Return to ChangeLog CVS log

Up to [Local Repository] / imach / src

*** empty log message ***

    1: 2006-03-16  LIEVRE Agnes  <lievre@PC_00697.ined.fr>
    2: 
    3: 	* imach.c (Module): Comments concerning covariates added
    4: 
    5: 	* noreg-setup.iss (Module): Version 0.98f
    6: 
    7: 	* Makefile (Module): Version 0.98f make _windows=1 imachcyg-setup
    8: 
    9: 	* imach.c (Module): refinements in the computation of lli if
   10: 	status=-2 in order to have more reliable computation if stepm is
   11: 	not 1 month. Version 0.98f
   12: 
   13: 2006-03-15  Brouard Nicolas  <brouard@localhost>
   14: 
   15: 	* imach.c (Module): Bug if status = -2, the loglikelihood was
   16: 	computed as likelihood omitting the logarithm. Version O.98e
   17: 
   18: 2006-03-14  Brouard Nicolas  <brouard@localhost>
   19: 
   20: 	* imach.c (Module): varevsij Comments added explaining the second
   21: 	table of variances if popbased=1 .
   22: 	(Module): Covariances of eij, ekl added, graphs fixed, new html link.
   23: 	(Module): Function pstamp added
   24: 	(Module): Version 0.98d 
   25: 
   26: 2006-03-06  Brouard Nicolas  <brouard@localhost>
   27: 
   28: 	* imach.c (Module): Variance-covariance wrong links and
   29: 	varian-covariance of ej. is needed (Saito).
   30: 
   31: 2006-02-27  Brouard Nicolas  <brouard@localhost>
   32: 
   33: 	* imach.c (Module): One freematrix added in mlikeli!
   34: 
   35: 2006-02-26  Brouard Nicolas  <brouard@localhost>
   36: 
   37: 	* imach.c (Module): Some improvements in processing parameter
   38: 	filename with strsep.
   39: 
   40: 2006-02-24  Brouard Nicolas  <brouard@localhost>
   41: 
   42: 	* imach.c (Module): Memory leaks checks with valgrind and:
   43: 	datafile was not closed, some imatrix were not freed and on matrix
   44: 	allocation too.
   45: 
   46: 2006-02-09  LIEVRE Agnes  <lievre@PC_00697.ined.fr>
   47: 
   48: 	* strsep.c (Module): Inclusion because not in mingw32!!!
   49: 
   50: 	* setup.iss.in (Module): IMACHSETUPVERSION added
   51: 
   52: 	* noreg-setup.iss (Module): added to cvs tree
   53: 
   54: 	* Makefile (Module): IMACHSETUPVERSION added
   55: 
   56: 2006-01-30  BROUARD Nicolas  <brouard@localhost>
   57: 
   58: 	* Makefile (Module, Module): Latest Makefile
   59: 
   60: 	* imach.c (Module): Back to gnuplot.exe instead of wgnuplot.exe
   61: 
   62: 2006-01-27  BROUARD Nicolas  <brouard@localhost>
   63: 
   64: 	* noreg-setup.iss (Module): ADDED
   65: 
   66: 2006-01-24  BROUARD Nicolas  <brouard@localhost>
   67: 
   68: 	* imach.c (Module): Comments (lines starting with a #) are allowed in data.
   69: 
   70: 2005-10-25  Nicolas Brouard  <brouard@dhcp171.recherche.ined.fr>
   71: 
   72: 	* Makefile (Module): Added Imach98.pmsp and howto crete a dmg
   73: 
   74: 	* createdmg.sh (Module): Creates a dmg disk for MaC OS/X
   75: 
   76: 2005-09-30    <lievre@PC_00595>
   77: 
   78: 	* imach.c (Module): sump fixed, loop imx fixed, and simplifications.
   79: 	(Module): If the status is missing at the last wave but we know
   80: 	that the person is alive, then we can code his/her status as -2
   81: 	(instead of missing=-1 in earlier versions) and his/her
   82: 	contributions to the likelihood is 1 - Prob of dying from last
   83: 	health status (= 1-p13= p11+p12 in the easiest case of somebody in
   84: 	the healthy state at last known wave). Version is 0.98 
   85: 
   86: 2004-05-20  Brouard Nicolas  <brouard@localhost>
   87: 
   88: 	* imach.c (Repository): 
   89: 	Agnes added a direct estimation of mortality (without the need of
   90: 	computing period prevalence and differential mortality). Thus here
   91: 	is version 0.97a which has been distributed to some people at
   92: 	REVES 16 in Brugge using an Inno setup.exe for PCs. Estimates of
   93: 	mortality using covariates is not done today. Estimating direct
   94: 	mortality is a very different process because it doesn't need
   95: 	interpolation because it is easy to get the lx from the force of
   96: 	the mortality mux in the simplest case as for a Gompertz (log mux
   97: 	= a + b*x . But we have been able to incorporate the new code
   98: 	within former imach program (0.96d) without deteriorating too much
   99: 	the understanding of the program. 
  100: 
  101: 2003-06-25    <brouard@BROUARD>
  102: 
  103: 	* imach.c (Module): On windows (cygwin) function asctime_r doesn't
  104: 	exist so I changed back to asctime which exists.
  105: 	(Module): Version 0.96b
  106: 
  107: 2003-06-24    <brouard@BROUARD>
  108: 
  109: 	* imach.c (Module): Some bugs corrected for windows. Also, when
  110: 	mle=-1 a template is output in file "or"mypar.txt with the design
  111: 	of the covariance matrix to be input.
  112: 
  113: 2003-06-25  Brouard Nicolas  <brouard@homebrou>
  114: 
  115: 	* imach.c (Repository): Duplicated warning errors corrected.
  116: 	(Repository): Elapsed time after each iteration is now output. It
  117: 	helps to forecast when convergence will be reached. Elapsed time
  118: 	is stamped in powell.  We created a new html file for the graphs
  119: 	concerning matrix of covariance. It has extension -cov.htm.
  120: 
  121: 2003-06-23  Brouard Nicolas  <brouard@homebrou>
  122: 
  123: 	* 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.
  124: 
  125: 2003-06-18  Brouard Nicolas  <brouard@homebrou>
  126: 
  127: 	* imach.c (Repository): Suppress abusive calls to append to file html (solution was to put fichtm as a global variable. Version 0.96
  128: 
  129: 2003-06-17  Brouard Nicolas  <brouard@homebrou>
  130: 
  131: 
  132: 	* imach.c (Repository): Check when date of death was earlier that
  133: 	current date of interview. It may happen when the death was just
  134: 	prior to the death. In this case, dh was negative and likelihood
  135: 	was wrong (infinity). We still send an "Error" but patch by
  136: 	assuming that the date of death was just one stepm after the
  137: 	interview.
  138: 	(Repository): Because some people have very long ID (first column)
  139: 	we changed int to long in num[] and we added a new lvector for
  140: 	memory allocation. But we also truncated to 8 characters (left
  141: 	truncation)
  142: 
  143: 	(Repository): No more line truncation errors.
  144: 
  145: 2003-06-14  Brouard Nicolas  <brouard@homebrou>
  146: 
  147: 	* imach.c (Repository): Add of a routine likelione (likelihood
  148: 	only once) which prints on a text file
  149: 	(ilk) the contributions to the likelihood for each
  150: 	individual/wave.
  151: 
  152: 2003-06-13  Brouard Nicolas  <brouard@homebrou>
  153: 	* imach.c (Repository): Replace "freqsummary" at a correct
  154: 	place. It differs from routine "prevalence" which may be called
  155: 	many times. Probs is memory consuming and must be used with
  156: 	parcimony.
  157: 
  158: 2003-06-17    <brouard@BROUARD>
  159: 
  160: 	* timeval.h (Module): Added included file to make use of
  161: 	gettimeofday working on win32 with cygwin.
  162: 
  163: 2003-05-16  Brouard Nicolas  <brouard@brouard>
  164: 	* imach.c (Module): Insert of a warning if the delay between two
  165: 	waves is negative.
  166: 
  167: 	* imach.c (Module): 
  168: 
  169: 2003-05-03    <brouard@BROUARD>
  170: 
  171: 	* imach.c: Some cleaning
  172: 
  173: 2003-03-28    <brouard@BROUARD>
  174: 	* imach.c (Module): In version up to 0.92 likelihood was computed
  175: 	as if date of death was unknown. Death was treated as any other
  176: 	health state: the date of the interview describes the actual state
  177: 	and not the date of a change in health state. The former idea was
  178: 	to consider that at each interview the state was recorded
  179: 	(healthy, disable or death) and IMaCh was corrected; but when we
  180: 	introduced the exact date of death then we should have modified
  181: 	the contribution of an exact death to the likelihood. This new
  182: 	contribution is smaller and very dependent of the step unit
  183: 	stepm. It is no more the probability to die between last interview
  184: 	and month of death but the probability to survive from last
  185: 	interview up to one month before death multiplied by the
  186: 	probability to die within a month. Thanks to Chris
  187: 	Jackson for correcting this bug.  Former versions increased
  188: 	mortality artificially. The bad side is that we add another loop
  189: 
  190: 
  191: 	which slows down the processing. The difference can be up to 10%
  192: 
  193: 
  194: 	lower mortality.
  195: 
  196: 
  197: 
  198: 
  199: 
  200: 2002-05-30    <brouard@BROUARD>
  201: 
  202: 
  203: 
  204: 
  205: 
  206: 	* 
  207: 
  208: 
  209: 
  210: 
  211: 
  212: 	* imach.c (Module): Add correlation matrix of one-step
  213: 
  214: 
  215: 	probabilities and covariance matrix
  216: 
  217: 
  218: 
  219: 
  220: