From: N. Brouard <brouard@ined.fr>
Date: Fri, 22 Jul 2022 12:27:48 +0000 (+0000)
Subject: *  imach.c (Module): Output of Wald test in the htm file and not only in the log.
X-Git-Tag: imach-099s7~113
X-Git-Url: https://henry.ined.fr/git/?a=commitdiff_plain;h=26524ebc5da5b99d4f3e455ec3daa866c2099e67;p=.git

*  imach.c (Module): Output of Wald test in the htm file and not only in the log.
---

diff --git a/src/imach.c b/src/imach.c
index ae19c93..315f2f1 100644
--- a/src/imach.c
+++ b/src/imach.c
@@ -1,6 +1,11 @@
 /* $Id$
   $State$
   $Log$
+  Revision 1.321  2022/07/22 12:04:24  brouard
+  Summary: r28
+
+  *  imach.c (Module): Output of Wald test in the htm file and not only in the log.
+
   Revision 1.320  2022/06/02 05:10:11  brouard
   *** empty log message ***
 
@@ -12551,7 +12556,7 @@ Please run with mle=-1 to get a correct covariance matrix.\n",ageminpar,agemaxpa
       hesscov(matcov, hess, p, npar, delti, ftolhess, func);
       printf("Parameters and 95%% confidence intervals\n W is simply the result of the division of the parameter by the square root of covariance of the parameter.\n And Wald-based confidence intervals plus and minus 1.96 * W .\n But be careful that parameters are highly correlated because incidence of disability is highly correlated to incidence of recovery.\n It might be better to visualize the covariance matrix. See the page 'Matrix of variance-covariance of one-step probabilities' and its graphs.\n");
       fprintf(ficlog, "Parameters, Wald tests and Wald-based confidence intervals\n W is simply the result of the division of the parameter by the square root of covariance of the parameter.\n And Wald-based confidence intervals plus and minus 1.96 * W \n  It might be better to visualize the covariance matrix. See the page 'Matrix of variance-covariance of one-step probabilities' and its graphs.\n");
-      fprintf(fichtm, "\n<p>Parameters, Wald tests and Wald-based confidence intervals\n</br> W is simply the result of the division of the parameter by the square root of covariance of the parameter.\n</br> And Wald-based confidence intervals plus and minus 1.96 * W \n </br> It might be better to visualize the covariance matrix. See the page 'Matrix of variance-covariance of one-step probabilities' and its graphs.\n</br>");
+      fprintf(fichtm, "\n<p>The Wald test results are output only if the maximimzation of the Likelihood is performed (mle=1)\n</br>Parameters, Wald tests and Wald-based confidence intervals\n</br> W is simply the result of the division of the parameter by the square root of covariance of the parameter.\n</br> And Wald-based confidence intervals plus and minus 1.96 * W \n </br> It might be better to visualize the covariance matrix. See the page '<a href=\"%s\">Matrix of variance-covariance of one-step probabilities and its graphs</a>'.\n</br>",optionfilehtmcov);
       fprintf(fichtm,"\n<table style=\"text-align:center; border: 1px solid\">");
       fprintf(fichtm, "\n<tr><th>Model=</th><th>1</th><th>+ age</th>");
       if(nagesqr==1){