/* 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
#define _CATEGORICALS__
#include <stddef.h>
-#include <data/missing-values.h>
+#include "data/missing-values.h"
struct categoricals;
struct variable;
struct ccase;
+struct interaction;
union value ;
typedef void *user_data_create_func (void *aux1, void *aux2);
-struct categoricals *categoricals_create (const struct variable *const *v, size_t n_vars,
+struct categoricals *categoricals_create (struct interaction *const *, size_t n_int,
const struct variable *wv, enum mv_class exclude,
user_data_create_func *udf,
update_func *update, void *aux1, void *aux2);
/* Return the number of categories (distinct values) for variable N */
size_t categoricals_n_count (const struct categoricals *cat, size_t n);
+size_t categoricals_df (const struct categoricals *cat, size_t n);
/* Return the total number of categories */
-size_t categoricals_total (const struct categoricals *cat);
+size_t categoricals_n_total (const struct categoricals *cat);
+
+/* Return the total degrees of freedom */
+size_t categoricals_df_total (const struct categoricals *cat);
+
/*
Return the total number of variables which participated in these categoricals.
*/
size_t categoricals_get_n_variables (const struct categoricals *cat);
-void categoricals_done (struct categoricals *cat);
-const struct variable * categoricals_get_variable_by_subscript (const struct categoricals *cat, int subscript);
+/*
+ Must be called (once) before any call to the *_by_subscript or *_by_category
+ functions, but AFTER any calls to categoricals_update
+*/
+void categoricals_done (const struct categoricals *cat);
-const union value * categoricals_get_value_by_subscript (const struct categoricals *cat, int subscript);
+/*
+ The *_by_subscript functions use the short map.
+ Their intended use is by covariance matrix routines, where normally 1 less than
+ the total number of distinct values of each categorical variable should
+ be considered.
+ */
double categoricals_get_weight_by_subscript (const struct categoricals *cat, int subscript);
+const struct interaction *categoricals_get_interaction_by_subscript (const struct categoricals *cat, int subscript);
double categoricals_get_sum_by_subscript (const struct categoricals *cat, int subscript);
-double categoricals_get_binary_by_subscript (const struct categoricals *cat, int subscript,
- const struct ccase *c);
+double categoricals_get_code_for_case (const struct categoricals *cat, int subscript, const struct ccase *c);
+
-void * categoricals_get_user_data_by_subscript (const struct categoricals *cat, int subscript);
+/* These use the long map. Useful for descriptive statistics. */
+/* Return the value corresponding to the N'th category */
+const union value * categoricals_get_value_by_category (const struct categoricals *cat, int n);
+void * categoricals_get_user_data_by_category (const struct categoricals *cat, int category);
+const struct ccase * categoricals_get_case_by_category (const struct categoricals *cat, int subscript);
#endif