Return to ChangeLog CVS log

Up to [Local Repository] / imach / src

*** empty log message ***

    1: 2003-06-25    <brouard@BROUARD>
    2: 
    3: 	* imach.c (Module): On windows (cygwin) function asctime_r doesn't
    4: 	exist so I changed back to asctime which exists.
    5: 	(Module): Version 0.96b
    6: 
    7: 2003-06-24    <brouard@BROUARD>
    8: 
    9: 	* imach.c (Module): Some bugs corrected for windows. Also, when
   10: 	mle=-1 a template is output in file "or"mypar.txt with the design
   11: 	of the covariance matrix to be input.
   12: 
   13: 2003-06-25  Brouard Nicolas  <brouard@homebrou>
   14: 
   15: 	* imach.c (Repository): Duplicated warning errors corrected.
   16: 	(Repository): Elapsed time after each iteration is now output. It
   17: 	helps to forecast when convergence will be reached. Elapsed time
   18: 	is stamped in powell.  We created a new html file for the graphs
   19: 	concerning matrix of covariance. It has extension -cov.htm.
   20: 
   21: 2003-06-23  Brouard Nicolas  <brouard@homebrou>
   22: 
   23: 	* 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.
   24: 
   25: 2003-06-18  Brouard Nicolas  <brouard@homebrou>
   26: 
   27: 	* imach.c (Repository): Suppress abusive calls to append to file html (solution was to put fichtm as a global variable. Version 0.96
   28: 
   29: 2003-06-17  Brouard Nicolas  <brouard@homebrou>
   30: 
   31: 
   32: 	* imach.c (Repository): Check when date of death was earlier that
   33: 	current date of interview. It may happen when the death was just
   34: 	prior to the death. In this case, dh was negative and likelihood
   35: 	was wrong (infinity). We still send an "Error" but patch by
   36: 	assuming that the date of death was just one stepm after the
   37: 	interview.
   38: 	(Repository): Because some people have very long ID (first column)
   39: 	we changed int to long in num[] and we added a new lvector for
   40: 	memory allocation. But we also truncated to 8 characters (left
   41: 	truncation)
   42: 
   43: 	(Repository): No more line truncation errors.
   44: 
   45: 2003-06-14  Brouard Nicolas  <brouard@homebrou>
   46: 
   47: 	* imach.c (Repository): Add of a routine likelione (likelihood
   48: 	only once) which prints on a text file
   49: 	(ilk) the contributions to the likelihood for each
   50: 	individual/wave.
   51: 
   52: 2003-06-13  Brouard Nicolas  <brouard@homebrou>
   53: 	* imach.c (Repository): Replace "freqsummary" at a correct
   54: 	place. It differs from routine "prevalence" which may be called
   55: 	many times. Probs is memory consuming and must be used with
   56: 	parcimony.
   57: 
   58: 2003-06-17    <brouard@BROUARD>
   59: 
   60: 	* timeval.h (Module): Added included file to make use of
   61: 	gettimeofday working on win32 with cygwin.
   62: 
   63: 2003-05-16  Brouard Nicolas  <brouard@brouard>
   64: 	* imach.c (Module): Insert of a warning if the delay between two
   65: 	waves is negative.
   66: 
   67: 	* imach.c (Module): 
   68: 
   69: 2003-05-03    <brouard@BROUARD>
   70: 
   71: 	* imach.c: Some cleaning
   72: 
   73: 2003-03-28    <brouard@BROUARD>
   74: 	* imach.c (Module): In version up to 0.92 likelihood was computed
   75: 	as if date of death was unknown. Death was treated as any other
   76: 	health state: the date of the interview describes the actual state
   77: 	and not the date of a change in health state. The former idea was
   78: 	to consider that at each interview the state was recorded
   79: 	(healthy, disable or death) and IMaCh was corrected; but when we
   80: 	introduced the exact date of death then we should have modified
   81: 	the contribution of an exact death to the likelihood. This new
   82: 	contribution is smaller and very dependent of the step unit
   83: 	stepm. It is no more the probability to die between last interview
   84: 	and month of death but the probability to survive from last
   85: 	interview up to one month before death multiplied by the
   86: 	probability to die within a month. Thanks to Chris
   87: 	Jackson for correcting this bug.  Former versions increased
   88: 	mortality artificially. The bad side is that we add another loop
   89: 
   90: 
   91: 	which slows down the processing. The difference can be up to 10%
   92: 
   93: 
   94: 	lower mortality.
   95: 
   96: 
   97: 
   98: 
   99: 
  100: 2002-05-30    <brouard@BROUARD>
  101: 
  102: 
  103: 
  104: 
  105: 
  106: 	* 
  107: 
  108: 
  109: 
  110: 
  111: 
  112: 	* imach.c (Module): Add correlation matrix of one-step
  113: 
  114: 
  115: 	probabilities and covariance matrix
  116: 
  117: 
  118: 
  119: 
  120: