X-Git-Url: https://pintos-os.org/cgi-bin/gitweb.cgi?a=blobdiff_plain;f=src%2Fmath%2Flinreg%2Flinreg.c;h=6bd450a1a9b3ecfe3833e0e11e6015ea43e83665;hb=da333d7456a56655ebf4ce0e16e6cf468ab1c1af;hp=9ba84f60ece9d8c3991be31cce09766129e886ef;hpb=43b1296aafe7582e7dbe6c2b6a8b478d7d9b0fcf;p=pspp-builds.git

diff --git a/src/math/linreg/linreg.c b/src/math/linreg/linreg.c
index 9ba84f60..6bd450a1 100644
--- a/src/math/linreg/linreg.c
+++ b/src/math/linreg/linreg.c
@@ -1,5 +1,5 @@
 /* PSPP - a program for statistical analysis.
-   Copyright (C) 2005 Free Software Foundation, Inc. Written by Jason H. Stover.
+   Copyright (C) 2005 Free Software Foundation, Inc. 
 
    This program is free software: you can redistribute it and/or modify
    it under the terms of the GNU General Public License as published by
@@ -110,7 +110,7 @@ pspp_linreg_get_vars (const void *c_, const struct variable **v)
   /*
      Start at c->coeff[1] to avoid the intercept.
    */
-  v[result] =  pspp_coeff_get_var (c->coeff[1], 0);
+  v[result] = pspp_coeff_get_var (c->coeff[1], 0);
   result = (v[result] == NULL) ? 0 : 1;
 
   for (coef = c->coeff[2]; coef < c->coeff[c->n_coeffs]; coef++)
@@ -184,10 +184,12 @@ pspp_linreg_cache_free (void *m)
       gsl_vector_free (c->indep_std);
       gsl_vector_free (c->ss_indeps);
       gsl_matrix_free (c->cov);
+      gsl_vector_free (c->ssx);
       for (i = 0; i < c->n_coeffs; i++)
 	{
 	  pspp_coeff_free (c->coeff[i]);
 	}
+      free (c->coeff);
       free (c);
     }
   return true;
@@ -203,7 +205,7 @@ pspp_linreg (const gsl_vector * Y, const gsl_matrix * X,
 	     const pspp_linreg_opts * opts, pspp_linreg_cache * cache)
 {
   int rc;
-  gsl_matrix *design;
+  gsl_matrix *design = NULL;
   gsl_matrix_view xtx;
   gsl_matrix_view xm;
   gsl_matrix_view xmxtx;
@@ -366,6 +368,18 @@ pspp_linreg (const gsl_vector * Y, const gsl_matrix * X,
 	}
       gsl_matrix_free (sw);
     }
+  else if (cache->method == PSPP_LINREG_CONDITIONAL_INVERSE)
+    {
+      /*
+	Use the SVD of X^T X to find a conditional inverse of X^TX. If
+	the SVD is X^T X = U D V^T, then set the conditional inverse
+	to (X^T X)^c = V D^- U^T. D^- is defined as follows: If entry
+	(i, i) has value sigma_i, then entry (i, i) of D^- is 1 /
+	sigma_i if sigma_i > 0, and 0 otherwise. Then solve the normal
+	equations by setting the estimated parameter vector to 
+	(X^TX)^c X^T Y.
+       */
+    }
   else
     {
       gsl_multifit_linear_workspace *wk;