--- /dev/null
+/* pspp - a program for statistical analysis.
+ Copyright (C) 2012 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
+ the Free Software Foundation, either version 3 of the License, or
+ (at your option) any later version.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ GNU General Public License for more details.
+
+ You should have received a copy of the GNU General Public License
+ along with this program. If not, see <http://www.gnu.org/licenses/>. */
+
+
+/*
+ References:
+ 1. "Coding Logistic Regression with Newton-Raphson", James McCaffrey
+ http://msdn.microsoft.com/en-us/magazine/jj618304.aspx
+
+ 2. "SPSS Statistical Algorithms" Chapter LOGISTIC REGRESSION Algorithms
+
+
+ The Newton Raphson method finds successive approximations to $\bf b$ where
+ approximation ${\bf b}_t$ is (hopefully) better than the previous ${\bf b}_{t-1}$.
+
+ $ {\bf b}_t = {\bf b}_{t -1} + ({\bf X}^T{\bf W}_{t-1}{\bf X})^{-1}{\bf X}^T({\bf y} - {\bf \pi}_{t-1})$
+ where:
+
+ $\bf X$ is the $n \times p$ design matrix, $n$ being the number of cases,
+ $p$ the number of parameters, \par
+ $\bf W$ is the diagonal matrix whose diagonal elements are
+ $\hat{\pi}_0(1 - \hat{\pi}_0), \, \hat{\pi}_1(1 - \hat{\pi}_2)\dots \hat{\pi}_{n-1}(1 - \hat{\pi}_{n-1})$
+ \par
+
+*/
+
+#include <config.h>
+
+#include <gsl/gsl_blas.h>
+
+#include <gsl/gsl_linalg.h>
+#include <gsl/gsl_cdf.h>
+#include <gsl/gsl_matrix.h>
+#include <gsl/gsl_vector.h>
+#include <math.h>
+
+#include "data/case.h"
+#include "data/casegrouper.h"
+#include "data/casereader.h"
+#include "data/dataset.h"
+#include "data/dictionary.h"
+#include "data/format.h"
+#include "data/value.h"
+#include "language/command.h"
+#include "language/commands/split-file.h"
+#include "language/lexer/lexer.h"
+#include "language/lexer/value-parser.h"
+#include "language/lexer/variable-parser.h"
+#include "libpspp/assertion.h"
+#include "libpspp/hash-functions.h"
+#include "libpspp/hmap.h"
+#include "libpspp/ll.h"
+#include "libpspp/message.h"
+#include "libpspp/misc.h"
+#include "math/categoricals.h"
+#include "math/interaction.h"
+#include "output/pivot-table.h"
+
+#include "gettext.h"
+#define N_(msgid) msgid
+#define _(msgid) gettext (msgid)
+
+
+
+
+#define PRINT_EACH_STEP 0x01
+#define PRINT_SUMMARY 0x02
+#define PRINT_CORR 0x04
+#define PRINT_ITER 0x08
+#define PRINT_GOODFIT 0x10
+#define PRINT_CI 0x20
+
+
+#define PRINT_DEFAULT (PRINT_SUMMARY | PRINT_EACH_STEP)
+
+/*
+ The constant parameters of the procedure.
+ That is, those which are set by the user.
+*/
+struct lr_spec
+{
+ /* The dependent variable */
+ const struct variable *dep_var;
+
+ /* The predictor variables (excluding categorical ones) */
+ const struct variable **predictor_vars;
+ size_t n_predictor_vars;
+
+ /* The categorical predictors */
+ struct interaction **cat_predictors;
+ size_t n_cat_predictors;
+
+
+ /* The union of the categorical and non-categorical variables */
+ const struct variable **indep_vars;
+ size_t n_indep_vars;
+
+
+ /* Which classes of missing vars are to be excluded */
+ enum mv_class exclude;
+
+ /* The weight variable */
+ const struct variable *wv;
+
+ /* The dictionary of the dataset */
+ const struct dictionary *dict;
+
+ /* True iff the constant (intercept) is to be included in the model */
+ bool constant;
+
+ /* Ths maximum number of iterations */
+ int max_iter;
+
+ /* Other iteration limiting conditions */
+ double bcon;
+ double min_epsilon;
+ double lcon;
+
+ /* The confidence interval (in percent) */
+ int confidence;
+
+ /* What results should be presented */
+ unsigned int print;
+
+ /* Inverse logit of the cut point */
+ double ilogit_cut_point;
+};
+
+
+/* The results and intermediate result of the procedure.
+ These are mutated as the procedure runs. Used for
+ temporary variables etc.
+*/
+struct lr_result
+{
+ /* Used to indicate if a pass should flag a warning when
+ invalid (ie negative or missing) weight values are encountered */
+ bool warn_bad_weight;
+
+ /* The two values of the dependent variable. */
+ union value y0;
+ union value y1;
+
+
+ /* The sum of caseweights */
+ double cc;
+
+ /* The number of missing and nonmissing cases */
+ casenumber n_missing;
+ casenumber n_nonmissing;
+
+
+ gsl_matrix *hessian;
+
+ /* The categoricals and their payload. Null if the analysis has no
+ categorical predictors */
+ struct categoricals *cats;
+ struct payload cp;
+
+
+ /* The estimates of the predictor coefficients */
+ gsl_vector *beta_hat;
+
+ /* The predicted classifications:
+ True Negative, True Positive, False Negative, False Positive */
+ double tn, tp, fn, fp;
+};
+
+
+/*
+ Convert INPUT into a dichotomous scalar, according to how the dependent variable's
+ values are mapped.
+ For simple cases, this is a 1:1 mapping
+ The return value is always either 0 or 1
+*/
+static double
+map_dependent_var (const struct lr_spec *cmd, const struct lr_result *res, const union value *input)
+{
+ const int width = var_get_width (cmd->dep_var);
+ if (value_equal (input, &res->y0, width))
+ return 0;
+
+ if (value_equal (input, &res->y1, width))
+ return 1;
+
+ /* This should never happen. If it does, then y0 and/or y1 have probably not been set */
+ NOT_REACHED ();
+
+ return SYSMIS;
+}
+
+static void output_classification_table (const struct lr_spec *cmd, const struct lr_result *res);
+
+static void output_categories (const struct lr_spec *cmd, const struct lr_result *res);
+
+static void output_depvarmap (const struct lr_spec *cmd, const struct lr_result *);
+
+static void output_variables (const struct lr_spec *cmd,
+ const struct lr_result *);
+
+static void output_model_summary (const struct lr_result *,
+ double initial_likelihood, double likelihood);
+
+static void case_processing_summary (const struct lr_result *);
+
+
+/* Return the value of case C corresponding to the INDEX'th entry in the
+ model */
+static double
+predictor_value (const struct ccase *c,
+ const struct variable **x, size_t n_x,
+ const struct categoricals *cats,
+ size_t index)
+{
+ /* Values of the scalar predictor variables */
+ if (index < n_x)
+ return case_num (c, x[index]);
+
+ /* Coded values of categorical predictor variables (or interactions) */
+ if (cats && index - n_x < categoricals_df_total (cats))
+ {
+ double x = categoricals_get_dummy_code_for_case (cats, index - n_x, c);
+ return x;
+ }
+
+ /* The constant term */
+ return 1.0;
+}
+
+
+/*
+ Return the probability beta_hat (that is the estimator logit(y))
+ corresponding to the coefficient estimator for case C
+*/
+static double
+pi_hat (const struct lr_spec *cmd,
+ const struct lr_result *res,
+ const struct variable **x, size_t n_x,
+ const struct ccase *c)
+{
+ int v0;
+ double pi = 0;
+ size_t n_coeffs = res->beta_hat->size;
+
+ if (cmd->constant)
+ {
+ pi += gsl_vector_get (res->beta_hat, res->beta_hat->size - 1);
+ n_coeffs--;
+ }
+
+ for (v0 = 0; v0 < n_coeffs; ++v0)
+ {
+ pi += gsl_vector_get (res->beta_hat, v0) *
+ predictor_value (c, x, n_x, res->cats, v0);
+ }
+
+ pi = 1.0 / (1.0 + exp(-pi));
+
+ return pi;
+}
+
+
+/*
+ Calculates the Hessian matrix X' V X,
+ where: X is the n by N_X matrix comprising the n cases in INPUT
+ V is a diagonal matrix { (pi_hat_0)(1 - pi_hat_0), (pi_hat_1)(1 - pi_hat_1), ... (pi_hat_{N-1})(1 - pi_hat_{N-1})}
+ (the partial derivative of the predicted values)
+
+ If ALL predicted values derivatives are close to zero or one, then CONVERGED
+ will be set to true.
+*/
+static void
+hessian (const struct lr_spec *cmd,
+ struct lr_result *res,
+ struct casereader *input,
+ const struct variable **x, size_t n_x,
+ bool *converged)
+{
+ struct casereader *reader;
+ struct ccase *c;
+
+ double max_w = -DBL_MAX;
+
+ gsl_matrix_set_zero (res->hessian);
+
+ for (reader = casereader_clone (input);
+ (c = casereader_read (reader)) != NULL; case_unref (c))
+ {
+ int v0, v1;
+ double pi = pi_hat (cmd, res, x, n_x, c);
+
+ double weight = dict_get_case_weight (cmd->dict, c, &res->warn_bad_weight);
+ double w = pi * (1 - pi);
+ if (w > max_w)
+ max_w = w;
+ w *= weight;
+
+ for (v0 = 0; v0 < res->beta_hat->size; ++v0)
+ {
+ double in0 = predictor_value (c, x, n_x, res->cats, v0);
+ for (v1 = 0; v1 < res->beta_hat->size; ++v1)
+ {
+ double in1 = predictor_value (c, x, n_x, res->cats, v1);
+ double *o = gsl_matrix_ptr (res->hessian, v0, v1);
+ *o += in0 * w * in1;
+ }
+ }
+ }
+ casereader_destroy (reader);
+
+ if (max_w < cmd->min_epsilon)
+ {
+ *converged = true;
+ msg (MN, _("All predicted values are either 1 or 0"));
+ }
+}
+
+
+/* Calculates the value X' (y - pi)
+ where X is the design model,
+ y is the vector of observed independent variables
+ pi is the vector of estimates for y
+
+ Side effects:
+ the likelihood is stored in LIKELIHOOD;
+ the predicted values are placed in the respective tn, fn, tp fp values in RES
+*/
+static gsl_vector *
+xt_times_y_pi (const struct lr_spec *cmd,
+ struct lr_result *res,
+ struct casereader *input,
+ const struct variable **x, size_t n_x,
+ const struct variable *y_var,
+ double *llikelihood)
+{
+ struct casereader *reader;
+ struct ccase *c;
+ gsl_vector *output = gsl_vector_calloc (res->beta_hat->size);
+
+ *llikelihood = 0.0;
+ res->tn = res->tp = res->fn = res->fp = 0;
+ for (reader = casereader_clone (input);
+ (c = casereader_read (reader)) != NULL; case_unref (c))
+ {
+ double pred_y = 0;
+ int v0;
+ double pi = pi_hat (cmd, res, x, n_x, c);
+ double weight = dict_get_case_weight (cmd->dict, c, &res->warn_bad_weight);
+
+
+ double y = map_dependent_var (cmd, res, case_data (c, y_var));
+
+ *llikelihood += (weight * y) * log (pi) + log (1 - pi) * weight * (1 - y);
+
+ for (v0 = 0; v0 < res->beta_hat->size; ++v0)
+ {
+ double in0 = predictor_value (c, x, n_x, res->cats, v0);
+ double *o = gsl_vector_ptr (output, v0);
+ *o += in0 * (y - pi) * weight;
+ pred_y += gsl_vector_get (res->beta_hat, v0) * in0;
+ }
+
+ /* Count the number of cases which would be correctly/incorrectly classified by this
+ estimated model */
+ if (pred_y <= cmd->ilogit_cut_point)
+ {
+ if (y == 0)
+ res->tn += weight;
+ else
+ res->fn += weight;
+ }
+ else
+ {
+ if (y == 0)
+ res->fp += weight;
+ else
+ res->tp += weight;
+ }
+ }
+
+ casereader_destroy (reader);
+
+ return output;
+}
+
+\f
+
+/* "payload" functions for the categoricals.
+ The only function is to accumulate the frequency of each
+ category.
+ */
+
+static void *
+frq_create (const void *aux1 UNUSED, void *aux2 UNUSED)
+{
+ return xzalloc (sizeof (double));
+}
+
+static void
+frq_update (const void *aux1 UNUSED, void *aux2 UNUSED,
+ void *ud, const struct ccase *c UNUSED , double weight)
+{
+ double *freq = ud;
+ *freq += weight;
+}
+
+static void
+frq_destroy (const void *aux1 UNUSED, void *aux2 UNUSED, void *user_data)
+{
+ free (user_data);
+}
+
+\f
+
+/*
+ Makes an initial pass though the data, doing the following:
+
+ * Checks that the dependent variable is dichotomous,
+ * Creates and initialises the categoricals,
+ * Accumulates summary results,
+ * Calculates necessary initial values.
+ * Creates an initial value for \hat\beta the vector of beta_hats of \beta
+
+ Returns true if successful
+*/
+static bool
+initial_pass (const struct lr_spec *cmd, struct lr_result *res, struct casereader *input)
+{
+ const int width = var_get_width (cmd->dep_var);
+
+ struct ccase *c;
+ struct casereader *reader;
+
+ double sum;
+ double sumA = 0.0;
+ double sumB = 0.0;
+
+ bool v0set = false;
+ bool v1set = false;
+
+ size_t n_coefficients = cmd->n_predictor_vars;
+ if (cmd->constant)
+ n_coefficients++;
+
+ /* Create categoricals if appropriate */
+ if (cmd->n_cat_predictors > 0)
+ {
+ res->cp.create = frq_create;
+ res->cp.update = frq_update;
+ res->cp.calculate = NULL;
+ res->cp.destroy = frq_destroy;
+
+ res->cats = categoricals_create (cmd->cat_predictors, cmd->n_cat_predictors,
+ cmd->wv, MV_ANY);
+
+ categoricals_set_payload (res->cats, &res->cp, cmd, res);
+ }
+
+ res->cc = 0;
+ for (reader = casereader_clone (input);
+ (c = casereader_read (reader)) != NULL; case_unref (c))
+ {
+ int v;
+ bool missing = false;
+ double weight = dict_get_case_weight (cmd->dict, c, &res->warn_bad_weight);
+ const union value *depval = case_data (c, cmd->dep_var);
+
+ if (var_is_value_missing (cmd->dep_var, depval) & cmd->exclude)
+ {
+ missing = true;
+ }
+ else
+ for (v = 0; v < cmd->n_indep_vars; ++v)
+ {
+ const union value *val = case_data (c, cmd->indep_vars[v]);
+ if (var_is_value_missing (cmd->indep_vars[v], val) & cmd->exclude)
+ {
+ missing = true;
+ break;
+ }
+ }
+
+ /* Accumulate the missing and non-missing counts */
+ if (missing)
+ {
+ res->n_missing++;
+ continue;
+ }
+ res->n_nonmissing++;
+
+ /* Find the values of the dependent variable */
+ if (!v0set)
+ {
+ value_clone (&res->y0, depval, width);
+ v0set = true;
+ }
+ else if (!v1set)
+ {
+ if (!value_equal (&res->y0, depval, width))
+ {
+ value_clone (&res->y1, depval, width);
+ v1set = true;
+ }
+ }
+ else
+ {
+ if (!value_equal (&res->y0, depval, width)
+ &&
+ !value_equal (&res->y1, depval, width)
+ )
+ {
+ msg (ME, _("Dependent variable's values are not dichotomous."));
+ case_unref (c);
+ goto error;
+ }
+ }
+
+ if (v0set && value_equal (&res->y0, depval, width))
+ sumA += weight;
+
+ if (v1set && value_equal (&res->y1, depval, width))
+ sumB += weight;
+
+
+ res->cc += weight;
+
+ categoricals_update (res->cats, c);
+ }
+ casereader_destroy (reader);
+
+ categoricals_done (res->cats);
+
+ sum = sumB;
+
+ /* Ensure that Y0 is less than Y1. Otherwise the mapping gets
+ inverted, which is confusing to users */
+ if (var_is_numeric (cmd->dep_var) && value_compare_3way (&res->y0, &res->y1, width) > 0)
+ {
+ union value tmp;
+ value_clone (&tmp, &res->y0, width);
+ value_copy (&res->y0, &res->y1, width);
+ value_copy (&res->y1, &tmp, width);
+ value_destroy (&tmp, width);
+ sum = sumA;
+ }
+
+ n_coefficients += categoricals_df_total (res->cats);
+ res->beta_hat = gsl_vector_calloc (n_coefficients);
+
+ if (cmd->constant)
+ {
+ double mean = sum / res->cc;
+ gsl_vector_set (res->beta_hat, res->beta_hat->size - 1, log (mean / (1 - mean)));
+ }
+
+ return true;
+
+ error:
+ casereader_destroy (reader);
+ return false;
+}
+
+
+
+/* Start of the logistic regression routine proper */
+static bool
+run_lr (const struct lr_spec *cmd, struct casereader *input,
+ const struct dataset *ds UNUSED)
+{
+ int i;
+
+ bool converged = false;
+
+ /* Set the log likelihoods to a sentinel value */
+ double log_likelihood = SYSMIS;
+ double prev_log_likelihood = SYSMIS;
+ double initial_log_likelihood = SYSMIS;
+
+ struct lr_result work;
+ work.n_missing = 0;
+ work.n_nonmissing = 0;
+ work.warn_bad_weight = true;
+ work.cats = NULL;
+ work.beta_hat = NULL;
+ work.hessian = NULL;
+
+ /* Get the initial estimates of \beta and their standard errors.
+ And perform other auxiliary initialisation. */
+ if (!initial_pass (cmd, &work, input))
+ goto error;
+
+ for (i = 0; i < cmd->n_cat_predictors; ++i)
+ {
+ if (1 >= categoricals_n_count (work.cats, i))
+ {
+ struct string str;
+ ds_init_empty (&str);
+
+ interaction_to_string (cmd->cat_predictors[i], &str);
+
+ msg (ME, _("Category %s does not have at least two distinct values. Logistic regression will not be run."),
+ ds_cstr(&str));
+ ds_destroy (&str);
+ goto error;
+ }
+ }
+
+ output_depvarmap (cmd, &work);
+
+ case_processing_summary (&work);
+
+
+ input = casereader_create_filter_missing (input,
+ cmd->indep_vars,
+ cmd->n_indep_vars,
+ cmd->exclude,
+ NULL,
+ NULL);
+
+ input = casereader_create_filter_missing (input,
+ &cmd->dep_var,
+ 1,
+ cmd->exclude,
+ NULL,
+ NULL);
+
+ work.hessian = gsl_matrix_calloc (work.beta_hat->size, work.beta_hat->size);
+
+ /* Start the Newton Raphson iteration process... */
+ for(i = 0; i < cmd->max_iter; ++i)
+ {
+ double min, max;
+ gsl_vector *v;
+
+
+ hessian (cmd, &work, input,
+ cmd->predictor_vars, cmd->n_predictor_vars,
+ &converged);
+
+ gsl_linalg_cholesky_decomp (work.hessian);
+ gsl_linalg_cholesky_invert (work.hessian);
+
+ v = xt_times_y_pi (cmd, &work, input,
+ cmd->predictor_vars, cmd->n_predictor_vars,
+ cmd->dep_var,
+ &log_likelihood);
+
+ {
+ /* delta = M.v */
+ gsl_vector *delta = gsl_vector_alloc (v->size);
+ gsl_blas_dgemv (CblasNoTrans, 1.0, work.hessian, v, 0, delta);
+ gsl_vector_free (v);
+
+
+ gsl_vector_add (work.beta_hat, delta);
+
+ gsl_vector_minmax (delta, &min, &max);
+
+ if (fabs (min) < cmd->bcon && fabs (max) < cmd->bcon)
+ {
+ msg (MN, _("Estimation terminated at iteration number %d because parameter estimates changed by less than %g"),
+ i + 1, cmd->bcon);
+ converged = true;
+ }
+
+ gsl_vector_free (delta);
+ }
+
+ if (i > 0)
+ {
+ if (-log_likelihood > -(1.0 - cmd->lcon) * prev_log_likelihood)
+ {
+ msg (MN, _("Estimation terminated at iteration number %d because Log Likelihood decreased by less than %g%%"), i + 1, 100 * cmd->lcon);
+ converged = true;
+ }
+ }
+ if (i == 0)
+ initial_log_likelihood = log_likelihood;
+ prev_log_likelihood = log_likelihood;
+
+ if (converged)
+ break;
+ }
+
+
+
+ if (!converged)
+ msg (MW, _("Estimation terminated at iteration number %d because maximum iterations has been reached"), i);
+
+
+ output_model_summary (&work, initial_log_likelihood, log_likelihood);
+
+ if (work.cats)
+ output_categories (cmd, &work);
+
+ output_classification_table (cmd, &work);
+ output_variables (cmd, &work);
+
+ casereader_destroy (input);
+ gsl_matrix_free (work.hessian);
+ gsl_vector_free (work.beta_hat);
+ categoricals_destroy (work.cats);
+
+ return true;
+
+ error:
+ casereader_destroy (input);
+ gsl_matrix_free (work.hessian);
+ gsl_vector_free (work.beta_hat);
+ categoricals_destroy (work.cats);
+
+ return false;
+}
+
+struct variable_node
+{
+ struct hmap_node node; /* Node in hash map. */
+ const struct variable *var; /* The variable */
+};
+
+static struct variable_node *
+lookup_variable (const struct hmap *map, const struct variable *var, unsigned int hash)
+{
+ struct variable_node *vn;
+ HMAP_FOR_EACH_WITH_HASH (vn, struct variable_node, node, hash, map)
+ if (vn->var == var)
+ return vn;
+
+ return NULL;
+}
+
+static void
+insert_variable (struct hmap *map, const struct variable *var, unsigned int hash)
+{
+ if (!lookup_variable (map, var, hash))
+ {
+ struct variable_node *vn = xmalloc (sizeof *vn);
+ *vn = (struct variable_node) { .var = var };
+ hmap_insert (map, &vn->node, hash);
+ }
+}
+
+/* Parse the LOGISTIC REGRESSION command syntax */
+int
+cmd_logistic (struct lexer *lexer, struct dataset *ds)
+{
+ /* Temporary location for the predictor variables.
+ These may or may not include the categorical predictors */
+ const struct variable **pred_vars = NULL;
+ size_t n_pred_vars = 0;
+ double cp = 0.5;
+
+ struct dictionary *dict = dataset_dict (ds);
+ struct lr_spec lr = {
+ .dict = dict,
+ .exclude = MV_ANY,
+ .wv = dict_get_weight (dict),
+ .max_iter = 20,
+ .lcon = 0.0000,
+ .bcon = 0.001,
+ .min_epsilon = 0.00000001,
+ .constant = true,
+ .confidence = 95,
+ .print = PRINT_DEFAULT,
+ };
+
+ if (lex_match_id (lexer, "VARIABLES"))
+ lex_match (lexer, T_EQUALS);
+
+ lr.dep_var = parse_variable_const (lexer, lr.dict);
+ if (!lr.dep_var)
+ goto error;
+
+ if (!lex_force_match (lexer, T_WITH))
+ goto error;
+
+ if (!parse_variables_const (lexer, lr.dict, &pred_vars, &n_pred_vars,
+ PV_NO_DUPLICATE))
+ goto error;
+
+ while (lex_token (lexer) != T_ENDCMD)
+ {
+ lex_match (lexer, T_SLASH);
+
+ if (lex_match_id (lexer, "MISSING"))
+ {
+ lex_match (lexer, T_EQUALS);
+ while (lex_token (lexer) != T_ENDCMD
+ && lex_token (lexer) != T_SLASH)
+ {
+ if (lex_match_id (lexer, "INCLUDE"))
+ lr.exclude = MV_SYSTEM;
+ else if (lex_match_id (lexer, "EXCLUDE"))
+ lr.exclude = MV_ANY;
+ else
+ {
+ lex_error_expecting (lexer, "INCLUDE", "EXCLUDE");
+ goto error;
+ }
+ }
+ }
+ else if (lex_match_id (lexer, "ORIGIN"))
+ lr.constant = false;
+ else if (lex_match_id (lexer, "NOORIGIN"))
+ lr.constant = true;
+ else if (lex_match_id (lexer, "NOCONST"))
+ lr.constant = false;
+ else if (lex_match_id (lexer, "EXTERNAL"))
+ {
+ /* This is for compatibility. It does nothing */
+ }
+ else if (lex_match_id (lexer, "CATEGORICAL"))
+ {
+ lex_match (lexer, T_EQUALS);
+ struct variable **cats;
+ size_t n_cats;
+ if (!parse_variables (lexer, lr.dict, &cats, &n_cats, PV_NO_DUPLICATE))
+ goto error;
+
+ lr.cat_predictors = xrealloc (lr.cat_predictors,
+ sizeof *lr.cat_predictors
+ * (n_cats + lr.n_cat_predictors));
+ for (size_t i = 0; i < n_cats; i++)
+ lr.cat_predictors[lr.n_cat_predictors++] = interaction_create (cats[i]);
+ free (cats);
+ }
+ else if (lex_match_id (lexer, "PRINT"))
+ {
+ lex_match (lexer, T_EQUALS);
+ while (lex_token (lexer) != T_ENDCMD && lex_token (lexer) != T_SLASH)
+ {
+ if (lex_match_id (lexer, "DEFAULT"))
+ lr.print |= PRINT_DEFAULT;
+ else if (lex_match_id (lexer, "SUMMARY"))
+ lr.print |= PRINT_SUMMARY;
+#if 0
+ else if (lex_match_id (lexer, "CORR"))
+ lr.print |= PRINT_CORR;
+ else if (lex_match_id (lexer, "ITER"))
+ lr.print |= PRINT_ITER;
+ else if (lex_match_id (lexer, "GOODFIT"))
+ lr.print |= PRINT_GOODFIT;
+#endif
+ else if (lex_match_id (lexer, "CI"))
+ {
+ lr.print |= PRINT_CI;
+ if (!lex_force_match (lexer, T_LPAREN)
+ || !lex_force_num (lexer))
+ goto error;
+ lr.confidence = lex_number (lexer);
+ lex_get (lexer);
+ if (!lex_force_match (lexer, T_RPAREN))
+ goto error;
+ }
+ else if (lex_match_id (lexer, "ALL"))
+ lr.print = ~0x0000;
+ else
+ {
+ lex_error_expecting (lexer, "DEFAULT", "SUMMARY",
+#if 0
+ "CORR", "ITER", "GOODFIT",
+#endif
+ "CI", "ALL");
+ goto error;
+ }
+ }
+ }
+ else if (lex_match_id (lexer, "CRITERIA"))
+ {
+ lex_match (lexer, T_EQUALS);
+ while (lex_token (lexer) != T_ENDCMD && lex_token (lexer) != T_SLASH)
+ {
+ if (lex_match_id (lexer, "BCON"))
+ {
+ if (!lex_force_match (lexer, T_LPAREN)
+ || !lex_force_num (lexer))
+ goto error;
+ lr.bcon = lex_number (lexer);
+ lex_get (lexer);
+ if (!lex_force_match (lexer, T_RPAREN))
+ goto error;
+ }
+ else if (lex_match_id (lexer, "ITERATE"))
+ {
+ if (!lex_force_match (lexer, T_LPAREN)
+ || !lex_force_int_range (lexer, "ITERATE", 0, INT_MAX))
+ goto error;
+ lr.max_iter = lex_integer (lexer);
+ lex_get (lexer);
+ if (!lex_force_match (lexer, T_RPAREN))
+ goto error;
+ }
+ else if (lex_match_id (lexer, "LCON"))
+ {
+ if (!lex_force_match (lexer, T_LPAREN)
+ || !lex_force_num (lexer))
+ goto error;
+ lr.lcon = lex_number (lexer);
+ lex_get (lexer);
+ if (!lex_force_match (lexer, T_RPAREN))
+ goto error;
+ }
+ else if (lex_match_id (lexer, "EPS"))
+ {
+ if (!lex_force_match (lexer, T_LPAREN)
+ || !lex_force_num (lexer))
+ goto error;
+ lr.min_epsilon = lex_number (lexer);
+ lex_get (lexer);
+ if (!lex_force_match (lexer, T_RPAREN))
+ goto error;
+ }
+ else if (lex_match_id (lexer, "CUT"))
+ {
+ if (!lex_force_match (lexer, T_LPAREN)
+ || !lex_force_num_range_closed (lexer, "CUT", 0, 1))
+ goto error;
+
+ cp = lex_number (lexer);
+
+ lex_get (lexer);
+ if (!lex_force_match (lexer, T_RPAREN))
+ goto error;
+ }
+ else
+ {
+ lex_error_expecting (lexer, "BCON", "ITERATE", "LCON", "EPS",
+ "CUT");
+ goto error;
+ }
+ }
+ }
+ else
+ {
+ lex_error_expecting (lexer, "MISSING", "ORIGIN", "NOORIGIN",
+ "NOCONST", "EXTERNAL", "CATEGORICAL",
+ "PRINT", "CRITERIA");
+ goto error;
+ }
+ }
+
+ lr.ilogit_cut_point = - log (1/cp - 1);
+
+ /* Copy the predictor variables from the temporary location into the
+ final one, dropping any categorical variables which appear there.
+ FIXME: This is O(NxM).
+ */
+ struct hmap allvars = HMAP_INITIALIZER (allvars);
+ size_t allocated_predictor_vars = 0;
+ for (size_t v = 0; v < n_pred_vars; ++v)
+ {
+ bool drop = false;
+ const struct variable *var = pred_vars[v];
+
+ unsigned int hash = hash_pointer (var, 0);
+ insert_variable (&allvars, var, hash);
+
+ for (size_t cv = 0; cv < lr.n_cat_predictors; ++cv)
+ {
+ const struct interaction *iact = lr.cat_predictors[cv];
+ for (size_t iv = 0; iv < iact->n_vars; ++iv)
+ {
+ const struct variable *ivar = iact->vars[iv];
+ unsigned int hash = hash_pointer (ivar, 0);
+ insert_variable (&allvars, ivar, hash);
+
+ if (var == ivar)
+ drop = true;
+ }
+ }
+
+ if (drop)
+ continue;
+
+ if (lr.n_predictor_vars >= allocated_predictor_vars)
+ lr.predictor_vars = x2nrealloc (lr.predictor_vars,
+ &allocated_predictor_vars,
+ sizeof *lr.predictor_vars);
+ lr.predictor_vars[lr.n_predictor_vars++] = var;
+ }
+
+ lr.n_indep_vars = hmap_count (&allvars);
+ lr.indep_vars = xmalloc (lr.n_indep_vars * sizeof *lr.indep_vars);
+
+ /* Interate over each variable and push it into the array */
+ size_t x = 0;
+ struct variable_node *vn, *next;
+ HMAP_FOR_EACH_SAFE (vn, next, struct variable_node, node, &allvars)
+ {
+ lr.indep_vars[x++] = vn->var;
+ hmap_delete (&allvars, &vn->node);
+ free (vn);
+ }
+ assert (x == lr.n_indep_vars);
+ hmap_destroy (&allvars);
+
+ /* Run logistical regression for each split group. */
+ struct casegrouper *grouper = casegrouper_create_splits (proc_open (ds), lr.dict);
+ struct casereader *group;
+ bool ok = true;
+ while (casegrouper_get_next_group (grouper, &group))
+ ok = run_lr (&lr, group, ds) && ok;
+ ok = casegrouper_destroy (grouper) && ok;
+ ok = proc_commit (ds) && ok;
+
+ for (size_t i = 0; i < lr.n_cat_predictors; ++i)
+ interaction_destroy (lr.cat_predictors[i]);
+ free (lr.predictor_vars);
+ free (lr.cat_predictors);
+ free (lr.indep_vars);
+ free (pred_vars);
+
+ return CMD_SUCCESS;
+
+ error:
+ for (size_t i = 0; i < lr.n_cat_predictors; ++i)
+ interaction_destroy (lr.cat_predictors[i]);
+ free (lr.predictor_vars);
+ free (lr.cat_predictors);
+ free (lr.indep_vars);
+ free (pred_vars);
+
+ return CMD_FAILURE;
+}
+
+
+\f
+
+/* Show the Dependent Variable Encoding box.
+ This indicates how the dependent variable
+ is mapped to the internal zero/one values.
+*/
+static void
+output_depvarmap (const struct lr_spec *cmd, const struct lr_result *res)
+{
+ struct pivot_table *table = pivot_table_create (
+ N_("Dependent Variable Encoding"));
+
+ pivot_dimension_create (table, PIVOT_AXIS_COLUMN, N_("Mapping"),
+ N_("Internal Value"));
+
+ struct pivot_dimension *original = pivot_dimension_create (
+ table, PIVOT_AXIS_ROW, N_("Original Value"));
+ original->root->show_label = true;
+
+ for (int i = 0; i < 2; i++)
+ {
+ const union value *v = i ? &res->y1 : &res->y0;
+ int orig_idx = pivot_category_create_leaf (
+ original->root, pivot_value_new_var_value (cmd->dep_var, v));
+ pivot_table_put2 (table, 0, orig_idx, pivot_value_new_number (
+ map_dependent_var (cmd, res, v)));
+ }
+
+ pivot_table_submit (table);
+}
+
+
+/* Show the Variables in the Equation box */
+static void
+output_variables (const struct lr_spec *cmd,
+ const struct lr_result *res)
+{
+ struct pivot_table *table = pivot_table_create (
+ N_("Variables in the Equation"));
+
+ struct pivot_dimension *statistics = pivot_dimension_create (
+ table, PIVOT_AXIS_COLUMN, N_("Statistics"),
+ N_("B"), PIVOT_RC_OTHER,
+ N_("S.E."), PIVOT_RC_OTHER,
+ N_("Wald"), PIVOT_RC_OTHER,
+ N_("df"), PIVOT_RC_INTEGER,
+ N_("Sig."), PIVOT_RC_SIGNIFICANCE,
+ N_("Exp(B)"), PIVOT_RC_OTHER);
+ if (cmd->print & PRINT_CI)
+ {
+ struct pivot_category *group = pivot_category_create_group__ (
+ statistics->root,
+ pivot_value_new_text_format (N_("%d%% CI for Exp(B)"),
+ cmd->confidence));
+ pivot_category_create_leaves (group, N_("Lower"), N_("Upper"));
+ }
+
+ struct pivot_dimension *variables = pivot_dimension_create (
+ table, PIVOT_AXIS_ROW, N_("Variables"));
+ struct pivot_category *step1 = pivot_category_create_group (
+ variables->root, N_("Step 1"));
+
+ int ivar = 0;
+ int idx_correction = 0;
+ int i = 0;
+
+ int nr = cmd->n_predictor_vars;
+ if (cmd->constant)
+ nr++;
+ if (res->cats)
+ nr += categoricals_df_total (res->cats) + cmd->n_cat_predictors;
+
+ for (int row = 0; row < nr; row++)
+ {
+ const int idx = row - idx_correction;
+
+ int var_idx;
+ if (idx < cmd->n_predictor_vars)
+ var_idx = pivot_category_create_leaf (
+ step1, pivot_value_new_variable (cmd->predictor_vars[idx]));
+ else if (i < cmd->n_cat_predictors)
+ {
+ const struct interaction *cat_predictors = cmd->cat_predictors[i];
+ struct string str = DS_EMPTY_INITIALIZER;
+ interaction_to_string (cat_predictors, &str);
+ if (ivar != 0)
+ ds_put_format (&str, "(%d)", ivar);
+ var_idx = pivot_category_create_leaf (
+ step1, pivot_value_new_user_text_nocopy (ds_steal_cstr (&str)));
+
+ int df = categoricals_df (res->cats, i);
+ bool summary = ivar == 0;
+ if (summary)
+ {
+ /* Calculate the Wald statistic,
+ which is \beta' C^-1 \beta .
+ where \beta is the vector of the coefficient estimates comprising this
+ categorial variable. and C is the corresponding submatrix of the
+ hessian matrix.
+ */
+ gsl_matrix_const_view mv =
+ gsl_matrix_const_submatrix (res->hessian, idx, idx, df, df);
+ gsl_matrix *subhessian = gsl_matrix_alloc (mv.matrix.size1, mv.matrix.size2);
+ gsl_vector_const_view vv = gsl_vector_const_subvector (res->beta_hat, idx, df);
+ gsl_vector *temp = gsl_vector_alloc (df);
+
+ gsl_matrix_memcpy (subhessian, &mv.matrix);
+ gsl_linalg_cholesky_decomp (subhessian);
+ gsl_linalg_cholesky_invert (subhessian);
+
+ gsl_blas_dgemv (CblasTrans, 1.0, subhessian, &vv.vector, 0, temp);
+ double wald;
+ gsl_blas_ddot (temp, &vv.vector, &wald);
+
+ double entries[] = { wald, df, gsl_cdf_chisq_Q (wald, df) };
+ for (size_t j = 0; j < sizeof entries / sizeof *entries; j++)
+ pivot_table_put2 (table, j + 2, var_idx,
+ pivot_value_new_number (entries[j]));
+
+ idx_correction++;
+ gsl_matrix_free (subhessian);
+ gsl_vector_free (temp);
+ }
+
+ if (ivar++ == df)
+ {
+ ++i; /* next interaction */
+ ivar = 0;
+ }
+
+ if (summary)
+ continue;
+ }
+ else
+ var_idx = pivot_category_create_leaves (step1, N_("Constant"));
+
+ double b = gsl_vector_get (res->beta_hat, idx);
+ double sigma2 = gsl_matrix_get (res->hessian, idx, idx);
+ double wald = pow2 (b) / sigma2;
+ double df = 1;
+ double wc = (gsl_cdf_ugaussian_Pinv (0.5 + cmd->confidence / 200.0)
+ * sqrt (sigma2));
+ bool show_ci = cmd->print & PRINT_CI && row < nr - cmd->constant;
+
+ double entries[] = {
+ b,
+ sqrt (sigma2),
+ wald,
+ df,
+ gsl_cdf_chisq_Q (wald, df),
+ exp (b),
+ show_ci ? exp (b - wc) : SYSMIS,
+ show_ci ? exp (b + wc) : SYSMIS,
+ };
+ for (size_t j = 0; j < sizeof entries / sizeof *entries; j++)
+ if (entries[j] != SYSMIS)
+ pivot_table_put2 (table, j, var_idx,
+ pivot_value_new_number (entries[j]));
+ }
+
+ pivot_table_submit (table);
+}
+
+
+/* Show the model summary box */
+static void
+output_model_summary (const struct lr_result *res,
+ double initial_log_likelihood, double log_likelihood)
+{
+ struct pivot_table *table = pivot_table_create (N_("Model Summary"));
+
+ pivot_dimension_create (table, PIVOT_AXIS_COLUMN, N_("Statistics"),
+ N_("-2 Log likelihood"), PIVOT_RC_OTHER,
+ N_("Cox & Snell R Square"), PIVOT_RC_OTHER,
+ N_("Nagelkerke R Square"), PIVOT_RC_OTHER);
+
+ struct pivot_dimension *step = pivot_dimension_create (
+ table, PIVOT_AXIS_ROW, N_("Step"));
+ step->root->show_label = true;
+ pivot_category_create_leaf (step->root, pivot_value_new_integer (1));
+
+ double cox = (1.0 - exp ((initial_log_likelihood - log_likelihood)
+ * (2 / res->cc)));
+ double entries[] = {
+ -2 * log_likelihood,
+ cox,
+ cox / (1.0 - exp(initial_log_likelihood * (2 / res->cc)))
+ };
+ for (size_t i = 0; i < sizeof entries / sizeof *entries; i++)
+ pivot_table_put2 (table, i, 0, pivot_value_new_number (entries[i]));
+
+ pivot_table_submit (table);
+}
+
+/* Show the case processing summary box */
+static void
+case_processing_summary (const struct lr_result *res)
+{
+ struct pivot_table *table = pivot_table_create (
+ N_("Case Processing Summary"));
+
+ pivot_dimension_create (table, PIVOT_AXIS_COLUMN, N_("Statistics"),
+ N_("N"), PIVOT_RC_COUNT,
+ N_("Percent"), PIVOT_RC_PERCENT);
+
+ struct pivot_dimension *cases = pivot_dimension_create (
+ table, PIVOT_AXIS_ROW, N_("Unweighted Cases"),
+ N_("Included in Analysis"), N_("Missing Cases"), N_("Total"));
+ cases->root->show_label = true;
+
+ double total = res->n_nonmissing + res->n_missing;
+ struct entry
+ {
+ int stat_idx;
+ int case_idx;
+ double x;
+ }
+ entries[] = {
+ { 0, 0, res->n_nonmissing },
+ { 0, 1, res->n_missing },
+ { 0, 2, total },
+ { 1, 0, 100.0 * res->n_nonmissing / total },
+ { 1, 1, 100.0 * res->n_missing / total },
+ { 1, 2, 100.0 },
+ };
+ for (size_t i = 0; i < sizeof entries / sizeof *entries; i++)
+ pivot_table_put2 (table, entries[i].stat_idx, entries[i].case_idx,
+ pivot_value_new_number (entries[i].x));
+
+ pivot_table_submit (table);
+}
+
+static void
+output_categories (const struct lr_spec *cmd, const struct lr_result *res)
+{
+ struct pivot_table *table = pivot_table_create (
+ N_("Categorical Variables' Codings"));
+ pivot_table_set_weight_var (table, dict_get_weight (cmd->dict));
+
+ int max_df = 0;
+ int total_cats = 0;
+ for (int i = 0; i < cmd->n_cat_predictors; ++i)
+ {
+ size_t n = categoricals_n_count (res->cats, i);
+ size_t df = categoricals_df (res->cats, i);
+ if (max_df < df)
+ max_df = df;
+ total_cats += n;
+ }
+
+ struct pivot_dimension *codings = pivot_dimension_create (
+ table, PIVOT_AXIS_COLUMN, N_("Codings"),
+ N_("Frequency"), PIVOT_RC_COUNT);
+ struct pivot_category *coding_group = pivot_category_create_group (
+ codings->root, N_("Parameter coding"));
+ for (int i = 0; i < max_df; ++i)
+ pivot_category_create_leaf_rc (
+ coding_group,
+ pivot_value_new_user_text_nocopy (xasprintf ("(%d)", i + 1)),
+ PIVOT_RC_INTEGER);
+
+ struct pivot_dimension *categories = pivot_dimension_create (
+ table, PIVOT_AXIS_ROW, N_("Categories"));
+
+ int cumulative_df = 0;
+ for (int v = 0; v < cmd->n_cat_predictors; ++v)
+ {
+ int cat;
+ const struct interaction *cat_predictors = cmd->cat_predictors[v];
+ int df = categoricals_df (res->cats, v);
+
+ struct string str = DS_EMPTY_INITIALIZER;
+ interaction_to_string (cat_predictors, &str);
+ struct pivot_category *var_group = pivot_category_create_group__ (
+ categories->root,
+ pivot_value_new_user_text_nocopy (ds_steal_cstr (&str)));
+
+ for (cat = 0; cat < categoricals_n_count (res->cats, v); ++cat)
+ {
+ const struct ccase *c = categoricals_get_case_by_category_real (
+ res->cats, v, cat);
+ struct string label = DS_EMPTY_INITIALIZER;
+ for (int x = 0; x < cat_predictors->n_vars; ++x)
+ {
+ if (!ds_is_empty (&label))
+ ds_put_byte (&label, ' ');
+
+ const union value *val = case_data (c, cat_predictors->vars[x]);
+ var_append_value_name (cat_predictors->vars[x], val, &label);
+ }
+ int cat_idx = pivot_category_create_leaf (
+ var_group,
+ pivot_value_new_user_text_nocopy (ds_steal_cstr (&label)));
+
+ double *freq = categoricals_get_user_data_by_category_real (
+ res->cats, v, cat);
+ pivot_table_put2 (table, 0, cat_idx, pivot_value_new_number (*freq));
+
+ for (int x = 0; x < df; ++x)
+ pivot_table_put2 (table, x + 1, cat_idx,
+ pivot_value_new_number (cat == x));
+ }
+ cumulative_df += df;
+ }
+
+ pivot_table_submit (table);
+}
+
+static void
+create_classification_dimension (const struct lr_spec *cmd,
+ const struct lr_result *res,
+ struct pivot_table *table,
+ enum pivot_axis_type axis_type,
+ const char *label, const char *total)
+{
+ struct pivot_dimension *d = pivot_dimension_create (
+ table, axis_type, label);
+ d->root->show_label = true;
+ struct pivot_category *pred_group = pivot_category_create_group__ (
+ d->root, pivot_value_new_variable (cmd->dep_var));
+ for (int i = 0; i < 2; i++)
+ {
+ const union value *y = i ? &res->y1 : &res->y0;
+ pivot_category_create_leaf_rc (
+ pred_group, pivot_value_new_var_value (cmd->dep_var, y),
+ PIVOT_RC_COUNT);
+ }
+ pivot_category_create_leaves (d->root, total, PIVOT_RC_PERCENT);
+}
+
+static void
+output_classification_table (const struct lr_spec *cmd, const struct lr_result *res)
+{
+ struct pivot_table *table = pivot_table_create (N_("Classification Table"));
+ pivot_table_set_weight_var (table, cmd->wv);
+
+ create_classification_dimension (cmd, res, table, PIVOT_AXIS_COLUMN,
+ N_("Predicted"), N_("Percentage Correct"));
+ create_classification_dimension (cmd, res, table, PIVOT_AXIS_ROW,
+ N_("Observed"), N_("Overall Percentage"));
+
+ pivot_dimension_create (table, PIVOT_AXIS_ROW, N_("Step"), N_("Step 1"));
+
+ struct entry
+ {
+ int pred_idx;
+ int obs_idx;
+ double x;
+ }
+ entries[] = {
+ { 0, 0, res->tn },
+ { 0, 1, res->fn },
+ { 1, 0, res->fp },
+ { 1, 1, res->tp },
+ { 2, 0, 100 * res->tn / (res->tn + res->fp) },
+ { 2, 1, 100 * res->tp / (res->tp + res->fn) },
+ { 2, 2,
+ 100 * (res->tp + res->tn) / (res->tp + res->tn + res->fp + res->fn)},
+ };
+ for (size_t i = 0; i < sizeof entries / sizeof *entries; i++)
+ {
+ const struct entry *e = &entries[i];
+ pivot_table_put3 (table, e->pred_idx, e->obs_idx, 0,
+ pivot_value_new_number (e->x));
+ }
+
+ pivot_table_submit (table);
+}
projects / pspp / blobdiff