imach/src/ChangeLog - view

File: [Local Repository] / imach / src / ChangeLog
Revision 1.69: download - view: text, annotated - select for diffs
Tue Apr 30 10:58:21 2024 UTC (6 months ago) by brouard
Branches: MAIN
CVS tags: HEAD

Summary: version 0.99s4 and std of e.j/e..

1: 2024-04-30 Nicolas Brouard <brouard@ined.fr> 2: 3: * (Module): In version 0.99s4, we incorporated the calculation of 4: the std error of the proportion of total life expectancy spent in 5: a specific state Var(e.j/e..) using the formula of Var(X/Y) 6: depending only of the variances of X and Y and expectancies. 7: 8: 2024-04-24 Nicolas Brouard <brouard@ined.fr> 9: 10: * (Module): This version comes late after having tested 11: successfully the praxis C version of Buckhardt. But Buckardt's 12: version was difficult to read and Gegenfurtner's version had a few 13: typos which made its results less reliable than Buckhardt's 14: results. The most important work consisted in retyping the Brent 15: original PRAXIS program written in Algol W (published with errors, 16: ommitting the transposition of matrix V before its QR reduction 17: from Golub. I used the recent "awe" compiler from Gkynn Webster. 18: The awe library had errors, for example in arc tangent function 19: which have been fixed. 20: 21: The main objective was to get identical results with the three 22: versions: (1) Algol W, (2) Buckhardt'C version as well (3) 23: Gegenfürtner C versions on the various test functions published by 24: Brent in 1973 in Algol W. 25: 26: Also, in order to compare them, the random function had to produce 27: the same sequence for the 3 softwares. The random function used in 28: imach corresponds to original Brent's random function written in 29: Algol W. Other point, in Algol W, the arrays of dimension n are 30: 'normal' mathematical arrays starting from 1 to n. But this is a 31: real issue in C where, by default, arrays are starting from 0 to 32: n-1. In Buckhardt, as well as in Gegenfürtner C code, it can be 33: seen that authors while trying to mimick original Brent Algol W 34: code are hesitating by changing either a loop originally from 1 to 35: n in a loop from 0 to n-1, or keeping Brent's loop from 1 to n and 36: shifting the index from original X(I) in Algol W to x[i-1] in C. 37: But as IMaCh is using, since the beginning, the Numerical Recipes 38: functions vector or matrix, I changed Geggenfürtner code to mimick 39: the original Algol W arrays. Thus the X(I) is translated in C as 40: x[i] which minimizes the errors. The Golub QR algorithm was 41: published in Algol with overflow errors which were reproduced in 42: Brent's Algol W code. Buckhardt code fixed these errors which are 43: much more problematic in C than in Algol W. Thus Buckhardt code 44: seems very safe, but i haven't chosen it for IMaCh because the C 45: style is horrible and almost unreadable compared to Gegenfürtner 46: CO code which is very close to Brent's original. Also what makes 47: Buckhardt code more difficult to read is, instead of passing the 48: minimum of parameters in the functions calls, as it is in Algol 49: Brent's code or Gegenfürtner's code, the list of parameters is 50: high. For example, the flin function LONG REAL PROCEDURE FLIN 51: (LONG REAL VALUE L) has only one parameter in Algol W, the 52: Gegenfürtner flin function had two parameters: static double 53: flin(l, j) double l; { int i; double tflin[N];} but Buckhardt 54: function has 14 parameters which makes the code unreadable and 55: useless. Gegenfürtner used a lot of static variables or functions 56: which I tried to minimize. Also in Gegefürtner, array dimensions 57: were fixed to N. In my adaptation the flin is static double 58: flin(double l, int j) and the parameter used are global variables. 59: 60: 2023-06-14 Nicolas Brouard <brouard@ined.fr> 61: 62: * imach.c (Module): Testing if conjugate gradient could be quicker 63: when lot of variables POWELLORIGINCONJUGATE 64: 65: 2023-05-23 Nicolas Brouard <brouard@ined.fr> 66: 67: * imach.c (Module): Fixed PROB_r 68: 69: 2023-05-22 Nicolas Brouard <brouard@ined.fr> 70: 71: * imach.c (Module): In the ILK....txt file, the number of columns 72: before the covariates values is dependent of the number of states (16+nlstate): 0.99r46 73: 74: 2023-05-08 Nicolas Brouard <brouard@ined.fr> 75: 76: * (Module): Error V0 when result:. and model 1+age+V1: fixed 77: 78: 2023-04-29 Nicolas Brouard <brouard@ined.fr> 79: 80: * imach.c (Module): Inverting the model equation and fixingg bugs 81: in the drawings 82: "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" 83: 84: 2023-04-24 Nicolas Brouard <brouard@ined.fr> 85: 86: * imach.c (Module): some bugs in printinggnuplot 87: 88: 2022-09-19 Nicolas Brouard <brouard@ined.fr> 89: 90: * imach.c (Module): Version 0.99r42 needed a newer version of 91: Gnuplot. But newer version 0.99r43 should run with the Gnuplot 92: version 5.0 or 5.1 distributed with IMaCh. 93: 94: 2022-09-18 Nicolas Brouard <brouard@ined.fr> 95: 96: * imach.c (Module): Fixing when a state in the data is 0 or higher 97: than lstate+ndeath. Fixing the plot of contribution to Likelihood. 98: 99: 2022-09-16 Nicolas Brouard <brouard@ined.fr> 100: 101: * imach.c (Module): 0.99r41 Was an error when product of timevarying and fixed. Using FixedV[of name] now. Thank you Feinuo 102: 103: 2022-09-14 Nicolas Brouard <brouard@ined.fr> 104: 105: * imach.c (Module): Fixing names of variables in T_ (thanks to Feinuo) 106: 107: 2022-09-14 Nicolas Brouard <brouard@ined.fr> 108: 109: * imach.c (Module): Version 0.99r39 with colored dummy covariates 110: (fixed or time varying), using new last columns of 111: ILK_parameter.txt file. 112: 113: 2022-09-11 Nicolas Brouard <brouard@ined.fr> 114: 115: * imach.c (Module): Adding timevarying products of any kinds, 116: should work before shifting cotvar from ncovcol+nqv columns in 117: order to have a correspondance between the column of cotvar and 118: the id of column. 119: (Module): Some cleaning and adding covariates in ILK.txt 120: 121: 2022-09-09 Nicolas Brouard <brouard@ined.fr> 122: 123: * imach.c (Module): Many improvements for fixing products of fixed 124: timevarying as well as fixed * fixed, and test with quantitative 125: covariate. 126: 127: 2022-09-04 Nicolas Brouard <brouard@ined.fr> 128: 129: * imach.c (Module): Now the easy runs i.e. without result or 130: model=1+age only did not work. The defautl combination should be 1 131: and not 0 because everything hasn't been tranformed yet. 132: 133: 2022-09-02 Nicolas Brouard <brouard@ined.fr> 134: 135: * imach.c: Version 0.99r35 because it outputs same results with 136: 1+age+V1+V1*age for females and 1+age for females only 137: (education=1 noweight) 138: 139: 2022-08-31 Nicolas Brouard <brouard@ined.fr> 140: 141: * imach.c (Module): Some improvments in fichtm and many verifications 0.99r34 142: 143: 2022-08-21 Nicolas Brouard <brouard@ined.fr> 144: 145: * imach.c (Module): Version 0.99r33 A lot of changes in 146: reassigning covariates: my first idea was that people will always 147: use the first covariate V1 into the model but in fact they are 148: producing data with many covariates and can use an equation model 149: with some of the covariate; it means that in a model V2+V3 instead 150: of codtabm(k,Tvaraff[j]) which calculates for combination k, for 151: three covariates (V1, V2, V3) the value of Tvaraff[j], but in fact 152: the equation model is restricted to two variables only (V2, V3) 153: and the combination for V2 should be codtabm(k,1) instead of 154: (codtabm(k,2), and the code should be 155: codtabm(k,TnsdVar[Tvaraff[j]]. Many many changes have been 156: made. All of these should be simplified once a day like we did in 157: hpxij() for example by using precov[nres] which is computed in 158: decoderesult for each nres of each resultline. Loop should be done 159: on the equation model globally by distinguishing only product with 160: age (which are changing with age) and no more on type of 161: covariates, single dummies, single covariates. 162: 163: 2022-08-06 Nicolas Brouard <brouard@ined.fr> 164: 165: * imach.c (Module): Version of imach using partly decoderesult to rebuild xpxij function 166: 167: 2022-08-03 Nicolas Brouard <brouard@ined.fr> 168: 169: * imach.c (Module): Many errors in graphs fixed with Vn*age covariates. 170: 171: 2022-07-25 Brouard Nicolas <brouard@brouard.name> 172: 173: * imach.c (Module): Error cptcovn instead of nsd in bmij (was 174: coredumped, revealed by Feiuno, thank you. 175: 176: 2022-07-23 Nicolas Brouard <brouard@ined.fr> 177: 178: * r29 W and not sqrt(Wald) 179: 180: 2022-07-22 Nicolas Brouard <brouard@ined.fr> 181: 182: * imach.c (Module): Output of Wald test in the htm file and not only in the log. 183: 184: 2022-06-02 Brouard Nicolas <brouard@brouard.name> 185: 186: * imach.c (Module): Adding the Wald tests from the log to the main 187: htm for better display of the maximum likelihood estimators. 188: 189: 2022-05-30 Brouard Nicolas <brouard@brouard.name> 190: 191: * imach.c: With products of covariates (age or dummies or 192: quantitatives), estimates of parameters were good but when 193: estimating variances the positions of newly created covariates 194: were wrongly assigned and results were wrong. Thank to Feinua_Sun! 195: 196: 2022-05-24 Nicolas Brouard <brouard@ined.fr> 197: 198: * imach.c (Module): Some attempts to find a bug of wrong estimates 199: of confidencce intervals with product in the equation modelC 200: 201: 2022-05-15 Nicolas Brouard <brouard@ined.fr> 202: 203: * imach.c (Module): Some minor improvements 204: 205: 2022-04-13 Brouard Nicolas <brouard@brouard.name> 206: 207: * imach.c (Module): Adding link to text data files 208: 209: 2022-04-11 Nicolas Brouard <brouard@ined.fr> 210: 211: * imach.c (Module): Error in rewriting the 'r' file with yearsfproj or yearsbproj fixed 212: 213: 2022-04-05 Brouard Nicolas <brouard@ined.fr> 214: 215: * imach.c (Module): Fixed covariates (dummy or quantitative) 216: with missing values have never been allowed but are ERRORS and 217: program quits. Standard deviations of fixed covariates were 218: wrongly computed. Mean and standard deviations of time varying 219: covariates are still not computed. 220: 221: 2022-03-17 Brouard Nicolas <brouard@ined.fr> 222: 223: * 99r25 Improvements in fixing discrepancies between covariates 224: missing in result line but in model. 225: 226: *Wed Mar 31 2021 Nicolas Brouard <brouard@ined.fr> 227: * imach.c (Module): Still bugs in the result loop. Thank to Holly Benett 228: 229: *2021-03-08 Nicolas Brouard <brouard@ined.fr> 230: 231: * ./ (Module): Fixed bug on result: 232: 233: *2021-02-20 Nicolas Brouard <brouard@ined.fr> 234: 235: * imach.c (Module): Fix bug on quitting after result lines! 236: (Module): Version 0.99r21 237: 238: 2021-02-12 Nicolas Brouard <brouard@ined.fr> 239: 240: * imach.c (Module): The use of a Windows BOM (huge) file is now an error 241: 242: 243: 2021-02-11 Nicolas Brouard <brouard@ined.fr> 244: 245: * (Module): imach.c Someone entered 'results:' instead of 'result:'. Now it is an error which is printed. 246: 247: 2020-02-22 Brouard Nicolas <brouard@ined.fr> 248: 249: * (Module): imach.c Update mle=-3 (for computing Life expectancy 250: and life table from the data without any state) 251: 252: 2019-05-20 Brouard Nicolas <brouard@ined.fr> 253: 254: * version.h: Summary: version 0.99.r19 255: 256: * imach.c: Summary: Projection syntax simplified 257: We can now start projections, forward or backward, from the mean date 258: of inteviews up to or down to a number of years of projection: 259: prevforecast=1 yearsfproj=15.3 mobil_average=0 260: or 261: prevforecast=1 starting-proj-date=1/1/2007 final-proj-date=12/31/2017 mobil_average=0 262: or 263: prevbackcast=1 yearsbproj=12.3 mobil_average=1 264: or 265: prevbackcast=1 starting-back-date=1/10/1999 final-back-date=1/1/1985 mobil_average=1 266: 267: 2019-05-18 Brouard Nicolas <brouard@ined.fr> 268: 269: * imach.c: Summary: doxygen tex bug 270: 271: 2019-05-16 Brouard Nicolas <brouard@ined.fr> 272: 273: * imach.c: Summary: There was some wrong lines added 274: 275: 2019-05-09 Brouard Nicolas <brouard@ined.fr> 276: 277: * Makefile, imach.c: *** empty log message *** 278: 279: * ChangeLog: Summary: C= 280: 281: * imach.c: Summary: Some updates 282: 283: * imach.c: Summary: Before ncovmax 284: 285: * version.h: *** empty log message *** 286: 287: * imach.c: Summary: 0.99r18 unlimited number of individuals 288: 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. 289: 290: 2018-12-13 Brouard Nicolas <brouard@ined.fr> 291: 292: * imach.c: Summary: Bug for young ages (<-30) will be in r17 293: 294: 2018-12-05 Brouard Nicolas <brouard@ined.fr> 295: 296: * Makefile: Summary: 0.99r16 297: 298: 2018-05-02 Brouard Nicolas <brouard@ined.fr> 299: 300: * imach.c: Summary: Some bugs fixed 301: 302: 2018-05-02 Brouard Nicolas <brouard@ined.fr> 303: 304: * imach.c: Summary: Bug for young ages (<-30) will be in r17 305: 306: 2018-05-02 Brouard Nicolas <brouard@ined.fr> 307: 308: * imach.c: Summary: Some bugs fixed 309: 310: 2018-05-01 Brouard Nicolas <brouard@ined.fr> 311: 312: * imach.c: 313: Summary: Bug fixed by providing frequencies only for non missing covariates 314: 315: 2018-04-27 Brouard Nicolas <brouard@ined.fr> 316: 317: * imach.c: Summary: some minor bugs 318: 319: 2018-04-21 Brouard Nicolas <brouard@ined.fr> 320: 321: * README.txt: *** empty log message *** 322: 323: * imach.c: Summary: Some bugs fixed, valgrind tested 324: 325: 2018-04-20 Brouard Nicolas <brouard@ined.fr> 326: 327: * Makefile, README.txt, setup.iss: Summary: imach 0.99r16 328: 329: * imach.c: 330: Summary: Computing mean and stdeviation of fixed quantitative variables 331: 332: 2018-04-19 Brouard Nicolas <brouard@ined.fr> 333: 334: * imach.c: Summary: Some minor bugs fixed 335: 336: * version.h: *** empty log message *** 337: 338: 2018-02-27 Brouard Nicolas <brouard@ined.fr> 339: 340: * imach.c: *** empty log message *** 341: 342: * imach.c: Summary: Adding second argument for quitting 343: 344: 2018-02-21 Brouard Nicolas <brouard@ined.fr> 345: 346: * Makefile, README.txt, imach.c: Summary: 0.99r15 347: New Makefile with recent VirtualBox 5.26. Bug in sqrt negatve in imach.c 348: 349: 2017-07-20 Brouard Nicolas <brouard@ined.fr> 350: 351: * imach.c: Summary: temporary working 352: 353: 2017-07-19 Brouard Nicolas <brouard@ined.fr> 354: 355: * imach.c: Summary: Bug for mobil_average=0 and prevforecast fixed(?) 356: 357: 2017-07-17 Brouard Nicolas <brouard@ined.fr> 358: 359: * imach.c: Summary: BOM files can be read now 360: 361: 2017-06-30 Brouard Nicolas <brouard@ined.fr> 362: 363: * imach.c: Summary: Graphs improvements 364: 365: * imach.c: Summary: Saito's color 366: 367: 2017-06-29 Brouard Nicolas <brouard@ined.fr> 368: 369: * imach.c: Summary: Version 0.99r14 370: 371: 2017-06-27 Brouard Nicolas <brouard@ined.fr> 372: 373: * imach.c: Summary: More documentation on projections 374: 375: * imach.c: Summary: Color of backprojection changed from 6 to 5(yellow) 376: 377: * imach.c: Summary: Some bug with rint 378: 379: 2017-05-24 Brouard Nicolas <brouard@ined.fr> 380: 381: * imach.c: *** empty log message *** 382: 383: 2017-05-23 Brouard Nicolas <brouard@ined.fr> 384: 385: * imach.c: Summary: Code into subroutine, cleanings 386: 387: 2017-05-18 Brouard Nicolas <brouard@ined.fr> 388: 389: * imach.c: 390: Summary: backprojection and confidence intervals of backprevalence 391: 392: 2017-05-13 Brouard Nicolas <brouard@ined.fr> 393: 394: * imach.c: Summary: temporary save for backprojection 395: 396: * imach.c: Summary: Version 0.99r13 (improvements and bugs fixed) 397: 398: 2017-04-26 Brouard Nicolas <brouard@ined.fr> 399: 400: * imach.c: Summary: imach 0.99r13 Some bugs fixed 401: 402: * imach.c: Summary: Labels in graphs 403: 404: 2017-04-24 Brouard Nicolas <brouard@ined.fr> 405: 406: * imach.c: Summary: to save 407: 408: 2017-04-18 Brouard Nicolas <brouard@ined.fr> 409: 410: * imach.c: *** empty log message *** 411: 412: 2017-04-05 Brouard Nicolas <brouard@ined.fr> 413: 414: * imach.c: Summary: Bug in E_ as well as in T_ fixed nres-1 vs k1-1 415: 416: 2017-04-04 Brouard Nicolas <brouard@ined.fr> 417: 418: * imach.c: Summary: Gnuplot indexations fixed (humm) 419: 420: * imach.c: 421: Summary: Some errors to warnings only if date of death is unknown but status is death we could set to pi3 422: 423: 2017-04-03 Brouard Nicolas <brouard@ined.fr> 424: 425: * imach.c: Summary: Version 0.99r12 426: Some cleanings, conformed with updated documentation. 427: 428: 2017-03-30 Brouard Nicolas <brouard@ined.fr> 429: 430: * version.h: *** empty log message *** 431: 432: * version.h: Summary:0.99r11 433: 434: 2017-03-29 Brouard Nicolas <brouard@ined.fr> 435: 436: * imach.c: Summary: Temp 437: 438: 2017-03-27 Brouard Nicolas <brouard@ined.fr> 439: 440: * imach.c: Summary: Temporary 441: 442: 2017-03-08 Brouard Nicolas <brouard@ined.fr> 443: 444: * imach.c: Summary: IMaCh version 0.99r10 bugs in gnuplot fixed 445: 446: * imach.c: Summary: Fixing data parameter line 447: 448: 2016-12-15 Brouard Nicolas <brouard@ined.fr> 449: 450: * imach.c: Summary: 0.99 in progress 451: 452: 2016-09-15 Brouard Nicolas <brouard@ined.fr> 453: 454: * imach.c: *** empty log message *** 455: 456: * imach.c: Summary: not working 457: 458: 2016-09-08 Brouard Nicolas <brouard@ined.fr> 459: 460: * imach.c: Summary: continue 461: 462: 2016-09-07 Brouard Nicolas <brouard@ined.fr> 463: 464: * imach.c: Summary: Starting values from frequencies 465: 466: * imach.c: *** empty log message *** 467: 468: 2016-09-02 Brouard Nicolas <brouard@ined.fr> 469: 470: * imach.c: *** empty log message *** 471: 472: 2016-08-30 Brouard Nicolas <brouard@ined.fr> 473: 474: * imach.c: Summary: Fixing a lots 475: 476: 2016-08-29 Brouard Nicolas <brouard@ined.fr> 477: 478: * imach.c: Summary: gnuplot problem in Back projection to fix 479: 480: * imach.c: Summary: Better 481: 482: 2016-08-26 Brouard Nicolas <brouard@ined.fr> 483: 484: * imach.c: 485: Summary: Improvement in Powell output in order to copy and paste 486: 487: * imach.c: Summary: Starting tests of 0.99 488: 489: * imach.c: Summary: to valgrind 490: 491: 2016-08-25 Brouard Nicolas <brouard@ined.fr> 492: 493: * imach.c: *** empty log message *** 494: 495: 2016-08-23 Brouard Nicolas <brouard@ined.fr> 496: 497: * imach.c: *** empty log message *** 498: 499: * imach.c: Summary: not working 500: 501: 2016-08-22 Brouard Nicolas <brouard@ined.fr> 502: 503: * imach.c: Summary: not working 504: 505: * imach.c: Summary: Not working 506: 507: 2016-07-23 Brouard Nicolas <brouard@ined.fr> 508: 509: * imach.c: Summary: Completing for func too 510: 511: 2016-07-22 Brouard Nicolas <brouard@ined.fr> 512: 513: * imach.c: Summary: Fixing some arrays, still debugging 514: 515: 2016-07-21 Brouard Nicolas <brouard@ined.fr> 516: 517: * imach.c: 518: Summary: 0.99 working (more or less) for Asian Workshop on multitate methods 519: 520: 2016-07-12 Brouard Nicolas <brouard@ined.fr> 521: 522: * imach.c: Summary: temp 523: 524: * imach.c: Summary: saving but not running 525: 526: 2016-07-01 Brouard Nicolas <brouard@ined.fr> 527: 528: * imach.c: Summary: Fixes 529: 530: 2016-02-19 Brouard Nicolas <brouard@ined.fr> 531: 532: * imach-0.98r.c: Summary: Kind of 0.98r? series, starting with r7 533: 534: * setup.iss, Makefile, version.h: *** empty log message *** 535: 536: * imach.c: Summary: temporary 537: 538: 2016-02-17 Brouard Nicolas <brouard@ined.fr> 539: 540: * imach.c: Summary: Probably last 0.98 stable version 0.98r6 541: 542: 2016-02-16 Brouard Nicolas <brouard@ined.fr> 543: 544: * imach.c: Summary: minor bug 545: 546: * imach.c: Summary: 0.99r2 547: 548: 2016-02-15 Brouard Nicolas <brouard@ined.fr> 549: 550: * imach.c: *** empty log message *** 551: 552: 2016-02-12 Brouard Nicolas <brouard@ined.fr> 553: 554: * imach.c: Summary: 0.99 Back projections 555: 556: 2015-12-23 Brouard Nicolas <brouard@ined.fr> 557: 558: * imach.c: Summary: Experimental backcast 559: 560: 2015-12-18 Brouard Nicolas <brouard@ined.fr> 561: 562: * imach.c: Summary: 0.98r4 Warning and status=-2 563: 564: Version 0.98r4 is now: 565: - displaying an error when status is -1, date of interview unknown and date of death known; 566: - permitting a status -2 when the vital status is unknown at a known date of right truncation. 567: Older changes concerning s=-2, dating from 2005 have been supersed. 568: 569: * Makefile: *** empty log message *** 570: 571: 2015-12-16 Brouard Nicolas <brouard@ined.fr> 572: 573: * Makefile: *** empty log message *** 574: 575: * imach.c: Summary: 0.98r4 working 576: 577: * Makefile: Summary: linuxrpm added 578: 579: * Makefile: *** empty log message *** 580: 581: * imach.c: Summary: temporary not working 582: 583: 2015-12-11 Brouard Nicolas <brouard@ined.fr> 584: 585: * imach.c: Summary: 0.98r4 586: 587: 2015-12-08 Brouard Nicolas <brouard@ined.fr> 588: 589: * Makefile: *** empty log message *** 590: 591: 2015-11-21 Brouard Nicolas <brouard@ined.fr> 592: 593: * Makefile: 594: Summary: VBoxManage change from execute --image to run --exe and --wait-exit doesn't exist any more 595: 596: * imach.c: Summary: minor typo 597: 598: * imach.c: Summary: 0.98r3 with some graph of projected cross-sectional 599: 600: Author: Nicolas Brouard 601: 602: 2015-11-18 Brouard Nicolas <brouard@ined.fr> 603: 604: * imach.c: Summary: Start working on projected prevalences 605: 606: 2015-11-17 Brouard Nicolas <brouard@ined.fr> 607: 608: * imach.c: Summary: Adding ftolpl parameter 609: Author: N Brouard 610: 611: We had difficulties to get smoothed confidence intervals. It was due 612: to the period prevalence which wasn't computed accurately. The inner 613: parameter ftolpl is now an outer parameter of the .imach parameter 614: file after estepm. If ftolpl is small 1.e-4 and estepm too, 615: computation are long. 616: 617: * version.h: *** empty log message *** 618: 619: * imach.c: Summary: temporary 620: 621: 2015-10-27 Brouard Nicolas <brouard@ined.fr> 622: 623: * Makefile: *** empty log message *** 624: 625: * setup.iss.in: Summary: some new subdirs 626: 627: * imach.c: *** empty log message *** 628: 629: * setup.iss.in: 630: Summary: Adding gnuplot 5.1 with bin and etc for pango fonts 631: 632: 2015-10-24 Brouard Nicolas <brouard@ined.fr> 633: 634: * imach.c: *** empty log message *** 635: 636: 2015-10-23 Brouard Nicolas <brouard@ined.fr> 637: 638: * imach.c: 639: Summary: 0.98r3 some clarification for graphs on likelihood contributions 640: 641: 2015-10-01 Brouard Nicolas <brouard@ined.fr> 642: 643: * imach.c: Summary: Some new graphs of contribution to likelihood 644: 645: 2015-09-30 Brouard Nicolas <brouard@ined.fr> 646: 647: * Makefile: 648: Summary: Version read from version.h, itself read with cmake 649: 650: * imach.c: Summary: looking at better estimation of the hessian 651: 652: Also a better criteria for convergence to the period prevalence And 653: therefore adding the number of years needed to converge. (The 654: prevalence in any alive state shold sum to one 655: 656: * Makefile: *** empty log message *** 657: 658: 2015-09-22 Brouard Nicolas <brouard@ined.fr> 659: 660: * imach.c: 661: Summary: Adding some overall graph on contribution to likelihood. Might change 662: 663: 2015-09-15 Brouard Nicolas <brouard@ined.fr> 664: 665: * Makefile: *** empty log message *** 666: 667: * imach.c: Summary: 0.98r0 668: 669: - Some new graphs like survival functions 670: - Some bugs fixed like model=1+age+V2. 671: 672: * ChangeLog, Makefile: Summary: 0.98r0 673: 674: * CMakeLists.txt: Summary: Some tests for setup.iss 675: 676: 2015-09-15 Nicolas J Brouard <nbrouard@tugault.ined.fr> 677: 678: * (Module): 0.98r0 Some new graphs, some bugs fixed: model=1+age+V2. etc 679: 680: 2015-08-18 Nicolas J Brouard <nbrouard@tugault.local> 681: 682: * (Module): Adding error when the covariance matrix doesn't contain the exact number of lines required by the model line. 683: 684: 2015-08-03 Nicolas J Brouard <nbrouard@tugault.ined.fr> 685: 686: * (Module): Changing Variance of one-step probabilities into Standard deviation (thanks to Yao-Chi Shih) 687: 688: 2014-12-16 Nicolas J Brouard <nbrouard@tugault.local> 689: 690: * imach.c (Module): Merging 1.61 to 1.162 691: 692: 2014-09-02 Nicolas J Brouard <nbrouard@njbrouard.ined.fr> 693: 694: * prevlim.h (Module): 695: 696: 2014-01-26 Nicolas J Brouard <nbrouard@ledermann-2.local> 697: 698: * imach.c (Module): Trying to merge old staffs together while being at Tokyo. Not tested... 699: (Module): Version 0.98nR Running ok, but output format still only works for three covariates. 700: 701: 2010-04-29 brouard <brouard@ined.fr> 702: 703: * imach.c (Module): Checking covariates for more complex models 704: than V1+V2. A lot of change to be done. Unstable. 705: 706: 2010-04-26 brouard <brouard@ined.fr> 707: 708: * imach.c (Module): merging some libgsl code. Fixing computation 709: of likelione (using inter/intrapolation if mle = 0) in order to 710: get same likelihood as if mle=1. 711: Some cleaning of code and comments added. 712: 713: 2009-10-29 brouard <brouard@ined.fr> 714: 715: * imach.c (Module): Now imach stops if date of birth, at least year of birth, is not given. Some cleaning of the code. 716: 717: 2006-06-30 Brouard Nicolas <brouard@localhost> 718: 719: * imach.c (Module): Clarifications on computing e.j 720: 721: 2006-04-28 Brouard Nicolas <brouard@localhost> 722: 723: * imach.c (Module): Yes the sum of survivors was wrong since 724: imach-114 because nhstepm was no more computed in the age 725: loop. Now we define nhstepma in the age loop. 726: (Module): In order to speed up (in case of numerous covariates) we 727: compute health expectancies (without variances) in a first step 728: and then all the health expectancies with variances or standard 729: deviation (needs data from the Hessian matrices) which slows the 730: computation. 731: In the future we should be able to stop the program is only health 732: expectancies and graph are needed without standard deviations. 733: 734: 2006-04-04 LIEVRE Agnes <lievre@PC_00697.ined.fr> 735: 736: * imach.iss (Module): 737: 738: * imach.c (Module): bugs in evsi + population file added in htm output file 739: 740: 2006-03-22 LIEVRE Agnes <lievre@PC_00697.ined.fr> 741: 742: * imach.c (Module): 743: 744: 2006-03-20 Brouard Nicolas <brouard@localhost> 745: 746: * imach.c (Module): <title> changed, corresponds to .htm file 747: name. <head> headers where missing. 748: 749: * Makefile (Module): 0.98g 750: 751: * imach.c (Module): Weights can have a decimal point as for 752: English (a comma might work with a correct LC_NUMERIC environment, 753: otherwise the weight is truncated). 754: Modification of warning when the covariates values are not 0 or 755: 1. 756: Version 0.98g 757: 758: 2006-03-16 LIEVRE Agnes <lievre@PC_00697.ined.fr> 759: 760: * imach.c (Module): Comments concerning covariates added 761: 762: * noreg-setup.iss (Module): Version 0.98f 763: 764: * Makefile (Module): Version 0.98f make _windows=1 imachcyg-setup 765: 766: * imach.c (Module): refinements in the computation of lli if 767: status=-2 in order to have more reliable computation if stepm is 768: not 1 month. Version 0.98f 769: 770: 2006-03-15 Brouard Nicolas <brouard@localhost> 771: 772: * imach.c (Module): Bug if status = -2, the loglikelihood was 773: computed as likelihood omitting the logarithm. Version O.98e 774: 775: 2006-03-14 Brouard Nicolas <brouard@localhost> 776: 777: * imach.c (Module): varevsij Comments added explaining the second 778: table of variances if popbased=1 . 779: (Module): Covariances of eij, ekl added, graphs fixed, new html link. 780: (Module): Function pstamp added 781: (Module): Version 0.98d 782: 783: 2006-03-06 Brouard Nicolas <brouard@localhost> 784: 785: * imach.c (Module): Variance-covariance wrong links and 786: varian-covariance of ej. is needed (Saito). 787: 788: 2006-02-27 Brouard Nicolas <brouard@localhost> 789: 790: * imach.c (Module): One freematrix added in mlikeli! 791: 792: 2006-02-26 Brouard Nicolas <brouard@localhost> 793: 794: * imach.c (Module): Some improvements in processing parameter 795: filename with strsep. 796: 797: 2006-02-24 Brouard Nicolas <brouard@localhost> 798: 799: * imach.c (Module): Memory leaks checks with valgrind and: 800: datafile was not closed, some imatrix were not freed and on matrix 801: allocation too. 802: 803: 2006-02-09 LIEVRE Agnes <lievre@PC_00697.ined.fr> 804: 805: * strsep.c (Module): Inclusion because not in mingw32!!! 806: 807: * setup.iss.in (Module): IMACHSETUPVERSION added 808: 809: * noreg-setup.iss (Module): added to cvs tree 810: 811: * Makefile (Module): IMACHSETUPVERSION added 812: 813: 2006-01-30 BROUARD Nicolas <brouard@localhost> 814: 815: * Makefile (Module, Module): Latest Makefile 816: 817: * imach.c (Module): Back to gnuplot.exe instead of wgnuplot.exe 818: 819: 2006-01-27 BROUARD Nicolas <brouard@localhost> 820: 821: * noreg-setup.iss (Module): ADDED 822: 823: 2006-01-24 BROUARD Nicolas <brouard@localhost> 824: 825: * imach.c (Module): Comments (lines starting with a #) are allowed in data. 826: 827: 2005-10-25 Nicolas Brouard <brouard@dhcp171.recherche.ined.fr> 828: 829: * Makefile (Module): Added Imach98.pmsp and howto crete a dmg 830: 831: * createdmg.sh (Module): Creates a dmg disk for MaC OS/X 832: 833: 2005-09-30 <lievre@PC_00595> 834: 835: * imach.c (Module): sump fixed, loop imx fixed, and simplifications. 836: (Module): If the status is missing at the last wave but we know 837: that the person is alive, then we can code his/her status as -2 838: (instead of missing=-1 in earlier versions) and his/her 839: contributions to the likelihood is 1 - Prob of dying from last 840: health status (= 1-p13= p11+p12 in the easiest case of somebody in 841: the healthy state at last known wave). Version is 0.98 842: 843: 2004-05-20 Brouard Nicolas <brouard@localhost> 844: 845: * imach.c (Repository): 846: Agnes added a direct estimation of mortality (without the need of 847: computing period prevalence and differential mortality). Thus here 848: is version 0.97a which has been distributed to some people at 849: REVES 16 in Brugge using an Inno setup.exe for PCs. Estimates of 850: mortality using covariates is not done today. Estimating direct 851: mortality is a very different process because it doesn't need 852: interpolation because it is easy to get the lx from the force of 853: the mortality mux in the simplest case as for a Gompertz (log mux 854: = a + b*x . But we have been able to incorporate the new code 855: within former imach program (0.96d) without deteriorating too much 856: the understanding of the program. 857: 858: 2003-06-25 <brouard@BROUARD> 859: 860: * imach.c (Module): On windows (cygwin) function asctime_r doesn't 861: exist so I changed back to asctime which exists. 862: (Module): Version 0.96b 863: 864: 2003-06-24 <brouard@BROUARD> 865: 866: * imach.c (Module): Some bugs corrected for windows. Also, when 867: mle=-1 a template is output in file "or"mypar.txt with the design 868: of the covariance matrix to be input. 869: 870: 2003-06-25 Brouard Nicolas <brouard@homebrou> 871: 872: * imach.c (Repository): Duplicated warning errors corrected. 873: (Repository): Elapsed time after each iteration is now output. It 874: helps to forecast when convergence will be reached. Elapsed time 875: is stamped in powell. We created a new html file for the graphs 876: concerning matrix of covariance. It has extension -cov.htm. 877: 878: 2003-06-23 Brouard Nicolas <brouard@homebrou> 879: 880: * 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. 881: 882: 2003-06-18 Brouard Nicolas <brouard@homebrou> 883: 884: * imach.c (Repository): Suppress abusive calls to append to file html (solution was to put fichtm as a global variable. Version 0.96 885: 886: 2003-06-17 Brouard Nicolas <brouard@homebrou> 887: 888: 889: * imach.c (Repository): Check when date of death was earlier that 890: current date of interview. It may happen when the death was just 891: prior to the death. In this case, dh was negative and likelihood 892: was wrong (infinity). We still send an "Error" but patch by 893: assuming that the date of death was just one stepm after the 894: interview. 895: (Repository): Because some people have very long ID (first column) 896: we changed int to long in num[] and we added a new lvector for 897: memory allocation. But we also truncated to 8 characters (left 898: truncation) 899: 900: (Repository): No more line truncation errors. 901: 902: 2003-06-14 Brouard Nicolas <brouard@homebrou> 903: 904: * imach.c (Repository): Add of a routine likelione (likelihood 905: only once) which prints on a text file 906: (ilk) the contributions to the likelihood for each 907: individual/wave. 908: 909: 2003-06-13 Brouard Nicolas <brouard@homebrou> 910: * imach.c (Repository): Replace "freqsummary" at a correct 911: place. It differs from routine "prevalence" which may be called 912: many times. Probs is memory consuming and must be used with 913: parcimony. 914: 915: 2003-06-17 <brouard@BROUARD> 916: 917: * timeval.h (Module): Added included file to make use of 918: gettimeofday working on win32 with cygwin. 919: 920: 2003-05-16 Brouard Nicolas <brouard@brouard> 921: * imach.c (Module): Insert of a warning if the delay between two 922: waves is negative. 923: 924: * imach.c (Module): 925: 926: 2003-05-03 <brouard@BROUARD> 927: 928: * imach.c: Some cleaning 929: 930: 2003-03-28 <brouard@BROUARD> 931: * imach.c (Module): In version up to 0.92 likelihood was computed 932: as if date of death was unknown. Death was treated as any other 933: health state: the date of the interview describes the actual state 934: and not the date of a change in health state. The former idea was 935: to consider that at each interview the state was recorded 936: (healthy, disable or death) and IMaCh was corrected; but when we 937: introduced the exact date of death then we should have modified 938: the contribution of an exact death to the likelihood. This new 939: contribution is smaller and very dependent of the step unit 940: stepm. It is no more the probability to die between last interview 941: and month of death but the probability to survive from last 942: interview up to one month before death multiplied by the 943: probability to die within a month. Thanks to Chris 944: Jackson for correcting this bug. Former versions increased 945: mortality artificially. The bad side is that we add another loop 946: which slows down the processing. The difference can be up to 10% 947: lower mortality. 948: 949: 2002-05-30 <brouard@BROUARD> 950: 951: * imach.c (Module): Add correlation matrix of one-step 952: probabilities and covariance matrix 953: