X-Git-Url: https://pintos-os.org/cgi-bin/gitweb.cgi?a=blobdiff_plain;ds=sidebyside;f=src%2Flanguage%2Fstats%2Fmann-whitney.c;h=e117cc366cbfd24aefacd5631eca45a1a81c2e0c;hb=ad4f9604cc877272fe504234e8b9078a7ce76e17;hp=4db82f8e904be5d7bf2a18460b319bc1b37717a1;hpb=81579d9e9f994fb2908f50af41c3eb033d216e58;p=pspp

diff --git a/src/language/stats/mann-whitney.c b/src/language/stats/mann-whitney.c
index 4db82f8e90..e117cc366c 100644
--- a/src/language/stats/mann-whitney.c
+++ b/src/language/stats/mann-whitney.c
@@ -22,9 +22,9 @@
 
 #include "data/case.h"
 #include "data/casereader.h"
+#include "data/dataset.h"
 #include "data/dictionary.h"
 #include "data/format.h"
-#include "data/procedure.h"
 #include "data/variable.h"
 #include "libpspp/cast.h"
 #include "libpspp/misc.h"
@@ -47,13 +47,33 @@ struct mw
 
   double u;  /* The Mann-Whitney U statistic */
   double w;  /* The Wilcoxon Rank Sum W statistic */
-  double z;  
+  double z;
 };
 
 static void show_ranks_box (const struct n_sample_test *nst, const struct mw *mw);
 static void show_statistics_box (const struct n_sample_test *nst, const struct mw *mw, bool exact);
 
 
+
+static bool
+belongs_to_test (const struct ccase *c, void *aux)
+{
+  const struct n_sample_test *nst = aux;
+
+  const union value *group = case_data (c, nst->indep_var);
+  const size_t group_var_width = var_get_width (nst->indep_var);
+
+  if ( value_equal (group, &nst->val1, group_var_width))
+    return true;
+
+  if ( value_equal (group, &nst->val2, group_var_width))
+    return true;
+
+  return false;
+}
+
+
+
 void
 mann_whitney_execute (const struct dataset *ds,
 		      struct casereader *input,
@@ -79,9 +99,19 @@ mann_whitney_execute (const struct dataset *ds,
       struct casereader *rr;
       struct ccase *c;
       const struct variable *var = nst->vars[i];
-      
+
       struct casereader *reader =
-	sort_execute_1var (casereader_clone (input), var);
+	casereader_create_filter_func (casereader_clone (input),
+				       belongs_to_test,
+				       NULL,
+				       CONST_CAST (struct n_sample_test *, nst),
+				       NULL);
+
+      reader = casereader_create_filter_missing (reader, &var, 1,
+						 exclude,
+						 NULL, NULL);
+
+      reader = sort_execute_1var (reader, var);
 
       rr = casereader_create_append_rank (reader, var,
 					  dict_get_weight (dict),
@@ -90,14 +120,10 @@ mann_whitney_execute (const struct dataset *ds,
 
       for (; (c = casereader_read (rr)); case_unref (c))
 	{
-	  const union value *val = case_data (c, var);
 	  const union value *group = case_data (c, nst->indep_var);
 	  const size_t group_var_width = var_get_width (nst->indep_var);
 	  const double rank = case_data_idx (c, rank_idx)->f;
 
-	  if ( var_is_value_missing (var, val, exclude))
-	    continue;
-
 	  if ( value_equal (group, &nst->val1, group_var_width))
 	    {
 	      mw[i].rank_sum[0] += rank;
@@ -133,7 +159,7 @@ mann_whitney_execute (const struct dataset *ds,
 	denominator -= tiebreaker;
 	denominator *= mwv->n[0] * mwv->n[1];
 	denominator /= n * (n - 1);
-      
+
 	mwv->z /= sqrt (denominator);
       }
     }
@@ -211,27 +237,27 @@ show_ranks_box (const struct n_sample_test *nst, const struct mw *mwv)
 		var_to_string (nst->vars[i]));
 
       tab_double (table, 1, column_headers + i, 0,
-		  mw->n[0], 0);
+		  mw->n[0], NULL, RC_OTHER);
 
       tab_double (table, 2, column_headers + i, 0,
-		  mw->n[1], 0);
+		  mw->n[1], NULL, RC_OTHER);
 
       tab_double (table, 3, column_headers + i, 0,
-		  mw->n[1] + mw->n[0], 0);
+		  mw->n[1] + mw->n[0], NULL, RC_OTHER);
 
       /* Mean Ranks */
       tab_double (table, 4, column_headers + i, 0,
-		  mw->rank_sum[0] / mw->n[0], 0);
+		  mw->rank_sum[0] / mw->n[0], NULL, RC_OTHER);
 
       tab_double (table, 5, column_headers + i, 0,
-		  mw->rank_sum[1] / mw->n[1], 0);
+		  mw->rank_sum[1] / mw->n[1], NULL, RC_OTHER);
 
       /* Sum of Ranks */
       tab_double (table, 6, column_headers + i, 0,
-		  mw->rank_sum[0], 0);
+		  mw->rank_sum[0], NULL, RC_OTHER);
 
       tab_double (table, 7, column_headers + i, 0,
-		  mw->rank_sum[1], 0);
+		  mw->rank_sum[1], NULL, RC_OTHER);
     }
 
   tab_submit (table);
@@ -266,7 +292,7 @@ show_statistics_box (const struct n_sample_test *nst, const struct mw *mwv, bool
   tab_text (table, 3, 0, TAT_TITLE | TAB_CENTER, _("Z"));
   tab_text (table, 4, 0, TAT_TITLE | TAB_CENTER, _("Asymp. Sig. (2-tailed)"));
 
-  if (exact) 
+  if (exact)
     {
       tab_text (table, 5, 0, TAT_TITLE | TAB_CENTER, _("Exact Sig. (2-tailed)"));
       tab_text (table, 6, 0, TAT_TITLE | TAB_CENTER, _("Point Probability"));
@@ -280,16 +306,16 @@ show_statistics_box (const struct n_sample_test *nst, const struct mw *mwv, bool
 		var_to_string (nst->vars[i]));
 
       tab_double (table, 1, column_headers + i, 0,
-		  mw->u, 0);
+		  mw->u, NULL, RC_OTHER);
 
       tab_double (table, 2, column_headers + i, 0,
-		  mw->w, 0);
+		  mw->w, NULL, RC_OTHER);
 
       tab_double (table, 3, column_headers + i, 0,
-		  mw->z, 0);
+		  mw->z, NULL, RC_OTHER);
 
       tab_double (table, 4, column_headers + i, 0,
-		  2.0 * gsl_cdf_ugaussian_P (mw->z), 0);
+		  2.0 * gsl_cdf_ugaussian_P (mw->z), NULL, RC_PVALUE);
     }
 
   tab_submit (table);