 351: 352: 2016-02-19 Brouard Nicolas <brouard@ined.fr> 353: 354: * imach-0.98r.c: Summary: Kind of 0.98r? series, starting with r7 355: 356: * setup.iss, Makefile, version.h: *** empty log message *** 357: 358: * imach.c: Summary: temporary 359: 360: 2016-02-17 Brouard Nicolas <brouard@ined.fr> 361: 362: * imach.c: Summary: Probably last 0.98 stable version 0.98r6 363: 364: 2016-02-16 Brouard Nicolas <brouard@ined.fr> 365: 366: * imach.c: Summary: minor bug 367: 368: * imach.c: Summary: 0.99r2 369: 370: 2016-02-15 Brouard Nicolas <brouard@ined.fr> 371: 372: * imach.c: *** empty log message *** 373: 374: 2016-02-12 Brouard Nicolas <brouard@ined.fr> 375: 376: * imach.c: Summary: 0.99 Back projections 377: 378: 2015-12-23 Brouard Nicolas <brouard@ined.fr> 379: 380: * imach.c: Summary: Experimental backcast 381: 382: 2015-12-18 Brouard Nicolas <brouard@ined.fr> 383: 384: * imach.c: Summary: 0.98r4 Warning and status=-2 385: 386: Version 0.98r4 is now: 387: - displaying an error when status is -1, date of interview unknown and date of death known; 388: - permitting a status -2 when the vital status is unknown at a known date of right truncation. 389: Older changes concerning s=-2, dating from 2005 have been supersed. 390: 391: * Makefile: *** empty log message *** 392: 393: 2015-12-16 Brouard Nicolas <brouard@ined.fr> 394: 395: * Makefile: *** empty log message *** 396: 397: * imach.c: Summary: 0.98r4 working 398: 399: * Makefile: Summary: linuxrpm added 400: 401: * Makefile: *** empty log message *** 402: 403: * imach.c: Summary: temporary not working 404: 405: 2015-12-11 Brouard Nicolas <brouard@ined.fr> 406: 407: * imach.c: Summary: 0.98r4 408: 409: 2015-12-08 Brouard Nicolas <brouard@ined.fr> 410: 411: * Makefile: *** empty log message *** 412: 413: 2015-11-21 Brouard Nicolas <brouard@ined.fr> 414: 415: * Makefile: 416: Summary: VBoxManage change from execute --image to run --exe and --wait-exit doesn't exist any more 417: 418: * imach.c: Summary: minor typo 419: 420: * imach.c: Summary: 0.98r3 with some graph of projected cross-sectional 421: 422: Author: Nicolas Brouard 423: 424: 2015-11-18 Brouard Nicolas <brouard@ined.fr> 425: 426: * imach.c: Summary: Start working on projected prevalences 427: 428: 2015-11-17 Brouard Nicolas <brouard@ined.fr> 429: 430: * imach.c: Summary: Adding ftolpl parameter 431: Author: N Brouard 432: 433: We had difficulties to get smoothed confidence intervals. It was due 434: to the period prevalence which wasn't computed accurately. The inner 435: parameter ftolpl is now an outer parameter of the .imach parameter 436: file after estepm. If ftolpl is small 1.e-4 and estepm too, 437: computation are long. 438: 439: * version.h: *** empty log message *** 440: 441: * imach.c: Summary: temporary 442: 443: 2015-10-27 Brouard Nicolas <brouard@ined.fr> 444: 445: * Makefile: *** empty log message *** 446: 447: * setup.iss.in: Summary: some new subdirs 448: 449: * imach.c: *** empty log message *** 450: 451: * setup.iss.in: 452: Summary: Adding gnuplot 5.1 with bin and etc for pango fonts 453: 454: 2015-10-24 Brouard Nicolas <brouard@ined.fr> 455: 456: * imach.c: *** empty log message *** 457: 458: 2015-10-23 Brouard Nicolas <brouard@ined.fr> 459: 460: * imach.c: 461: Summary: 0.98r3 some clarification for graphs on likelihood contributions 462: 463: 2015-10-01 Brouard Nicolas <brouard@ined.fr> 464: 465: * imach.c: Summary: Some new graphs of contribution to likelihood 466: 467: 2015-09-30 Brouard Nicolas <brouard@ined.fr> 468: 469: * Makefile: 470: Summary: Version read from version.h, itself read with cmake 471: 472: * imach.c: Summary: looking at better estimation of the hessian 473: 474: Also a better criteria for convergence to the period prevalence And 475: therefore adding the number of years needed to converge. (The 476: prevalence in any alive state shold sum to one 477: 478: * Makefile: *** empty log message *** 479: 480: 2015-09-22 Brouard Nicolas <brouard@ined.fr> 481: 482: * imach.c: 483: Summary: Adding some overall graph on contribution to likelihood. Might change 484: 485: 2015-09-15 Brouard Nicolas <brouard@ined.fr> 486: 487: * Makefile: *** empty log message *** 488: 489: * imach.c: Summary: 0.98r0 490: 491: - Some new graphs like survival functions 492: - Some bugs fixed like model=1+age+V2. 493: 494: * ChangeLog, Makefile: Summary: 0.98r0 495: 496: * CMakeLists.txt: Summary: Some tests for setup.iss 497: 498: 2015-09-15 Nicolas J Brouard <nbrouard@tugault.ined.fr> 499: 500: * (Module): 0.98r0 Some new graphs, some bugs fixed: model=1+age+V2. etc 501: 502: 2015-08-18 Nicolas J Brouard <nbrouard@tugault.local> 503: 504: * (Module): Adding error when the covariance matrix doesn't contain the exact number of lines required by the model line. 505: 506: 2015-08-03 Nicolas J Brouard <nbrouard@tugault.ined.fr> 507: 508: * (Module): Changing Variance of one-step probabilities into Standard deviation (thanks to Yao-Chi Shih) 509: 510: 2014-12-16 Nicolas J Brouard <nbrouard@tugault.local> 511: 512: * imach.c (Module): Merging 1.61 to 1.162 513: 514: 2014-09-02 Nicolas J Brouard <nbrouard@njbrouard.ined.fr> 515: 516: * prevlim.h (Module): 517: 518: 2014-01-26 Nicolas J Brouard <nbrouard@ledermann-2.local> 519: 520: * imach.c (Module): Trying to merge old staffs together while being at Tokyo. Not tested... 521: (Module): Version 0.98nR Running ok, but output format still only works for three covariates. 522: 523: 2010-04-29 brouard <brouard@ined.fr> 524: 525: * imach.c (Module): Checking covariates for more complex models 526: than V1+V2. A lot of change to be done. Unstable. 527: 528: 2010-04-26 brouard <brouard@ined.fr> 529: 530: * imach.c (Module): merging some libgsl code. Fixing computation 531: of likelione (using inter/intrapolation if mle = 0) in order to 532: get same likelihood as if mle=1. 533: Some cleaning of code and comments added. 534: 535: 2009-10-29 brouard <brouard@ined.fr> 536: 537: * imach.c (Module): Now imach stops if date of birth, at least year of birth, is not given. Some cleaning of the code. 538: 539: 2006-06-30 Brouard Nicolas <brouard@localhost> 540: 541: * imach.c (Module): Clarifications on computing e.j 542: 543: 2006-04-28 Brouard Nicolas <brouard@localhost> 544: 545: * imach.c (Module): Yes the sum of survivors was wrong since 546: imach-114 because nhstepm was no more computed in the age 547: loop. Now we define nhstepma in the age loop. 548: (Module): In order to speed up (in case of numerous covariates) we 549: compute health expectancies (without variances) in a first step 550: and then all the health expectancies with variances or standard 551: deviation (needs data from the Hessian matrices) which slows the 552: computation. 553: In the future we should be able to stop the program is only health 554: expectancies and graph are needed without standard deviations. 555: 556: 2006-04-04 LIEVRE Agnes <lievre@PC_00697.ined.fr> 557: 558: * imach.iss (Module): 559: 560: * imach.c (Module): bugs in evsi + population file added in htm output file 561: 562: 2006-03-22 LIEVRE Agnes <lievre@PC_00697.ined.fr> 563: 564: * imach.c (Module): 565: 566: 2006-03-20 Brouard Nicolas <brouard@localhost> 567: 568: * imach.c (Module): <title> changed, corresponds to .htm file 569: name. <head> headers where missing. 570: 571: * Makefile (Module): 0.98g 572: 573: * imach.c (Module): Weights can have a decimal point as for 574: English (a comma might work with a correct LC_NUMERIC environment, 575: otherwise the weight is truncated). 576: Modification of warning when the covariates values are not 0 or 577: 1. 578: Version 0.98g 579: 580: 2006-03-16 LIEVRE Agnes <lievre@PC_00697.ined.fr> 581: 582: * imach.c (Module): Comments concerning covariates added 583: 584: * noreg-setup.iss (Module): Version 0.98f 585: 586: * Makefile (Module): Version 0.98f make _windows=1 imachcyg-setup 587: 588: * imach.c (Module): refinements in the computation of lli if 589: status=-2 in order to have more reliable computation if stepm is 590: not 1 month. Version 0.98f 591: 592: 2006-03-15 Brouard Nicolas <brouard@localhost> 593: 594: * imach.c (Module): Bug if status = -2, the loglikelihood was 595: computed as likelihood omitting the logarithm. Version O.98e 596: 597: 2006-03-14 Brouard Nicolas <brouard@localhost> 598: 599: * imach.c (Module): varevsij Comments added explaining the second 600: table of variances if popbased=1 . 601: (Module): Covariances of eij, ekl added, graphs fixed, new html link. 602: (Module): Function pstamp added 603: (Module): Version 0.98d 604: 605: 2006-03-06 Brouard Nicolas <brouard@localhost> 606: 607: * imach.c (Module): Variance-covariance wrong links and 608: varian-covariance of ej. is needed (Saito). 609: 610: 2006-02-27 Brouard Nicolas <brouard@localhost> 611: 612: * imach.c (Module): One freematrix added in mlikeli! 613: 614: 2006-02-26 Brouard Nicolas <brouard@localhost> 615: 616: * imach.c (Module): Some improvements in processing parameter 617: filename with strsep. 618: 619: 2006-02-24 Brouard Nicolas <brouard@localhost> 620: 621: * imach.c (Module): Memory leaks checks with valgrind and: 622: datafile was not closed, some imatrix were not freed and on matrix 623: allocation too. 624: 625: 2006-02-09 LIEVRE Agnes <lievre@PC_00697.ined.fr> 626: 627: * strsep.c (Module): Inclusion because not in mingw32!!! 628: 629: * setup.iss.in (Module): IMACHSETUPVERSION added 630: 631: * noreg-setup.iss (Module): added to cvs tree 632: 633: * Makefile (Module): IMACHSETUPVERSION added 634: 635: 2006-01-30 BROUARD Nicolas <brouard@localhost> 636: 637: * Makefile (Module, Module): Latest Makefile 638: 639: * imach.c (Module): Back to gnuplot.exe instead of wgnuplot.exe 640: 641: 2006-01-27 BROUARD Nicolas <brouard@localhost> 642: 643: * noreg-setup.iss (Module): ADDED 644: 645: 2006-01-24 BROUARD Nicolas <brouard@localhost> 646: 647: * imach.c (Module): Comments (lines starting with a #) are allowed in data. 648: 649: 2005-10-25 Nicolas Brouard <brouard@dhcp171.recherche.ined.fr> 650: 651: * Makefile (Module): Added Imach98.pmsp and howto crete a dmg 652: 653: * createdmg.sh (Module): Creates a dmg disk for MaC OS/X 654: 655: 2005-09-30 <lievre@PC_00595> 656: 657: * imach.c (Module): sump fixed, loop imx fixed, and simplifications. 658: (Module): If the status is missing at the last wave but we know 659: that the person is alive, then we can code his/her status as -2 660: (instead of missing=-1 in earlier versions) and his/her 661: contributions to the likelihood is 1 - Prob of dying from last 662: health status (= 1-p13= p11+p12 in the easiest case of somebody in 663: the healthy state at last known wave). Version is 0.98 664: 665: 2004-05-20 Brouard Nicolas <brouard@localhost> 666: 667: * imach.c (Repository): 668: Agnes added a direct estimation of mortality (without the need of 669: computing period prevalence and differential mortality). Thus here 670: is version 0.97a which has been distributed to some people at 671: REVES 16 in Brugge using an Inno setup.exe for PCs. Estimates of 672: mortality using covariates is not done today. Estimating direct 673: mortality is a very different process because it doesn't need 674: interpolation because it is easy to get the lx from the force of 675: the mortality mux in the simplest case as for a Gompertz (log mux 676: = a + b*x . But we have been able to incorporate the new code 677: within former imach program (0.96d) without deteriorating too much 678: the understanding of the program. 679: 680: 2003-06-25 <brouard@BROUARD> 681: 682: * imach.c (Module): On windows (cygwin) function asctime_r doesn't 683: exist so I changed back to asctime which exists. 684: (Module): Version 0.96b 685: 686: 2003-06-24 <brouard@BROUARD> 687: 688: * imach.c (Module): Some bugs corrected for windows. Also, when 689: mle=-1 a template is output in file "or"mypar.txt with the design 690: of the covariance matrix to be input. 691: 692: 2003-06-25 Brouard Nicolas <brouard@homebrou> 693: 694: * imach.c (Repository): Duplicated warning errors corrected. 695: (Repository): Elapsed time after each iteration is now output. It 696: helps to forecast when convergence will be reached. Elapsed time 697: is stamped in powell. We created a new html file for the graphs 698: concerning matrix of covariance. It has extension -cov.htm. 699: 700: 2003-06-23 Brouard Nicolas <brouard@homebrou> 701: 702: * 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. 703: 704: 2003-06-18 Brouard Nicolas <brouard@homebrou> 705: 706: * imach.c (Repository): Suppress abusive calls to append to file html (solution was to put fichtm as a global variable. Version 0.96 707: 708: 2003-06-17 Brouard Nicolas <brouard@homebrou> 709: 710: 711: * imach.c (Repository): Check when date of death was earlier that 712: current date of interview. It may happen when the death was just 713: prior to the death. In this case, dh was negative and likelihood 714: was wrong (infinity). We still send an "Error" but patch by 715: assuming that the date of death was just one stepm after the 716: interview. 717: (Repository): Because some people have very long ID (first column) 718: we changed int to long in num[] and we added a new lvector for 719: memory allocation. But we also truncated to 8 characters (left 720: truncation) 721: 722: (Repository): No more line truncation errors. 723: 724: 2003-06-14 Brouard Nicolas <brouard@homebrou> 725: 726: * imach.c (Repository): Add of a routine likelione (likelihood 727: only once) which prints on a text file 728: (ilk) the contributions to the likelihood for each 729: individual/wave. 730: 731: 2003-06-13 Brouard Nicolas <brouard@homebrou> 732: * imach.c (Repository): Replace "freqsummary" at a correct 733: place. It differs from routine "prevalence" which may be called 734: many times. Probs is memory consuming and must be used with 735: parcimony. 736: 737: 2003-06-17 <brouard@BROUARD> 738: 739: * timeval.h (Module): Added included file to make use of 740: gettimeofday working on win32 with cygwin. 741: 742: 2003-05-16 Brouard Nicolas <brouard@brouard> 743: * imach.c (Module): Insert of a warning if the delay between two 744: waves is negative. 745: 746: * imach.c (Module): 747: 748: 2003-05-03 <brouard@BROUARD> 749: 750: * imach.c: Some cleaning 751: 752: 2003-03-28 <brouard@BROUARD> 753: * imach.c (Module): In version up to 0.92 likelihood was computed 754: as if date of death was unknown. Death was treated as any other 755: health state: the date of the interview describes the actual state 756: and not the date of a change in health state. The former idea was 757: to consider that at each interview the state was recorded 758: (healthy, disable or death) and IMaCh was corrected; but when we 759: introduced the exact date of death then we should have modified 760: the contribution of an exact death to the likelihood. This new 761: contribution is smaller and very dependent of the step unit 762: stepm. It is no more the probability to die between last interview 763: and month of death but the probability to survive from last 764: interview up to one month before death multiplied by the 765: probability to die within a month. Thanks to Chris 766: Jackson for correcting this bug. Former versions increased 767: mortality artificially. The bad side is that we add another loop 768: which slows down the processing. The difference can be up to 10% 769: lower mortality. 770: 771: 2002-05-30 <brouard@BROUARD> 772: * 773: * imach.c (Module): Add correlation matrix of one-step 774: probabilities and covariance matrix 775: 776: 777: 778: 779: 780: ======= 781: 2022-07-23 Nicolas Brouard <brouard@ined.fr> 782: 783: * r29 W and not sqrt(Wald) 784: 785: 2022-07-22 Nicolas Brouard <brouard@ined.fr> 786: 787: * imach.c (Module): Output of Wald test in the htm file and not only in the log. 788: 789: 2022-05-30 Brouard Nicolas <brouard@brouard.name> 790: 791: * imach.c: With products of covariates (age or dummies or 792: quantitatives), estimates of parameters were good but when 793: estimating variances the positions of newly created covariates 794: were wrongly assigned and results were wrong. Thank to Feinua_Sun! 795: 796: 2022-05-24 Nicolas Brouard <brouard@ined.fr> 797: 798: * imach.c (Module): Some attempts to find a bug of wrong estimates 799: of confidencce intervals with product in the equation modelC 800: 801: 2022-05-15 Nicolas Brouard <brouard@ined.fr> 802: 803: * imach.c (Module): Some minor improvements 804: 805: 2022-04-13 Brouard Nicolas <brouard@brouard.name> 806: 807: * imach.c (Module): Adding link to text data files 808: 809: 2022-04-11 Nicolas Brouard <brouard@ined.fr> 810: 811: * imach.c (Module): Error in rewriting the 'r' file with yearsfproj or yearsbproj fixed 812: 813: 2022-04-05 Brouard Nicolas <brouard@ined.fr> 814: 815: * imach.c (Module): Fixed covariates (dummy or quantitative) 816: with missing values have never been allowed but are ERRORS and 817: program quits. Standard deviations of fixed covariates were 818: wrongly computed. Mean and standard deviations of time varying 819: covariates are still not computed. 820: 821: 2022-03-17 Brouard Nicolas <brouard@ined.fr> 822: 823: * 99r25 Improvements in fixing discrepancies between covariates 824: missing in result line but in model. 825: 826: *Wed Mar 31 2021 Nicolas Brouard <brouard@ined.fr> 827: * imach.c (Module): Still bugs in the result loop. Thank to Holly Benett 828: 829: *2021-03-08 Nicolas Brouard <brouard@ined.fr> 830: 831: * ./ (Module): Fixed bug on result: 832: 833: *2021-02-20 Nicolas Brouard <brouard@ined.fr> 834: 835: * imach.c (Module): Fix bug on quitting after result lines! 836: (Module): Version 0.99r21 837: 838: 2021-02-12 Nicolas Brouard <brouard@ined.fr> 839: 840: * imach.c (Module): The use of a Windows BOM (huge) file is now an error 841: 842: 843: 2021-02-11 Nicolas Brouard <brouard@ined.fr> 844: 845: * (Module): imach.c Someone entered 'results:' instead of 'result:'. Now it is an error which is printed. 846: 847: 2020-02-22 Brouard Nicolas <brouard@ined.fr> 848: 849: * (Module): imach.c Update mle=-3 (for computing Life expectancy 850: and life table from the data without any state) 851: 852: 2019-05-20 Brouard Nicolas <brouard@ined.fr> 853: 854: * version.h: Summary: version 0.99.r19 855: 856: * imach.c: Summary: Projection syntax simplified 857: We can now start projections, forward or backward, from the mean date 858: of inteviews up to or down to a number of years of projection: 859: prevforecast=1 yearsfproj=15.3 mobil_average=0 860: or 861: prevforecast=1 starting-proj-date=1/1/2007 final-proj-date=12/31/2017 mobil_average=0 862: or 863: prevbackcast=1 yearsbproj=12.3 mobil_average=1 864: or 865: prevbackcast=1 starting-back-date=1/10/1999 final-back-date=1/1/1985 mobil_average=1 866: 867: 2019-05-18 Brouard Nicolas <brouard@ined.fr> 868: 869: * imach.c: Summary: doxygen tex bug 870: 871: 2019-05-16 Brouard Nicolas <brouard@ined.fr> 872: 873: * imach.c: Summary: There was some wrong lines added 874: 875: 2019-05-09 Brouard Nicolas <brouard@ined.fr> 876: 877: * Makefile, imach.c: *** empty log message *** 878: 879: * ChangeLog: Summary: C= 880: 881: * imach.c: Summary: Some updates 882: 883: * imach.c: Summary: Before ncovmax 884: 885: * version.h: *** empty log message *** 886: 887: * imach.c: Summary: 0.99r18 unlimited number of individuals 888: 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. 889: 890: 2018-12-13 Brouard Nicolas <brouard@ined.fr> 891: 892: * imach.c: Summary: Bug for young ages (<-30) will be in r17 893: 894: 2018-12-05 Brouard Nicolas <brouard@ined.fr> 895: 896: * Makefile: Summary: 0.99r16 897: 898: 2018-05-02 Brouard Nicolas <brouard@ined.fr> 899: 900: * imach.c: Summary: Some bugs fixed 901: 902: 2018-05-02 Brouard Nicolas <brouard@ined.fr> 903: 904: * imach.c: Summary: Bug for young ages (<-30) will be in r17 905: 906: 2018-05-02 Brouard Nicolas <brouard@ined.fr> 907: 908: * imach.c: Summary: Some bugs fixed 909: 910: 2018-05-01 Brouard Nicolas <brouard@ined.fr> 911: 912: * imach.c: 913: Summary: Bug fixed by providing frequencies only for non missing covariates 914: 915: 2018-04-27 Brouard Nicolas <brouard@ined.fr> 916: 917: * imach.c: Summary: some minor bugs 918: 919: 2018-04-21 Brouard Nicolas <brouard@ined.fr> 920: 921: * README.txt: *** empty log message *** 922: 923: * imach.c: Summary: Some bugs fixed, valgrind tested 924: 925: 2018-04-20 Brouard Nicolas <brouard@ined.fr> 926: 927: * Makefile, README.txt, setup.iss: Summary: imach 0.99r16 928: 929: * imach.c: 930: Summary: Computing mean and stdeviation of fixed quantitative variables 931: 932: 2018-04-19 Brouard Nicolas <brouard@ined.fr> 933: 934: * imach.c: Summary: Some minor bugs fixed 935: 936: * version.h: *** empty log message *** 937: 938: 2018-02-27 Brouard Nicolas <brouard@ined.fr> 939: 940: * imach.c: *** empty log message *** 941: 942: * imach.c: Summary: Adding second argument for quitting 943: 944: 2018-02-21 Brouard Nicolas <brouard@ined.fr> 945: 946: * Makefile, README.txt, imach.c: Summary: 0.99r15 947: New Makefile with recent VirtualBox 5.26. Bug in sqrt negatve in imach.c 948: 949: 2017-07-20 Brouard Nicolas <brouard@ined.fr> 950: 951: * imach.c: Summary: temporary working 952: 953: 2017-07-19 Brouard Nicolas <brouard@ined.fr> 954: 955: * imach.c: Summary: Bug for mobil_average=0 and prevforecast fixed(?) 956: 957: 2017-07-17 Brouard Nicolas <brouard@ined.fr> 958: 959: * imach.c: Summary: BOM files can be read now 960: 961: 2017-06-30 Brouard Nicolas <brouard@ined.fr> 962: 963: * imach.c: Summary: Graphs improvements 964: 965: * imach.c: Summary: Saito's color 966: 967: 2017-06-29 Brouard Nicolas <brouard@ined.fr> 968: 969: * imach.c: Summary: Version 0.99r14 970: 971: 2017-06-27 Brouard Nicolas <brouard@ined.fr> 972: 973: * imach.c: Summary: More documentation on projections 974: 975: * imach.c: Summary: Color of backprojection changed from 6 to 5(yellow) 976: 977: * imach.c: Summary: Some bug with rint 978: 979: 2017-05-24 Brouard Nicolas <brouard@ined.fr> 980: 981: * imach.c: *** empty log message *** 982: 983: 2017-05-23 Brouard Nicolas <brouard@ined.fr> 984: 985: * imach.c: Summary: Code into subroutine, cleanings 986: 987: 2017-05-18 Brouard Nicolas <brouard@ined.fr> 988: 989: * imach.c: 990: Summary: backprojection and confidence intervals of backprevalence 991: 992: 2017-05-13 Brouard Nicolas <brouard@ined.fr> 993: 994: * imach.c: Summary: temporary save for backprojection 995: 996: * imach.c: Summary: Version 0.99r13 (improvements and bugs fixed) 997: 998: 2017-04-26 Brouard Nicolas <brouard@ined.fr> 999: 1000: * imach.c: Summary: imach 0.99r13 Some bugs fixed 1001: 1002: * imach.c: Summary: Labels in graphs 1003: 1004: 2017-04-24 Brouard Nicolas <brouard@ined.fr> 1005: 1006: * imach.c: Summary: to save 1007: 1008: 2017-04-18 Brouard Nicolas <brouard@ined.fr> 1009: 1010: * imach.c: *** empty log message *** 1011: 1012: 2017-04-05 Brouard Nicolas <brouard@ined.fr> 1013: 1014: * imach.c: Summary: Bug in E_ as well as in T_ fixed nres-1 vs k1-1 1015: 1016: 2017-04-04 Brouard Nicolas <brouard@ined.fr> 1017: 1018: * imach.c: Summary: Gnuplot indexations fixed (humm) 1019: 1020: * imach.c: 1021: Summary: Some errors to warnings only if date of death is unknown but status is death we could set to pi3 1022: 1023: 2017-04-03 Brouard Nicolas <brouard@ined.fr> 1024: 1025: * imach.c: Summary: Version 0.99r12 1026: Some cleanings, conformed with updated documentation. 1027: 1028: 2017-03-30 Brouard Nicolas <brouard@ined.fr> 1029: 1030: * version.h: *** empty log message *** 1031: 1032: * version.h: Summary:0.99r11 1033: 1034: 2017-03-29 Brouard Nicolas <brouard@ined.fr> 1035: 1036: * imach.c: Summary: Temp 1037: 1038: 2017-03-27 Brouard Nicolas <brouard@ined.fr> 1039: 1040: * imach.c: Summary: Temporary 1041: 1042: 2017-03-08 Brouard Nicolas <brouard@ined.fr> 1043: 1044: * imach.c: Summary: IMaCh version 0.99r10 bugs in gnuplot fixed 1045: 1046: * imach.c: Summary: Fixing data parameter line 1047: 1048: 2016-12-15 Brouard Nicolas <brouard@ined.fr> 1049: 1050: * imach.c: Summary: 0.99 in progress 1051: 1052: 2016-09-15 Brouard Nicolas <brouard@ined.fr> 1053: 1054: * imach.c: *** empty log message *** 1055: 1056: * imach.c: Summary: not working 1057: 1058: 2016-09-08 Brouard Nicolas <brouard@ined.fr> 1059: 1060: * imach.c: Summary: continue 1061: 1062: 2016-09-07 Brouard Nicolas <brouard@ined.fr> 1063: 1064: * imach.c: Summary: Starting values from frequencies 1065: 1066: * imach.c: *** empty log message *** 1067: 1068: 2016-09-02 Brouard Nicolas <brouard@ined.fr> 1069: 1070: * imach.c: *** empty log message *** 1071: 1072: 2016-08-30 Brouard Nicolas <brouard@ined.fr> 1073: 1074: * imach.c: Summary: Fixing a lots 1075: 1076: 2016-08-29 Brouard Nicolas <brouard@ined.fr> 1077: 1078: * imach.c: Summary: gnuplot problem in Back projection to fix 1079: 1080: * imach.c: Summary: Better 1081: 1082: 2016-08-26 Brouard Nicolas <brouard@ined.fr> 1083: 1084: * imach.c: 1085: Summary: Improvement in Powell output in order to copy and paste 1086: 1087: * imach.c: Summary: Starting tests of 0.99 1088: 1089: * imach.c: Summary: to valgrind 1090: 1091: 2016-08-25 Brouard Nicolas <brouard@ined.fr> 1092: 1093: * imach.c: *** empty log message *** 1094: 1095: 2016-08-23 Brouard Nicolas <brouard@ined.fr> 1096: 1097: * imach.c: *** empty log message *** 1098: 1099: * imach.c: Summary: not working 1100: 1101: 2016-08-22 Brouard Nicolas <brouard@ined.fr> 1102: 1103: * imach.c: Summary: not working 1104: 1105: * imach.c: Summary: Not working 1106: 1107: 2016-07-23 Brouard Nicolas <brouard@ined.fr> 1108: 1109: * imach.c: Summary: Completing for func too 1110: 1111: 2016-07-22 Brouard Nicolas <brouard@ined.fr> 1112: 1113: * imach.c: Summary: Fixing some arrays, still debugging 1114: 1115: 2016-07-21 Brouard Nicolas <brouard@ined.fr> 1116: 1117: * imach.c: 1118: Summary: 0.99 working (more or less) for Asian Workshop on multitate methods 1119: 1120: 2016-07-12 Brouard Nicolas <brouard@ined.fr> 1121: 1122: * imach.c: Summary: temp 1123: 1124: * imach.c: Summary: saving but not running 1125: 1126: 2016-07-01 Brouard Nicolas <brouard@ined.fr> 1127: 1128: * imach.c: Summary: Fixes 1129: 1130: 2016-02-19 Brouard Nicolas <brouard@ined.fr> 1131: 1132: * imach-0.98r.c: Summary: Kind of 0.98r? series, starting with r7 1133: 1134: * setup.iss, Makefile, version.h: *** empty log message *** 1135: 1136: * imach.c: Summary: temporary 1137: 1138: 2016-02-17 Brouard Nicolas <brouard@ined.fr> 1139: 1140: * imach.c: Summary: Probably last 0.98 stable version 0.98r6 1141: 1142: 2016-02-16 Brouard Nicolas <brouard@ined.fr> 1143: 1144: * imach.c: Summary: minor bug 1145: 1146: * imach.c: Summary: 0.99r2 1147: 1148: 2016-02-15 Brouard Nicolas <brouard@ined.fr> 1149: 1150: * imach.c: *** empty log message *** 1151: 1152: 2016-02-12 Brouard Nicolas <brouard@ined.fr> 1153: 1154: * imach.c: Summary: 0.99 Back projections 1155: 1156: 2015-12-23 Brouard Nicolas <brouard@ined.fr> 1157: 1158: * imach.c: Summary: Experimental backcast 1159: 1160: 2015-12-18 Brouard Nicolas <brouard@ined.fr> 1161: 1162: * imach.c: Summary: 0.98r4 Warning and status=-2 1163: 1164: Version 0.98r4 is now: 1165: - displaying an error when status is -1, date of interview unknown and date of death known; 1166: - permitting a status -2 when the vital status is unknown at a known date of right truncation. 1167: Older changes concerning s=-2, dating from 2005 have been supersed. 1168: 1169: * Makefile: *** empty log message *** 1170: 1171: 2015-12-16 Brouard Nicolas <brouard@ined.fr> 1172: 1173: * Makefile: *** empty log message *** 1174: 1175: * imach.c: Summary: 0.98r4 working 1176: 1177: * Makefile: Summary: linuxrpm added 1178: 1179: * Makefile: *** empty log message *** 1180: 1181: * imach.c: Summary: temporary not working 1182: 1183: 2015-12-11 Brouard Nicolas <brouard@ined.fr> 1184: 1185: * imach.c: Summary: 0.98r4 1186: 1187: 2015-12-08 Brouard Nicolas <brouard@ined.fr> 1188: 1189: * Makefile: *** empty log message *** 1190: 1191: 2015-11-21 Brouard Nicolas <brouard@ined.fr> 1192: 1193: * Makefile: 1194: Summary: VBoxManage change from execute --image to run --exe and --wait-exit doesn't exist any more 1195: 1196: * imach.c: Summary: minor typo 1197: 1198: * imach.c: Summary: 0.98r3 with some graph of projected cross-sectional 1199: 1200: Author: Nicolas Brouard 1201: 1202: 2015-11-18 Brouard Nicolas <brouard@ined.fr> 1203: 1204: * imach.c: Summary: Start working on projected prevalences 1205: 1206: 2015-11-17 Brouard Nicolas <brouard@ined.fr> 1207: 1208: * imach.c: Summary: Adding ftolpl parameter 1209: Author: N Brouard 1210: 1211: We had difficulties to get smoothed confidence intervals. It was due 1212: to the period prevalence which wasn't computed accurately. The inner 1213: parameter ftolpl is now an outer parameter of the .imach parameter 1214: file after estepm. If ftolpl is small 1.e-4 and estepm too, 1215: computation are long. 1216: 1217: * version.h: *** empty log message *** 1218: 1219: * imach.c: Summary: temporary 1220: 1221: 2015-10-27 Brouard Nicolas <brouard@ined.fr> 1222: 1223: * Makefile: *** empty log message *** 1224: 1225: * setup.iss.in: Summary: some new subdirs 1226: 1227: * imach.c: *** empty log message *** 1228: 1229: * setup.iss.in: 1230: Summary: Adding gnuplot 5.1 with bin and etc for pango fonts 1231: 1232: 2015-10-24 Brouard Nicolas <brouard@ined.fr> 1233: 1234: * imach.c: *** empty log message *** 1235: 1236: 2015-10-23 Brouard Nicolas <brouard@ined.fr> 1237: 1238: * imach.c: 1239: Summary: 0.98r3 some clarification for graphs on likelihood contributions 1240: 1241: 2015-10-01 Brouard Nicolas <brouard@ined.fr> 1242: 1243: * imach.c: Summary: Some new graphs of contribution to likelihood 1244: 1245: 2015-09-30 Brouard Nicolas <brouard@ined.fr> 1246: 1247: * Makefile: 1248: Summary: Version read from version.h, itself read with cmake 1249: 1250: * imach.c: Summary: looking at better estimation of the hessian 1251: 1252: Also a better criteria for convergence to the period prevalence And 1253: therefore adding the number of years needed to converge. (The 1254: prevalence in any alive state shold sum to one 1255: 1256: * Makefile: *** empty log message *** 1257: 1258: 2015-09-22 Brouard Nicolas <brouard@ined.fr> 1259: 1260: * imach.c: 1261: Summary: Adding some overall graph on contribution to likelihood. Might change 1262: 1263: 2015-09-15 Brouard Nicolas <brouard@ined.fr> 1264: 1265: * Makefile: *** empty log message *** 1266: 1267: * imach.c: Summary: 0.98r0 1268: 1269: - Some new graphs like survival functions 1270: - Some bugs fixed like model=1+age+V2. 1271: 1272: * ChangeLog, Makefile: Summary: 0.98r0 1273: 1274: * CMakeLists.txt: Summary: Some tests for setup.iss 1275: 1276: 2015-09-15 Nicolas J Brouard <nbrouard@tugault.ined.fr> 1277: 1278: * (Module): 0.98r0 Some new graphs, some bugs fixed: model=1+age+V2. etc 1279: 1280: 2015-08-18 Nicolas J Brouard <nbrouard@tugault.local> 1281: 1282: * (Module): Adding error when the covariance matrix doesn't contain the exact number of lines required by the model line. 1283: 1284: 2015-08-03 Nicolas J Brouard <nbrouard@tugault.ined.fr> 1285: 1286: * (Module): Changing Variance of one-step probabilities into Standard deviation (thanks to Yao-Chi Shih) 1287: 1288: 2014-12-16 Nicolas J Brouard <nbrouard@tugault.local> 1289: 1290: * imach.c (Module): Merging 1.61 to 1.162 1291: 1292: 2014-09-02 Nicolas J Brouard <nbrouard@njbrouard.ined.fr> 1293: 1294: * prevlim.h (Module): 1295: 1296: 2014-01-26 Nicolas J Brouard <nbrouard@ledermann-2.local> 1297: 1298: * imach.c (Module): Trying to merge old staffs together while being at Tokyo. Not tested... 1299: (Module): Version 0.98nR Running ok, but output format still only works for three covariates. 1300: 1301: 2010-04-29 brouard <brouard@ined.fr> 1302: 1303: * imach.c (Module): Checking covariates for more complex models 1304: than V1+V2. A lot of change to be done. Unstable. 1305: 1306: 2010-04-26 brouard <brouard@ined.fr> 1307: 1308: * imach.c (Module): merging some libgsl code. Fixing computation 1309: of likelione (using inter/intrapolation if mle = 0) in order to 1310: get same likelihood as if mle=1. 1311: Some cleaning of code and comments added. 1312: 1313: 2009-10-29 brouard <brouard@ined.fr> 1314: 1315: * imach.c (Module): Now imach stops if date of birth, at least year of birth, is not given. Some cleaning of the code. 1316: 1317: 2006-06-30 Brouard Nicolas <brouard@localhost> 1318: 1319: * imach.c (Module): Clarifications on computing e.j 1320: 1321: 2006-04-28 Brouard Nicolas <brouard@localhost> 1322: 1323: * imach.c (Module): Yes the sum of survivors was wrong since 1324: imach-114 because nhstepm was no more computed in the age 1325: loop. Now we define nhstepma in the age loop. 1326: (Module): In order to speed up (in case of numerous covariates) we 1327: compute health expectancies (without variances) in a first step 1328: and then all the health expectancies with variances or standard 1329: deviation (needs data from the Hessian matrices) which slows the 1330: computation. 1331: In the future we should be able to stop the program is only health 1332: expectancies and graph are needed without standard deviations. 1333: 1334: 2006-04-04 LIEVRE Agnes <lievre@PC_00697.ined.fr> 1335: 1336: * imach.iss (Module): 1337: 1338: * imach.c (Module): bugs in evsi + population file added in htm output file 1339: 1340: 2006-03-22 LIEVRE Agnes <lievre@PC_00697.ined.fr> 1341: 1342: * imach.c (Module): 1343: 1344: 2006-03-20 Brouard Nicolas <brouard@localhost> 1345: 1346: * imach.c (Module): <title> changed, corresponds to .htm file 1347: name. <head> headers where missing. 1348: 1349: * Makefile (Module): 0.98g 1350: 1351: * imach.c (Module): Weights can have a decimal point as for 1352: English (a comma might work with a correct LC_NUMERIC environment, 1353: otherwise the weight is truncated). 1354: Modification of warning when the covariates values are not 0 or 1355: 1. 1356: Version 0.98g 1357: 1358: 2006-03-16 LIEVRE Agnes <lievre@PC_00697.ined.fr> 1359: 1360: * imach.c (Module): Comments concerning covariates added 1361: 1362: * noreg-setup.iss (Module): Version 0.98f 1363: 1364: * Makefile (Module): Version 0.98f make _windows=1 imachcyg-setup 1365: 1366: * imach.c (Module): refinements in the computation of lli if 1367: status=-2 in order to have more reliable computation if stepm is 1368: not 1 month. Version 0.98f 1369: 1370: 2006-03-15 Brouard Nicolas <brouard@localhost> 1371: 1372: * imach.c (Module): Bug if status = -2, the loglikelihood was 1373: computed as likelihood omitting the logarithm. Version O.98e 1374: 1375: 2006-03-14 Brouard Nicolas <brouard@localhost> 1376: 1377: * imach.c (Module): varevsij Comments added explaining the second 1378: table of variances if popbased=1 . 1379: (Module): Covariances of eij, ekl added, graphs fixed, new html link. 1380: (Module): Function pstamp added 1381: (Module): Version 0.98d 1382: 1383: 2006-03-06 Brouard Nicolas <brouard@localhost> 1384: 1385: * imach.c (Module): Variance-covariance wrong links and 1386: varian-covariance of ej. is needed (Saito). 1387: 1388: 2006-02-27 Brouard Nicolas <brouard@localhost> 1389: 1390: * imach.c (Module): One freematrix added in mlikeli! 1391: 1392: 2006-02-26 Brouard Nicolas <brouard@localhost> 1393: 1394: * imach.c (Module): Some improvements in processing parameter 1395: filename with strsep. 1396: 1397: 2006-02-24 Brouard Nicolas <brouard@localhost> 1398: 1399: * imach.c (Module): Memory leaks checks with valgrind and: 1400: datafile was not closed, some imatrix were not freed and on matrix 1401: allocation too. 1402: 1403: 2006-02-09 LIEVRE Agnes <lievre@PC_00697.ined.fr> 1404: 1405: * strsep.c (Module): Inclusion because not in mingw32!!! 1406: 1407: * setup.iss.in (Module): IMACHSETUPVERSION added 1408: 1409: * noreg-setup.iss (Module): added to cvs tree 1410: 1411: * Makefile (Module): IMACHSETUPVERSION added 1412: 1413: 2006-01-30 BROUARD Nicolas <brouard@localhost> 1414: 1415: * Makefile (Module, Module): Latest Makefile 1416: 1417: * imach.c (Module): Back to gnuplot.exe instead of wgnuplot.exe 1418: 1419: 2006-01-27 BROUARD Nicolas <brouard@localhost> 1420: 1421: * noreg-setup.iss (Module): ADDED 1422: 1423: 2006-01-24 BROUARD Nicolas <brouard@localhost> 1424: 1425: * imach.c (Module): Comments (lines starting with a #) are allowed in data. 1426: 1427: 2005-10-25 Nicolas Brouard <brouard@dhcp171.recherche.ined.fr> 1428: 1429: * Makefile (Module): Added Imach98.pmsp and howto crete a dmg 1430: 1431: * createdmg.sh (Module): Creates a dmg disk for MaC OS/X 1432: 1433: 2005-09-30 <lievre@PC_00595> 1434: 1435: * imach.c (Module): sump fixed, loop imx fixed, and simplifications. 1436: (Module): If the status is missing at the last wave but we know 1437: that the person is alive, then we can code his/her status as -2 1438: (instead of missing=-1 in earlier versions) and his/her 1439: contributions to the likelihood is 1 - Prob of dying from last 1440: health status (= 1-p13= p11+p12 in the easiest case of somebody in 1441: the healthy state at last known wave). Version is 0.98 1442: 1443: 2004-05-20 Brouard Nicolas <brouard@localhost> 1444: 1445: * imach.c (Repository): 1446: Agnes added a direct estimation of mortality (without the need of 1447: computing period prevalence and differential mortality). Thus here 1448: is version 0.97a which has been distributed to some people at 1449: REVES 16 in Brugge using an Inno setup.exe for PCs. Estimates of 1450: mortality using covariates is not done today. Estimating direct 1451: mortality is a very different process because it doesn't need 1452: interpolation because it is easy to get the lx from the force of 1453: the mortality mux in the simplest case as for a Gompertz (log mux 1454: = a + b*x . But we have been able to incorporate the new code 1455: within former imach program (0.96d) without deteriorating too much 1456: the understanding of the program. 1457: 1458: 2003-06-25 <brouard@BROUARD> 1459: 1460: * imach.c (Module): On windows (cygwin) function asctime_r doesn't 1461: exist so I changed back to asctime which exists. 1462: (Module): Version 0.96b 1463: 1464: 2003-06-24 <brouard@BROUARD> 1465: 1466: * imach.c (Module): Some bugs corrected for windows. Also, when 1467: mle=-1 a template is output in file "or"mypar.txt with the design 1468: of the covariance matrix to be input. 1469: 1470: 2003-06-25 Brouard Nicolas <brouard@homebrou> 1471: 1472: * imach.c (Repository): Duplicated warning errors corrected. 1473: (Repository): Elapsed time after each iteration is now output. It 1474: helps to forecast when convergence will be reached. Elapsed time 1475: is stamped in powell. We created a new html file for the graphs 1476: concerning matrix of covariance. It has extension -cov.htm. 1477: 1478: 2003-06-23 Brouard Nicolas <brouard@homebrou> 1479: 1480: * 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. 1481: 1482: 2003-06-18 Brouard Nicolas <brouard@homebrou> 1483: 1484: * imach.c (Repository): Suppress abusive calls to append to file html (solution was to put fichtm as a global variable. Version 0.96 1485: 1486: 2003-06-17 Brouard Nicolas <brouard@homebrou> 1487: 1488: 1489: * imach.c (Repository): Check when date of death was earlier that 1490: current date of interview. It may happen when the death was just 1491: prior to the death. In this case, dh was negative and likelihood 1492: was wrong (infinity). We still send an "Error" but patch by 1493: assuming that the date of death was just one stepm after the 1494: interview. 1495: (Repository): Because some people have very long ID (first column) 1496: we changed int to long in num[] and we added a new lvector for 1497: memory allocation. But we also truncated to 8 characters (left 1498: truncation) 1499: 1500: (Repository): No more line truncation errors. 1501: 1502: 2003-06-14 Brouard Nicolas <brouard@homebrou> 1503: 1504: * imach.c (Repository): Add of a routine likelione (likelihood 1505: only once) which prints on a text file 1506: (ilk) the contributions to the likelihood for each 1507: individual/wave. 1508: 1509: 2003-06-13 Brouard Nicolas <brouard@homebrou> 1510: * imach.c (Repository): Replace "freqsummary" at a correct 1511: place. It differs from routine "prevalence" which may be called 1512: many times. Probs is memory consuming and must be used with 1513: parcimony. 1514: 1515: 2003-06-17 <brouard@BROUARD> 1516: 1517: * timeval.h (Module): Added included file to make use of 1518: gettimeofday working on win32 with cygwin. 1519: 1520: 2003-05-16 Brouard Nicolas <brouard@brouard> 1521: * imach.c (Module): Insert of a warning if the delay between two 1522: waves is negative. 1523: 1524: * imach.c (Module): 1525: 1526: 2003-05-03 <brouard@BROUARD> 1527: 1528: * imach.c: Some cleaning 1529: 1530: 2003-03-28 <brouard@BROUARD> 1531: * imach.c (Module): In version up to 0.92 likelihood was computed 1532: as if date of death was unknown. Death was treated as any other 1533: health state: the date of the interview describes the actual state 1534: and not the date of a change in health state. The former idea was 1535: to consider that at each interview the state was recorded 1536: (healthy, disable or death) and IMaCh was corrected; but when we 1537: introduced the exact date of death then we should have modified 1538: the contribution of an exact death to the likelihood. This new 1539: contribution is smaller and very dependent of the step unit 1540: stepm. It is no more the probability to die between last interview 1541: and month of death but the probability to survive from last 1542: interview up to one month before death multiplied by the 1543: probability to die within a month. Thanks to Chris 1544: Jackson for correcting this bug. Former versions increased 1545: mortality artificially. The bad side is that we add another loop 1546: which slows down the processing. The difference can be up to 10% 1547: lower mortality. 1548: 1549: 2002-05-30 <brouard@BROUARD> 1550: * 1551: * imach.c (Module): Add correlation matrix of one-step 1552: probabilities and covariance matrix 1553: 1554: 1555: 1556: 1557: