Return to ChangeLog CVS log

Up to [Local Repository] / imach / src

*** empty log message ***

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