#include <data/case.h>
#include <data/variable.h>
#include <libpspp/misc.h>
+#include "categoricals.h"
#define n_MOMENTS (MOMENT_VARIANCE + 1)
+/* Create a new matrix of NEW_SIZE x NEW_SIZE and copy the elements of
+ matrix IN into it. IN must be a square matrix, and in normal usage
+ it will be smaller than NEW_SIZE.
+ IN is destroyed by this function. The return value must be destroyed
+ when no longer required.
+*/
+static gsl_matrix *
+resize_matrix (gsl_matrix *in, size_t new_size)
+{
+ size_t i, j;
+
+ gsl_matrix *out = NULL;
+
+ assert (in->size1 == in->size2);
+
+ if (new_size <= in->size1)
+ return in;
+
+ out = gsl_matrix_calloc (new_size, new_size);
+
+ for (i = 0; i < in->size1; ++i)
+ {
+ for (j = 0; j < in->size2; ++j)
+ {
+ double x = gsl_matrix_get (in, i, j);
+
+ gsl_matrix_set (out, i, j, x);
+ }
+ }
+
+ gsl_matrix_free (in);
+
+ return out;
+}
+
struct covariance
{
/* The variables for which the covariance matrix is to be calculated. */
const struct variable **vars;
/* Categorical variables. */
- size_t n_catvars;
- const struct variable **catvars;
+ struct categoricals *categoricals;
/* Array containing number of categories per categorical variable. */
size_t *n_categories;
/* Create a covariance struct.
*/
struct covariance *
-covariance_create (size_t n_vars, const struct variable **vars,
- const struct variable *weight, enum mv_class exclude,
- short passes)
+covariance_1pass_create (size_t n_vars, const struct variable **vars,
+ const struct variable *weight, enum mv_class exclude)
{
size_t i;
struct covariance *cov = xmalloc (sizeof *cov);
- assert (passes == 1 || passes == 2);
- cov->passes = passes;
+
+ cov->passes = 1;
cov->state = 0;
cov->pass_one_first_case_seen = cov->pass_two_first_case_seen = false;
- cov->vars = xmalloc (sizeof *cov->vars * n_vars);
+ cov->vars = vars;
cov->wv = weight;
cov->n_vars = n_vars;
cov->dim = n_vars;
- for (i = 0; i < n_vars; ++i)
- cov->vars[i] = vars[i];
-
cov->moments = xmalloc (sizeof *cov->moments * n_MOMENTS);
for (i = 0; i < n_MOMENTS; ++i)
struct covariance *
covariance_2pass_create (size_t n_vars, const struct variable **vars,
size_t n_catvars, const struct variable **catvars,
- const struct variable *weight, enum mv_class exclude)
+ const struct variable *wv, enum mv_class exclude)
{
size_t i;
struct covariance *cov = xmalloc (sizeof *cov);
- cov->vars = xmalloc (sizeof *cov->vars * n_vars);
- cov->catvars = xnmalloc (n_catvars, sizeof (*cov->catvars));
- cov->n_categories = xnmalloc (n_catvars, sizeof (cov->n_categories));
- cov->wv = weight;
- cov->n_vars = n_vars;
- cov->n_catvars = n_catvars;
-
- for (i = 0; i < n_vars; ++i)
- cov->vars[i] = vars[i];
+ cov->passes = 2;
+ cov->state = 0;
+ cov->pass_one_first_case_seen = cov->pass_two_first_case_seen = false;
+
+ cov->vars = vars;
- for (i = 0; i < n_catvars; i++)
- {
- cov->catvars[i] = catvars[i];
- cov->n_categories[i] = 0;
- }
+ cov->wv = wv;
+ cov->n_vars = n_vars;
+ cov->dim = n_vars;
cov->moments = xmalloc (sizeof *cov->moments * n_MOMENTS);
+ for (i = 0; i < n_MOMENTS; ++i)
+ cov->moments[i] = gsl_matrix_calloc (n_vars, n_vars);
+
cov->exclude = exclude;
+ cov->n_cm = - 1;
+ cov->cm = NULL;
+
+ cov->categoricals = categoricals_create (catvars, n_catvars, wv);
+
return cov;
}
cov->state = 1;
}
+ categoricals_update (cov->categoricals, c);
+
for (i = 0 ; i < cov->n_vars; ++i)
{
const union value *val1 = case_data (c, cov->vars[i]);
assert (cov->state == 1);
cov->state = 2;
+ cov->dim = cov->n_vars + categoricals_total (cov->categoricals);
+ cov->n_cm = (cov->dim * (cov->dim - 1) ) / 2;
+ cov->cm = xcalloc (sizeof *cov->cm, cov->n_cm);
+
+ /* Grow the moment matrices so that they're large enough to accommodate the
+ categorical elements */
+ for (i = 0; i < n_MOMENTS; ++i)
+ {
+ cov->moments[i] = resize_matrix (cov->moments[i], cov->dim);
+ }
+
/* Divide the means by the number of samples */
for (i = 0; i < cov->n_vars; ++i)
{
{
double *x = gsl_matrix_ptr (cov->moments[MOMENT_MEAN], i, j);
*x /= gsl_matrix_get (cov->moments[MOMENT_NONE], i, j);
- }
+ }
}
}
{
size_t i;
free (cov->vars);
+ categoricals_destroy (cov->categoricals);
for (i = 0; i < n_MOMENTS; ++i)
gsl_matrix_free (cov->moments[i]);