X-Git-Url: https://pintos-os.org/cgi-bin/gitweb.cgi?a=blobdiff_plain;f=src%2Flanguage%2Fstats%2Fdescriptives.c;h=50d52d3cff03f68d2bac2d9d0a84097b961ae032;hb=81579d9e9f994fb2908f50af41c3eb033d216e58;hp=4b1ea6e2fa8bc0974fd8252b1bc233cc05597665;hpb=f5c108becd49d78f4898cab11352291f5689d24e;p=pspp-builds.git diff --git a/src/language/stats/descriptives.c b/src/language/stats/descriptives.c index 4b1ea6e2..50d52d3c 100644 --- a/src/language/stats/descriptives.c +++ b/src/language/stats/descriptives.c @@ -1,20 +1,18 @@ -/* PSPP - computes sample statistics. - Copyright (C) 1997-9, 2000 Free Software Foundation, Inc. +/* PSPP - a program for statistical analysis. + Copyright (C) 1997-9, 2000, 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 the Free Software Foundation; either version 2 of the - License, or (at your option) any later version. + 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. + 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, write to the Free Software - Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA - 02110-1301, USA. */ + along with this program. If not, see . */ #include @@ -22,25 +20,24 @@ #include #include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include +#include "data/casegrouper.h" +#include "data/casereader.h" +#include "data/dictionary.h" +#include "data/procedure.h" +#include "data/transformations.h" +#include "data/variable.h" +#include "language/command.h" +#include "language/dictionary/split-file.h" +#include "language/lexer/lexer.h" +#include "language/lexer/variable-parser.h" +#include "libpspp/array.h" +#include "libpspp/compiler.h" +#include "libpspp/message.h" +#include "libpspp/assertion.h" +#include "math/moments.h" +#include "output/tab.h" + +#include "gl/xalloc.h" #include "gettext.h" #define _(msgid) gettext (msgid) @@ -102,13 +99,13 @@ struct dsc_statistic_info static const struct dsc_statistic_info dsc_info[DSC_N_STATS] = { {"MEAN", N_("Mean"), MOMENT_MEAN}, - {"SEMEAN", N_("S E Mean"), MOMENT_VARIANCE}, + {"SEMEAN", N_("S.E. Mean"), MOMENT_VARIANCE}, {"STDDEV", N_("Std Dev"), MOMENT_VARIANCE}, {"VARIANCE", N_("Variance"), MOMENT_VARIANCE}, {"KURTOSIS", N_("Kurtosis"), MOMENT_KURTOSIS}, - {"SEKURTOSIS", N_("S E Kurt"), MOMENT_NONE}, + {"SEKURTOSIS", N_("S.E. Kurt"), MOMENT_NONE}, {"SKEWNESS", N_("Skewness"), MOMENT_SKEWNESS}, - {"SESKEWNESS", N_("S E Skew"), MOMENT_NONE}, + {"SESKEWNESS", N_("S.E. Skew"), MOMENT_NONE}, {"RANGE", N_("Range"), MOMENT_NONE}, {"MINIMUM", N_("Minimum"), MOMENT_NONE}, {"MAXIMUM", N_("Maximum"), MOMENT_NONE}, @@ -125,7 +122,7 @@ static const struct dsc_statistic_info dsc_info[DSC_N_STATS] = struct dsc_var { const struct variable *v; /* Variable to calculate on. */ - char z_name[LONG_NAME_LEN + 1]; /* Name for z-score variable. */ + char *z_name; /* Name for z-score variable. */ double valid, missing; /* Valid, missing counts. */ struct moments *moments; /* Moments. */ double min, max; /* Maximum and mimimum values. */ @@ -171,9 +168,9 @@ static void free_dsc_proc (struct dsc_proc *); /* Z-score functions. */ static bool try_name (const struct dictionary *dict, struct dsc_proc *dsc, const char *name); -static bool generate_z_varname (const struct dictionary *dict, - struct dsc_proc *dsc, char *z_name, - const char *name, int *z_cnt); +static char *generate_z_varname (const struct dictionary *dict, + struct dsc_proc *dsc, + const char *name, int *z_cnt); static void dump_z_table (struct dsc_proc *); static void setup_z_trns (struct dsc_proc *, struct dataset *); @@ -217,12 +214,12 @@ cmd_descriptives (struct lexer *lexer, struct dataset *ds) dsc->show_stats = dsc->calc_stats = DEFAULT_STATS; /* Parse DESCRIPTIVES. */ - while (lex_token (lexer) != '.') + while (lex_token (lexer) != T_ENDCMD) { if (lex_match_id (lexer, "MISSING")) { - lex_match (lexer, '='); - while (lex_token (lexer) != '.' && lex_token (lexer) != '/') + lex_match (lexer, T_EQUALS); + while (lex_token (lexer) != T_ENDCMD && lex_token (lexer) != T_SLASH) { if (lex_match_id (lexer, "VARIABLE")) dsc->missing_type = DSC_VARIABLE; @@ -235,15 +232,15 @@ cmd_descriptives (struct lexer *lexer, struct dataset *ds) lex_error (lexer, NULL); goto error; } - lex_match (lexer, ','); + lex_match (lexer, T_COMMA); } } else if (lex_match_id (lexer, "SAVE")) save_z_scores = 1; else if (lex_match_id (lexer, "FORMAT")) { - lex_match (lexer, '='); - while (lex_token (lexer) != '.' && lex_token (lexer) != '/') + lex_match (lexer, T_EQUALS); + while (lex_token (lexer) != T_ENDCMD && lex_token (lexer) != T_SLASH) { if (lex_match_id (lexer, "LABELS")) dsc->show_var_labels = 1; @@ -262,14 +259,14 @@ cmd_descriptives (struct lexer *lexer, struct dataset *ds) lex_error (lexer, NULL); goto error; } - lex_match (lexer, ','); + lex_match (lexer, T_COMMA); } } else if (lex_match_id (lexer, "STATISTICS")) { - lex_match (lexer, '='); + lex_match (lexer, T_EQUALS); dsc->show_stats = 0; - while (lex_token (lexer) != '.' && lex_token (lexer) != '/') + while (lex_token (lexer) != T_ENDCMD && lex_token (lexer) != T_SLASH) { if (lex_match (lexer, T_ALL)) dsc->show_stats |= (1ul << DSC_N_STATS) - 1; @@ -277,14 +274,14 @@ cmd_descriptives (struct lexer *lexer, struct dataset *ds) dsc->show_stats |= DEFAULT_STATS; else dsc->show_stats |= 1ul << (match_statistic (lexer)); - lex_match (lexer, ','); + lex_match (lexer, T_COMMA); } if (dsc->show_stats == 0) dsc->show_stats = DEFAULT_STATS; } else if (lex_match_id (lexer, "SORT")) { - lex_match (lexer, '='); + lex_match (lexer, T_EQUALS); if (lex_match_id (lexer, "NAME")) dsc->sort_by_stat = DSC_NAME; else @@ -293,7 +290,7 @@ cmd_descriptives (struct lexer *lexer, struct dataset *ds) if (dsc->sort_by_stat == DSC_NONE ) dsc->sort_by_stat = DSC_MEAN; } - if (lex_match (lexer, '(')) + if (lex_match (lexer, T_LPAREN)) { if (lex_match_id (lexer, "A")) dsc->sort_ascending = 1; @@ -301,18 +298,18 @@ cmd_descriptives (struct lexer *lexer, struct dataset *ds) dsc->sort_ascending = 0; else lex_error (lexer, NULL); - lex_force_match (lexer, ')'); + lex_force_match (lexer, T_RPAREN); } } else if (var_cnt == 0) { - if (lex_look_ahead (lexer) == '=') + if (lex_look_ahead (lexer) == T_EQUALS) { lex_match_id (lexer, "VARIABLES"); - lex_match (lexer, '='); + lex_match (lexer, T_EQUALS); } - while (lex_token (lexer) != '.' && lex_token (lexer) != '/') + while (lex_token (lexer) != T_ENDCMD && lex_token (lexer) != T_SLASH) { int i; @@ -325,28 +322,29 @@ cmd_descriptives (struct lexer *lexer, struct dataset *ds) { struct dsc_var *dv = &dsc->vars[i]; dv->v = vars[i]; - dv->z_name[0] = '\0'; + dv->z_name = NULL; dv->moments = NULL; } dsc->var_cnt = var_cnt; - if (lex_match (lexer, '(')) + if (lex_match (lexer, T_LPAREN)) { if (lex_token (lexer) != T_ID) { lex_error (lexer, NULL); goto error; } - if (try_name (dict, dsc, lex_tokid (lexer))) + if (try_name (dict, dsc, lex_tokcstr (lexer))) { - strcpy (dsc->vars[dsc->var_cnt - 1].z_name, lex_tokid (lexer)); + struct dsc_var *dsc_var = &dsc->vars[dsc->var_cnt - 1]; + dsc_var->z_name = xstrdup (lex_tokcstr (lexer)); z_cnt++; } else msg (SE, _("Z-score variable name %s would be" - " a duplicate variable name."), lex_tokid (lexer)); + " a duplicate variable name."), lex_tokcstr (lexer)); lex_get (lexer); - if (!lex_force_match (lexer, ')')) + if (!lex_force_match (lexer, T_RPAREN)) goto error; } } @@ -357,7 +355,7 @@ cmd_descriptives (struct lexer *lexer, struct dataset *ds) goto error; } - lex_match (lexer, '/'); + lex_match (lexer, T_SLASH); } if (var_cnt == 0) { @@ -373,14 +371,19 @@ cmd_descriptives (struct lexer *lexer, struct dataset *ds) int gen_cnt = 0; for (i = 0; i < dsc->var_cnt; i++) - if (dsc->vars[i].z_name[0] == 0) - { - if (!generate_z_varname (dict, dsc, dsc->vars[i].z_name, - var_get_name (dsc->vars[i].v), - &gen_cnt)) - goto error; - z_cnt++; - } + { + struct dsc_var *dsc_var = &dsc->vars[i]; + if (dsc_var->z_name == NULL) + { + const char *name = var_get_name (dsc_var->v); + dsc_var->z_name = generate_z_varname (dict, dsc, name, + &gen_cnt); + if (dsc_var->z_name == NULL) + goto error; + + z_cnt++; + } + } } dump_z_table (dsc); } @@ -412,7 +415,7 @@ cmd_descriptives (struct lexer *lexer, struct dataset *ds) for (i = 0; i < dsc->var_cnt; i++) dsc->vars[i].moments = moments_create (dsc->max_moment); - /* Data pass. FIXME: error handling. */ + /* Data pass. */ grouper = casegrouper_create_splits (proc_open (ds), dict); while (casegrouper_get_next_group (grouper, &group)) calc_descriptives (dsc, group, ds); @@ -466,7 +469,11 @@ free_dsc_proc (struct dsc_proc *dsc) return; for (i = 0; i < dsc->var_cnt; i++) - moments_destroy (dsc->vars[i].moments); + { + struct dsc_var *dsc_var = &dsc->vars[i]; + free (dsc_var->z_name); + moments_destroy (dsc_var->moments); + } free (dsc->vars); free (dsc); } @@ -484,29 +491,29 @@ try_name (const struct dictionary *dict, struct dsc_proc *dsc, if (dict_lookup_var (dict, name) != NULL) return false; for (i = 0; i < dsc->var_cnt; i++) - if (!strcasecmp (dsc->vars[i].z_name, name)) - return false; + { + struct dsc_var *dsc_var = &dsc->vars[i]; + if (dsc_var->z_name != NULL && !strcasecmp (dsc_var->z_name, name)) + return false; + } return true; } /* Generates a name for a Z-score variable based on a variable named VAR_NAME, given that *Z_CNT generated variable names are - known to already exist. If successful, returns true and - copies the new name into Z_NAME. On failure, returns false. */ -static bool -generate_z_varname (const struct dictionary *dict, struct dsc_proc *dsc, char *z_name, + known to already exist. If successful, returns the new name + as a dynamically allocated string. On failure, returns NULL. */ +static char * +generate_z_varname (const struct dictionary *dict, struct dsc_proc *dsc, const char *var_name, int *z_cnt) { - char name[LONG_NAME_LEN + 1]; + char name[VAR_NAME_LEN + 1]; /* Try a name based on the original variable name. */ name[0] = 'Z'; str_copy_trunc (name + 1, sizeof name - 1, var_name); if (try_name (dict, dsc, name)) - { - strcpy (z_name, name); - return true; - } + return xstrdup (name); /* Generate a synthetic name. */ for (;;) @@ -526,14 +533,11 @@ generate_z_varname (const struct dictionary *dict, struct dsc_proc *dsc, char *z msg (SE, _("Ran out of generic names for Z-score variables. " "There are only 126 generic names: ZSC001-ZSC0999, " "STDZ01-STDZ09, ZZZZ01-ZZZZ09, ZQZQ01-ZQZQ09.")); - return false; + return NULL; } if (try_name (dict, dsc, name)) - { - strcpy (z_name, name); - return true; - } + return xstrdup (name); } NOT_REACHED(); } @@ -550,25 +554,23 @@ dump_z_table (struct dsc_proc *dsc) size_t i; for (i = 0; i < dsc->var_cnt; i++) - if (dsc->vars[i].z_name[0] != '\0') + if (dsc->vars[i].z_name != NULL) cnt++; } - t = tab_create (2, cnt + 1, 0); + t = tab_create (2, cnt + 1); tab_title (t, _("Mapping of variables to corresponding Z-scores.")); - tab_columns (t, SOM_COL_DOWN, 1); tab_headers (t, 0, 0, 1, 0); tab_box (t, TAL_1, TAL_1, TAL_0, TAL_1, 0, 0, 1, cnt); tab_hline (t, TAL_2, 0, 1, 1); tab_text (t, 0, 0, TAB_CENTER | TAT_TITLE, _("Source")); tab_text (t, 1, 0, TAB_CENTER | TAT_TITLE, _("Target")); - tab_dim (t, tab_natural_dimensions); { size_t i, y; for (i = 0, y = 1; i < dsc->var_cnt; i++) - if (dsc->vars[i].z_name[0] != '\0') + if (dsc->vars[i].z_name != NULL) { tab_text (t, 0, y, TAB_LEFT, var_get_name (dsc->vars[i].v)); tab_text (t, 1, y++, TAB_LEFT, dsc->vars[i].z_name); @@ -585,7 +587,7 @@ dump_z_table (struct dsc_proc *dsc) (either system or user-missing values that weren't included). */ static int -descriptives_trns_proc (void *trns_, struct ccase * c, +descriptives_trns_proc (void *trns_, struct ccase **c, casenumber case_idx UNUSED) { struct dsc_trns *t = trns_; @@ -598,7 +600,7 @@ descriptives_trns_proc (void *trns_, struct ccase * c, assert(t->vars); for (vars = t->vars; vars < t->vars + t->var_cnt; vars++) { - double score = case_num (c, *vars); + double score = case_num (*c, *vars); if (var_is_num_missing (*vars, score, t->exclude)) { all_sysmis = 1; @@ -607,10 +609,11 @@ descriptives_trns_proc (void *trns_, struct ccase * c, } } + *c = case_unshare (*c); for (z = t->z_scores; z < t->z_scores + t->z_score_cnt; z++) { - double input = case_num (c, z->src_var); - double *output = &case_data_rw (c, z->z_var)->f; + double input = case_num (*c, z->src_var); + double *output = &case_data_rw (*c, z->z_var)->f; if (z->mean == SYSMIS || z->std_dev == SYSMIS || all_sysmis || var_is_num_missing (z->src_var, input, t->exclude)) @@ -641,7 +644,7 @@ setup_z_trns (struct dsc_proc *dsc, struct dataset *ds) size_t cnt, i; for (cnt = i = 0; i < dsc->var_cnt; i++) - if (dsc->vars[i].z_name[0] != '\0') + if (dsc->vars[i].z_name != NULL) cnt++; t = xmalloc (sizeof *t); @@ -665,7 +668,7 @@ setup_z_trns (struct dsc_proc *dsc, struct dataset *ds) for (cnt = i = 0; i < dsc->var_cnt; i++) { struct dsc_var *dv = &dsc->vars[i]; - if (dv->z_name[0] != '\0') + if (dv->z_name != NULL) { struct dsc_z_score *z; struct variable *dst_var; @@ -697,20 +700,23 @@ calc_descriptives (struct dsc_proc *dsc, struct casereader *group, struct dataset *ds) { struct casereader *pass1, *pass2; - struct ccase c; + struct ccase *c; size_t i; - if (!casereader_peek (group, 0, &c)) - return; - output_split_file_values (ds, &c); - case_destroy (&c); + c = casereader_peek (group, 0); + if (c == NULL) + { + casereader_destroy (group); + return; + } + output_split_file_values (ds, c); + case_unref (c); group = casereader_create_filter_weight (group, dataset_dict (ds), NULL, NULL); - casereader_split (group, &pass1, &pass2); - if (dsc->max_moment <= MOMENT_MEAN) - casereader_destroy (pass2); + pass1 = group; + pass2 = dsc->max_moment <= MOMENT_MEAN ? NULL : casereader_clone (pass1); for (i = 0; i < dsc->var_cnt; i++) { @@ -726,12 +732,12 @@ calc_descriptives (struct dsc_proc *dsc, struct casereader *group, dsc->valid = 0.; /* First pass to handle most of the work. */ - for (; casereader_read (pass1, &c); case_destroy (&c)) + for (; (c = casereader_read (pass1)) != NULL; case_unref (c)) { - double weight = dict_get_case_weight (dataset_dict (ds), &c, NULL); + double weight = dict_get_case_weight (dataset_dict (ds), c, NULL); /* Check for missing values. */ - if (listwise_missing (dsc, &c)) + if (listwise_missing (dsc, c)) { dsc->missing_listwise += weight; if (dsc->missing_type == DSC_LISTWISE) @@ -742,7 +748,7 @@ calc_descriptives (struct dsc_proc *dsc, struct casereader *group, for (i = 0; i < dsc->var_cnt; i++) { struct dsc_var *dv = &dsc->vars[i]; - double x = case_num (&c, dv->v); + double x = case_num (c, dv->v); if (var_is_num_missing (dv->v, x, dsc->exclude)) { @@ -760,23 +766,26 @@ calc_descriptives (struct dsc_proc *dsc, struct casereader *group, } } if (!casereader_destroy (pass1)) - return; + { + casereader_destroy (pass2); + return; + } /* Second pass for higher-order moments. */ if (dsc->max_moment > MOMENT_MEAN) { - for (; casereader_read (pass2, &c); case_destroy (&c)) + for (; (c = casereader_read (pass2)) != NULL; case_unref (c)) { - double weight = dict_get_case_weight (dataset_dict (ds), &c, NULL); + double weight = dict_get_case_weight (dataset_dict (ds), c, NULL); /* Check for missing values. */ - if (dsc->missing_type == DSC_LISTWISE && listwise_missing (dsc, &c)) + if (dsc->missing_type == DSC_LISTWISE && listwise_missing (dsc, c)) continue; for (i = 0; i < dsc->var_cnt; i++) { struct dsc_var *dv = &dsc->vars[i]; - double x = case_num (&c, dv->v); + double x = case_num (c, dv->v); if (var_is_num_missing (dv->v, x, dsc->exclude)) continue; @@ -868,13 +877,12 @@ display (struct dsc_proc *dsc) sort (dsc->vars, dsc->var_cnt, sizeof *dsc->vars, descriptives_compare_dsc_vars, dsc); - t = tab_create (nc, dsc->var_cnt + 1, 0); + t = tab_create (nc, dsc->var_cnt + 1); tab_headers (t, 1, 0, 1, 0); tab_box (t, TAL_1, TAL_1, -1, -1, 0, 0, nc - 1, dsc->var_cnt); tab_box (t, -1, -1, -1, TAL_1, 1, 0, nc - 1, dsc->var_cnt); tab_hline (t, TAL_2, 0, nc - 1, 1); tab_vline (t, TAL_2, 1, 0, dsc->var_cnt); - tab_dim (t, tab_natural_dimensions); nc = 0; tab_text (t, nc++, 0, TAB_LEFT | TAT_TITLE, _("Variable")); @@ -900,12 +908,13 @@ display (struct dsc_proc *dsc) nc = 0; tab_text (t, nc++, i + 1, TAB_LEFT, var_get_name (dv->v)); - tab_text (t, nc++, i + 1, TAT_PRINTF, "%g", dv->valid); + tab_text_format (t, nc++, i + 1, 0, "%g", dv->valid); if (dsc->format == DSC_SERIAL) - tab_text (t, nc++, i + 1, TAT_PRINTF, "%g", dv->missing); + tab_text_format (t, nc++, i + 1, 0, "%g", dv->missing); + for (j = 0; j < DSC_N_STATS; j++) if (dsc->show_stats & (1ul << j)) - tab_float (t, nc++, i + 1, TAB_NONE, dv->stats[j], 10, 3); + tab_double (t, nc++, i + 1, TAB_NONE, dv->stats[j], NULL); } tab_title (t, _("Valid cases = %g; cases with missing value(s) = %g."),