/* PSPP - a program for statistical analysis.
- Copyright (C) 2009 Free Software Foundation, Inc.
+ Copyright (C) 2009, 2010, 2011 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
#include <config.h>
+#include "math/categoricals.h"
+
#include <stdio.h>
-#include "categoricals.h"
+#include "data/case.h"
+#include "data/value.h"
+#include "data/variable.h"
+#include "libpspp/array.h"
+#include "libpspp/hmap.h"
+#include "libpspp/pool.h"
+#include "libpspp/str.h"
+
+#include "gl/xalloc.h"
+
+struct value_node
+{
+ struct hmap_node node; /* Node in hash map. */
+ union value value; /* The value being labeled. */
+ double cc; /* The total of the weights of cases with this value */
+
+ void *user_data; /* A pointer to data which the caller can store stuff */
+
+ int subscript; /* A zero based integer, unique within the variable.
+ Can be used as an index into an array */
+};
+
+struct var_params
+{
+ /* A map indexed by a union values */
+ struct hmap map;
+
+ const struct variable *var;
+
+ int base_subscript;
+
+ /* The number of distinct values of this variable */
+ int n_cats;
+
+ /* A map of values indexed by subscript */
+ const struct value_node **reverse_value_map;
+
+ /* Total of the weights of this variable */
+ double cc;
+};
+
+
+/* Comparison function to sort the reverse_value_map in ascending order */
+static int
+compare_value_node (const void *vn1_, const void *vn2_, const void *aux)
+{
+ const struct value_node * const *vn1 = vn1_;
+ const struct value_node * const *vn2 = vn2_;
+ const struct var_params *vp = aux;
+
+ return value_compare_3way (&(*vn1)->value, &(*vn2)->value, var_get_width (vp->var));
+}
-#include <gl/xalloc.h>
-#include <data/variable.h>
-#include <data/case.h>
-#include <data/value.h>
-#include <libpspp/hmap.h>
struct categoricals
{
- const struct variable **vars;
+ /* The weight variable */
+ const struct variable *wv;
+
+
+ /* An array of var_params */
+ struct var_params *vp;
+
+ /* The size of VP. (ie, the number of variables involved.) */
+ size_t n_vp;
+
+ /* The number of categorical variables which contain entries.
+ In the absence of missing values, this will be equal to N_VP */
size_t n_vars;
- const struct variable *wv;
+ /* A map to enable the lookup of variables indexed by subscript */
+ int *reverse_variable_map;
+
+ size_t n_cats_total;
+
+ struct pool *pool;
- struct hmap *map;
+ /* Missing values to be excluded */
+ enum mv_class exclude;
- int *next_index;
+ /* Function to be called on each update */
+ update_func *update;
- size_t n_cats;
+ /* Function specified by the caller to create user_data */
+ user_data_create_func *user_data_create;
+
+ /* Auxilliary data to be passed to update and user_data_create_func*/
+ void *aux1;
+ void *aux2;
};
-struct value_node
- {
- struct hmap_node node; /* Node in hash map. */
- union value value; /* The value being labeled. */
- double cc; /* The total of the weights of cases with this value */
- int index; /* A zero based integer, unique within the variable.
- Can be used as an index into an array */
- };
+void
+categoricals_destroy ( struct categoricals *cat)
+{
+ int i;
+ if (cat != NULL)
+ {
+ for (i = 0 ; i < cat->n_vp; ++i)
+ hmap_destroy (&cat->vp[i].map);
+
+ pool_destroy (cat->pool);
+ free (cat);
+ }
+}
+
+
+#if 0
+void
+categoricals_dump (const struct categoricals *cat)
+{
+ int v;
+
+ for (v = 0 ; v < cat->n_vp; ++v)
+ {
+ const struct var_params *vp = &cat->vp[v];
+ const struct hmap *m = &vp->map;
+ struct hmap_node *node ;
+ int x;
+
+ printf ("\n%s (%d) CC=%g n_cats=%d:\n",
+ var_get_name (vp->var), vp->base_subscript, vp->cc, vp->n_cats);
+
+ printf ("Reverse map\n");
+ for (x = 0 ; x < vp->n_cats; ++x)
+ {
+ struct string s;
+ const struct value_node *vn = vp->reverse_value_map[x];
+ ds_init_empty (&s);
+ var_append_value_name (vp->var, &vn->value, &s);
+ printf ("Value for %d is %s\n", x, ds_cstr(&s));
+ ds_destroy (&s);
+ }
+
+ printf ("\nForward map\n");
+ for (node = hmap_first (m); node; node = hmap_next (m, node))
+ {
+ struct string s;
+ const struct value_node *vn = HMAP_DATA (node, struct value_node, node);
+ ds_init_empty (&s);
+ var_append_value_name (vp->var, &vn->value, &s);
+ printf ("Value: %s; Index %d; CC %g\n",
+ ds_cstr (&s),
+ vn->subscript, vn->cc);
+ ds_destroy (&s);
+ }
+ }
+
+ assert (cat->n_vars <= cat->n_vp);
+
+ printf ("\n");
+ printf ("Number of categorical variables: %d\n", cat->n_vp);
+ printf ("Number of non-empty categorical variables: %d\n", cat->n_vars);
+ printf ("Total number of categories: %d\n", cat->n_cats_total);
+
+ printf ("\nReverse variable map:\n");
+
+ for (v = 0 ; v < cat->n_cats_total - cat->n_vars; ++v)
+ printf ("%d ", cat->reverse_variable_map[v]);
+ printf ("\n");
+}
+#endif
+
static struct value_node *
{
if (value_equal (val, &foo->value, width))
break;
+
+ fprintf (stderr, "Warning: Hash table collision\n");
}
return foo;
struct categoricals *
-categoricals_create (const struct variable **v, size_t n_vars, const struct variable *wv)
+categoricals_create (const struct variable *const *v, size_t n_vars,
+ const struct variable *wv, enum mv_class exclude,
+ user_data_create_func *udf,
+ update_func *update, void *aux1, void *aux2
+ )
{
size_t i;
struct categoricals *cat = xmalloc (sizeof *cat);
- cat->vars = v;
- cat->n_vars = n_vars;
+ cat->n_vp = n_vars;
cat->wv = wv;
- cat->n_cats = 0;
+ cat->n_cats_total = 0;
+ cat->n_vars = 0;
+ cat->reverse_variable_map = NULL;
+ cat->pool = pool_create ();
+ cat->exclude = exclude;
+ cat->update = update;
+ cat->user_data_create = udf;
+
+ cat->aux1 = aux1;
+ cat->aux2 = aux2;
+
- cat->map = xmalloc (sizeof *cat->map * n_vars);
- cat->next_index = xcalloc (sizeof *cat->next_index, n_vars);
+ cat->vp = pool_calloc (cat->pool, cat->n_vp, sizeof *cat->vp);
- for (i = 0 ; i < cat->n_vars; ++i)
+ for (i = 0 ; i < cat->n_vp; ++i)
{
- hmap_init (&cat->map[i]);
+ hmap_init (&cat->vp[i].map);
+ cat->vp[i].var = v[i];
}
return cat;
}
+
void
categoricals_update (struct categoricals *cat, const struct ccase *c)
{
const double weight = cat->wv ? case_data (c, cat->wv)->f : 1.0;
- for (i = 0 ; i < cat->n_vars; ++i)
+ assert (NULL == cat->reverse_variable_map);
+
+ for (i = 0 ; i < cat->n_vp; ++i)
{
- unsigned int width = var_get_width (cat->vars[i]);
- const union value *val = case_data (c, cat->vars[i]);
- size_t hash = value_hash (val, width, 0);
+ const struct variable *var = cat->vp[i].var;
+ unsigned int width = var_get_width (var);
+ const union value *val = case_data (c, var);
+ size_t hash;
+ struct value_node *node ;
- struct value_node *node = lookup_value (&cat->map[i], cat->vars[i], val);
+ if ( var_is_value_missing (var, val, cat->exclude))
+ continue;
+
+ hash = value_hash (val, width, 0);
+ node = lookup_value (&cat->vp[i].map, var, val);
if ( NULL == node)
{
- node = xmalloc (sizeof *node);
+ node = pool_malloc (cat->pool, sizeof *node);
value_init (&node->value, width);
value_copy (&node->value, val, width);
node->cc = 0.0;
- hmap_insert (&cat->map[i], &node->node, hash);
- cat->n_cats ++;
- node->index = cat->next_index[i]++ ;
+ hmap_insert (&cat->vp[i].map, &node->node, hash);
+ cat->n_cats_total++;
+
+ if ( 0 == cat->vp[i].n_cats)
+ cat->n_vars++;
+
+ node->subscript = cat->vp[i].n_cats++ ;
+
+ if (cat->user_data_create)
+ node->user_data = cat->user_data_create (cat->aux1, cat->aux2);
}
node->cc += weight;
+ cat->vp[i].cc += weight;
+
+ if (cat->update)
+ cat->update (node->user_data, cat->exclude, cat->wv, var, c, cat->aux1, cat->aux2);
}
}
size_t
categoricals_n_count (const struct categoricals *cat, size_t n)
{
- return hmap_count (&cat->map[n]);
+ return hmap_count (&cat->vp[n].map);
+}
+
+
+/* Return the total number of categories */
+size_t
+categoricals_total (const struct categoricals *cat)
+{
+ return cat->n_cats_total;
}
-/* Return the index for value VAL in the Nth variable */
-int
-categoricals_index (const struct categoricals *cat, size_t n, const union value *val)
+/* This function must be called *before* any call to categoricals_get_*_by subscript an
+ *after* all calls to categoricals_update */
+void
+categoricals_done (const struct categoricals *cat_)
{
- struct value_node *vn = lookup_value (&cat->map[n], cat->vars[n], val);
+ /* Implementation Note: Whilst this function is O(n) in cat->n_cats_total, in most
+ uses it will be more efficient that using a tree based structure, since it
+ is called only once, and means that subsequent lookups will be O(1).
+
+ 1 call of O(n) + 10^9 calls of O(1) is better than 10^9 calls of O(log n).
+ */
+ struct categoricals *cat = CONST_CAST (struct categoricals *, cat_);
+ int v;
+ int idx = 0;
+ cat->reverse_variable_map = pool_calloc (cat->pool,
+ cat->n_cats_total - cat->n_vars,
+ sizeof *cat->reverse_variable_map);
+
+ for (v = 0 ; v < cat->n_vp; ++v)
+ {
+ int i;
+ struct var_params *vp = &cat->vp[v];
+ int n_cats_total = categoricals_n_count (cat, v);
+ struct hmap_node *node ;
+
+ vp->reverse_value_map = pool_calloc (cat->pool, n_cats_total, sizeof *vp->reverse_value_map);
+
+ vp->base_subscript = idx;
+
+ for (node = hmap_first (&vp->map); node; node = hmap_next (&vp->map, node))
+ {
+ const struct value_node *vn = HMAP_DATA (node, struct value_node, node);
+ vp->reverse_value_map[vn->subscript] = vn;
+ }
- if ( vn == NULL)
- return -1;
+ /* For some purposes (eg CONTRASTS in ONEWAY) the values need to be sorted */
+ sort (vp->reverse_value_map, vp->n_cats, sizeof (const struct value_node *),
+ compare_value_node, vp);
- return vn->index;
+ /* Populate the reverse variable map.
+ */
+ for (i = 0; i < vp->n_cats - 1; ++i)
+ cat->reverse_variable_map[idx++] = v;
+ }
+
+ assert (cat->n_vars <= cat->n_vp);
}
-/* Return the total number of categories */
-size_t
-categoricals_total (const struct categoricals *cat)
+static int
+reverse_variable_lookup (const struct categoricals *cat, int subscript)
+{
+ assert (cat->reverse_variable_map);
+ assert (subscript >= 0);
+ assert (subscript < cat->n_cats_total - cat->n_vars);
+
+ return cat->reverse_variable_map[subscript];
+}
+
+
+/* Return the categorical variable corresponding to SUBSCRIPT */
+const struct variable *
+categoricals_get_variable_by_subscript (const struct categoricals *cat, int subscript)
+{
+ int index = reverse_variable_lookup (cat, subscript);
+
+ return cat->vp[index].var;
+}
+
+/* Return the value corresponding to SUBSCRIPT */
+const union value *
+categoricals_get_value_by_subscript (const struct categoricals *cat, int subscript)
+{
+ int vindex = reverse_variable_lookup (cat, subscript);
+ const struct var_params *vp = &cat->vp[vindex];
+ const struct value_node *vn = vp->reverse_value_map [subscript - vp->base_subscript];
+
+ return &vn->value;
+}
+
+
+double
+categoricals_get_weight_by_subscript (const struct categoricals *cat, int subscript)
+{
+ int vindex = reverse_variable_lookup (cat, subscript);
+ const struct var_params *vp = &cat->vp[vindex];
+
+ return vp->cc;
+}
+
+double
+categoricals_get_sum_by_subscript (const struct categoricals *cat, int subscript)
+{
+ int vindex = reverse_variable_lookup (cat, subscript);
+ const struct var_params *vp = &cat->vp[vindex];
+
+ const struct value_node *vn = vp->reverse_value_map [subscript - vp->base_subscript];
+ return vn->cc;
+}
+
+
+/* Returns unity if the value in case C at SUBSCRIPT is equal to the category
+ for that subscript */
+double
+categoricals_get_binary_by_subscript (const struct categoricals *cat, int subscript,
+ const struct ccase *c)
{
- return cat->n_cats;
+ const struct variable *var = categoricals_get_variable_by_subscript (cat, subscript);
+ int width = var_get_width (var);
+
+ const union value *val = case_data (c, var);
+
+ return value_equal (val, categoricals_get_value_by_subscript (cat, subscript), width);
}
+size_t
+categoricals_get_n_variables (const struct categoricals *cat)
+{
+ return cat->n_vars;
+}
+void *
+categoricals_get_user_data_by_subscript (const struct categoricals *cat, int subscript)
+{
+ int vindex = reverse_variable_lookup (cat, subscript);
+ const struct var_params *vp = &cat->vp[vindex];
+ const struct value_node *vn = vp->reverse_value_map [subscript - vp->base_subscript];
+ return vn->user_data;
+}