X-Git-Url: https://pintos-os.org/cgi-bin/gitweb.cgi?a=blobdiff_plain;ds=sidebyside;f=src%2Flanguage%2Fstats%2Fdescriptives.c;h=6afb6b53a4f336d29e72fa2b0fffdefec7ded38c;hb=5862de297bb487c81660beb3796d5c612eeb12b7;hp=de0e417c81fdb59b397159e9bf549e8ea2455747;hpb=f43378497b8400e9c22a3485c534693dc1bc9554;p=pspp-builds.git diff --git a/src/language/stats/descriptives.c b/src/language/stats/descriptives.c index de0e417c..6afb6b53 100644 --- a/src/language/stats/descriptives.c +++ b/src/language/stats/descriptives.c @@ -1,6 +1,5 @@ /* PSPP - computes sample statistics. Copyright (C) 1997-9, 2000 Free Software Foundation, Inc. - Written by Ben Pfaff . This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as @@ -64,11 +63,10 @@ enum dsc_missing_type calculating a Z-score. */ struct dsc_z_score { - int src_idx; /* Source index into case data. */ - int dst_idx; /* Destination index into case data. */ + struct variable *src_var; /* Variable on which z-score is based. */ + struct variable *z_var; /* New z-score variable. */ double mean; /* Distribution mean. */ double std_dev; /* Distribution standard deviation. */ - struct variable *v; /* Variable on which z-score is based. */ }; /* DESCRIPTIVES transformation (for calculating Z-scores). */ @@ -169,12 +167,12 @@ struct dsc_proc }; /* Parsing. */ -static enum dsc_statistic match_statistic (void); +static enum dsc_statistic match_statistic (struct lexer *); static void free_dsc_proc (struct dsc_proc *); /* Z-score functions. */ static bool try_name (const struct dictionary *dict, - struct dsc_proc *dsc, char *name); + 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, size_t *z_cnt); @@ -191,7 +189,7 @@ static void display (struct dsc_proc *dsc); /* Handles DESCRIPTIVES. */ int -cmd_descriptives (struct dataset *ds) +cmd_descriptives (struct lexer *lexer, struct dataset *ds) { struct dictionary *dict = dataset_dict (ds); struct dsc_proc *dsc; @@ -219,106 +217,106 @@ cmd_descriptives (struct dataset *ds) dsc->show_stats = dsc->calc_stats = DEFAULT_STATS; /* Parse DESCRIPTIVES. */ - while (token != '.') + while (lex_token (lexer) != '.') { - if (lex_match_id ("MISSING")) + if (lex_match_id (lexer, "MISSING")) { - lex_match ('='); - while (token != '.' && token != '/') + lex_match (lexer, '='); + while (lex_token (lexer) != '.' && lex_token (lexer) != '/') { - if (lex_match_id ("VARIABLE")) + if (lex_match_id (lexer, "VARIABLE")) dsc->missing_type = DSC_VARIABLE; - else if (lex_match_id ("LISTWISE")) + else if (lex_match_id (lexer, "LISTWISE")) dsc->missing_type = DSC_LISTWISE; - else if (lex_match_id ("INCLUDE")) + else if (lex_match_id (lexer, "INCLUDE")) dsc->include_user_missing = 1; else { - lex_error (NULL); + lex_error (lexer, NULL); goto error; } - lex_match (','); + lex_match (lexer, ','); } } - else if (lex_match_id ("SAVE")) + else if (lex_match_id (lexer, "SAVE")) save_z_scores = 1; - else if (lex_match_id ("FORMAT")) + else if (lex_match_id (lexer, "FORMAT")) { - lex_match ('='); - while (token != '.' && token != '/') + lex_match (lexer, '='); + while (lex_token (lexer) != '.' && lex_token (lexer) != '/') { - if (lex_match_id ("LABELS")) + if (lex_match_id (lexer, "LABELS")) dsc->show_var_labels = 1; - else if (lex_match_id ("NOLABELS")) + else if (lex_match_id (lexer, "NOLABELS")) dsc->show_var_labels = 0; - else if (lex_match_id ("INDEX")) + else if (lex_match_id (lexer, "INDEX")) dsc->show_index = 1; - else if (lex_match_id ("NOINDEX")) + else if (lex_match_id (lexer, "NOINDEX")) dsc->show_index = 0; - else if (lex_match_id ("LINE")) + else if (lex_match_id (lexer, "LINE")) dsc->format = DSC_LINE; - else if (lex_match_id ("SERIAL")) + else if (lex_match_id (lexer, "SERIAL")) dsc->format = DSC_SERIAL; else { - lex_error (NULL); + lex_error (lexer, NULL); goto error; } - lex_match (','); + lex_match (lexer, ','); } } - else if (lex_match_id ("STATISTICS")) + else if (lex_match_id (lexer, "STATISTICS")) { - lex_match ('='); + lex_match (lexer, '='); dsc->show_stats = 0; - while (token != '.' && token != '/') + while (lex_token (lexer) != '.' && lex_token (lexer) != '/') { - if (lex_match (T_ALL)) + if (lex_match (lexer, T_ALL)) dsc->show_stats |= (1ul << DSC_N_STATS) - 1; - else if (lex_match_id ("DEFAULT")) + else if (lex_match_id (lexer, "DEFAULT")) dsc->show_stats |= DEFAULT_STATS; else - dsc->show_stats |= 1ul << (match_statistic ()); - lex_match (','); + dsc->show_stats |= 1ul << (match_statistic (lexer)); + lex_match (lexer, ','); } if (dsc->show_stats == 0) dsc->show_stats = DEFAULT_STATS; } - else if (lex_match_id ("SORT")) + else if (lex_match_id (lexer, "SORT")) { - lex_match ('='); - if (lex_match_id ("NAME")) + lex_match (lexer, '='); + if (lex_match_id (lexer, "NAME")) dsc->sort_by_stat = DSC_NAME; else { - dsc->sort_by_stat = match_statistic (); + dsc->sort_by_stat = match_statistic (lexer); if (dsc->sort_by_stat == DSC_NONE ) dsc->sort_by_stat = DSC_MEAN; } - if (lex_match ('(')) + if (lex_match (lexer, '(')) { - if (lex_match_id ("A")) + if (lex_match_id (lexer, "A")) dsc->sort_ascending = 1; - else if (lex_match_id ("D")) + else if (lex_match_id (lexer, "D")) dsc->sort_ascending = 0; else - lex_error (NULL); - lex_force_match (')'); + lex_error (lexer, NULL); + lex_force_match (lexer, ')'); } } else if (var_cnt == 0) { - if (lex_look_ahead () == '=') + if (lex_look_ahead (lexer) == '=') { - lex_match_id ("VARIABLES"); - lex_match ('='); + lex_match_id (lexer, "VARIABLES"); + lex_match (lexer, '='); } - while (token != '.' && token != '/') + while (lex_token (lexer) != '.' && lex_token (lexer) != '/') { int i; - if (!parse_variables (dataset_dict (ds), &vars, &var_cnt, + if (!parse_variables (lexer, dataset_dict (ds), &vars, &var_cnt, PV_APPEND | PV_NO_DUPLICATE | PV_NUMERIC)) goto error; @@ -332,34 +330,34 @@ cmd_descriptives (struct dataset *ds) } dsc->var_cnt = var_cnt; - if (lex_match ('(')) + if (lex_match (lexer, '(')) { - if (token != T_ID) + if (lex_token (lexer) != T_ID) { - lex_error (NULL); + lex_error (lexer, NULL); goto error; } - if (try_name (dict, dsc, tokid)) + if (try_name (dict, dsc, lex_tokid (lexer))) { - strcpy (dsc->vars[dsc->var_cnt - 1].z_name, tokid); + strcpy (dsc->vars[dsc->var_cnt - 1].z_name, lex_tokid (lexer)); z_cnt++; } else msg (SE, _("Z-score variable name %s would be" - " a duplicate variable name."), tokid); - lex_get (); - if (!lex_force_match (')')) + " a duplicate variable name."), lex_tokid (lexer)); + lex_get (lexer); + if (!lex_force_match (lexer, ')')) goto error; } } } else { - lex_error (NULL); + lex_error (lexer, NULL); goto error; } - lex_match ('/'); + lex_match (lexer, '/'); } if (var_cnt == 0) { @@ -378,7 +376,8 @@ cmd_descriptives (struct dataset *ds) if (dsc->vars[i].z_name[0] == 0) { if (!generate_z_varname (dict, dsc, dsc->vars[i].z_name, - dsc->vars[i].v->name, &gen_cnt)) + var_get_name (dsc->vars[i].v), + &gen_cnt)) goto error; z_cnt++; } @@ -436,18 +435,18 @@ cmd_descriptives (struct dataset *ds) specifiers). Emits an error if the current token ID does not name a statistic. */ static enum dsc_statistic -match_statistic (void) +match_statistic (struct lexer *lexer) { - if (token == T_ID) + if (lex_token (lexer) == T_ID) { enum dsc_statistic stat; for (stat = 0; stat < DSC_N_STATS; stat++) - if (lex_match_id (dsc_info[stat].identifier)) + if (lex_match_id (lexer, dsc_info[stat].identifier)) return stat; - lex_get(); - lex_error (_("expecting statistic name: reverting to default")); + lex_get (lexer); + lex_error (lexer, _("expecting statistic name: reverting to default")); } return DSC_NONE; @@ -473,7 +472,8 @@ free_dsc_proc (struct dsc_proc *dsc) /* Returns false if NAME is a duplicate of any existing variable name or of any previously-declared z-var name; otherwise returns true. */ static bool -try_name (const struct dictionary *dict, struct dsc_proc *dsc, char *name) +try_name (const struct dictionary *dict, struct dsc_proc *dsc, + const char *name) { size_t i; @@ -566,7 +566,7 @@ dump_z_table (struct dsc_proc *dsc) for (i = 0, y = 1; i < dsc->var_cnt; i++) if (dsc->vars[i].z_name[0] != '\0') { - tab_text (t, 0, y, TAB_LEFT, dsc->vars[i].v->name); + 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); } } @@ -594,10 +594,10 @@ 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)->fv); + double score = case_num (c, *vars); if ( score == SYSMIS || (!t->include_user_missing - && mv_is_num_user_missing (&(*vars)->miss, score))) + && var_is_num_user_missing (*vars, score))) { all_sysmis = 1; break; @@ -607,13 +607,13 @@ descriptives_trns_proc (void *trns_, struct ccase * c, for (z = t->z_scores; z < t->z_scores + t->z_score_cnt; z++) { - double input = case_num (c, z->src_idx); - double *output = &case_data_rw (c, z->dst_idx)->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 || input == SYSMIS || (!t->include_user_missing - && mv_is_num_user_missing (&z->v->miss, input))) + && var_is_num_user_missing (z->src_var, input))) *output = SYSMIS; else *output = (input - z->mean) / z->std_dev; @@ -668,29 +668,17 @@ setup_z_trns (struct dsc_proc *dsc, struct dataset *ds) if (dv->z_name[0] != '\0') { struct dsc_z_score *z; - char *cp; struct variable *dst_var; dst_var = dict_create_var_assert (dataset_dict (ds), dv->z_name, 0); - if (dv->v->label) - { - dst_var->label = xmalloc (strlen (dv->v->label) + 12); - cp = stpcpy (dst_var->label, _("Z-score of ")); - strcpy (cp, dv->v->label); - } - else - { - dst_var->label = xmalloc (strlen (dv->v->name) + 12); - cp = stpcpy (dst_var->label, _("Z-score of ")); - strcpy (cp, dv->v->name); - } + var_set_label (dst_var, xasprintf (_("Z-score of %s"), + var_to_string (dv->v))); z = &t->z_scores[cnt++]; - z->src_idx = dv->v->fv; - z->dst_idx = dst_var->fv; + z->src_var = dv->v; + z->z_var = dst_var; z->mean = dv->stats[DSC_MEAN]; z->std_dev = dv->stats[DSC_STDDEV]; - z->v = dv->v; } } @@ -750,12 +738,12 @@ calc_descriptives (const struct ccase *first, for (i = 0; i < dsc->var_cnt; i++) { struct dsc_var *dv = &dsc->vars[i]; - double x = case_num (&c, dv->v->fv); + double x = case_num (&c, dv->v); if (dsc->missing_type != DSC_LISTWISE && (x == SYSMIS || (!dsc->include_user_missing - && mv_is_num_user_missing (&dv->v->miss, x)))) + && var_is_num_user_missing (dv->v, x)))) { dv->missing += weight; continue; @@ -792,12 +780,12 @@ calc_descriptives (const struct ccase *first, for (i = 0; i < dsc->var_cnt; i++) { struct dsc_var *dv = &dsc->vars[i]; - double x = case_num (&c, dv->v->fv); + double x = case_num (&c, dv->v); if (dsc->missing_type != DSC_LISTWISE && (x == SYSMIS || (!dsc->include_user_missing - && mv_is_num_user_missing (&dv->v->miss, x)))) + && var_is_num_user_missing (dv->v, x)))) continue; if (dv->moments != NULL) @@ -859,11 +847,11 @@ listwise_missing (struct dsc_proc *dsc, const struct ccase *c) for (i = 0; i < dsc->var_cnt; i++) { struct dsc_var *dv = &dsc->vars[i]; - double x = case_num (c, dv->v->fv); + double x = case_num (c, dv->v); if (x == SYSMIS || (!dsc->include_user_missing - && mv_is_num_user_missing (&dv->v->miss, x))) + && var_is_num_user_missing (dv->v, x))) return true; } return false; @@ -921,7 +909,7 @@ display (struct dsc_proc *dsc) size_t j; nc = 0; - tab_text (t, nc++, i + 1, TAB_LEFT, dv->v->name); + tab_text (t, nc++, i + 1, TAB_LEFT, var_get_name (dv->v)); tab_text (t, nc++, i + 1, TAT_PRINTF, "%g", dv->valid); if (dsc->format == DSC_SERIAL) tab_text (t, nc++, i + 1, TAT_PRINTF, "%g", dv->missing); @@ -948,7 +936,7 @@ descriptives_compare_dsc_vars (const void *a_, const void *b_, const void *dsc_) int result; if (dsc->sort_by_stat == DSC_NAME) - result = strcasecmp (a->v->name, b->v->name); + result = strcasecmp (var_get_name (a->v), var_get_name (b->v)); else { double as = a->stats[dsc->sort_by_stat];