X-Git-Url: https://pintos-os.org/cgi-bin/gitweb.cgi?a=blobdiff_plain;ds=sidebyside;f=src%2Flanguage%2Fstats%2Ffactor.c;h=0679c5f506bff96006e576a0e2d2c3968427bfc0;hb=c74561b1a11c7ee5540336646f1f928bcbb77a05;hp=9d3d944e8fd1ae7d929e17e5127640af45a5f375;hpb=73ee9eab1bd4159ad1982f3012abf3631131052e;p=pspp-builds.git diff --git a/src/language/stats/factor.c b/src/language/stats/factor.c index 9d3d944e..0679c5f5 100644 --- a/src/language/stats/factor.c +++ b/src/language/stats/factor.c @@ -1,5 +1,5 @@ /* PSPP - a program for statistical analysis. - Copyright (C) 2009 Free Software Foundation, Inc. + Copyright (C) 2009, 2010 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 @@ -44,8 +44,10 @@ #include #include -#include +#include +#include +#include #include "gettext.h" #define _(msgid) gettext (msgid) @@ -70,6 +72,12 @@ enum extraction_method EXTRACTION_PAF, }; +enum plot_opts + { + PLOT_SCREE = 0x0001, + PLOT_ROTATION = 0x0002 + }; + enum print_opts { PRINT_UNIVARIATE = 0x0001, @@ -100,6 +108,7 @@ struct cmd_factor enum mv_class exclude; enum print_opts print; enum extraction_method extraction; + enum plot_opts plot; /* Extraction Criteria */ int n_factors; @@ -526,6 +535,7 @@ cmd_factor (struct lexer *lexer, struct dataset *ds) factor.econverge = 0.001; factor.blank = 0; factor.sort = false; + factor.plot = 0; factor.wv = dict_get_weight (dict); @@ -542,11 +552,13 @@ cmd_factor (struct lexer *lexer, struct dataset *ds) PV_NO_DUPLICATE | PV_NUMERIC)) goto error; + if (factor.n_vars < 2) + msg (MW, _("Factor analysis on a single variable is not useful.")); + while (lex_token (lexer) != '.') { lex_match (lexer, '/'); -#if FACTOR_FULLY_IMPLEMENTED if (lex_match_id (lexer, "PLOT")) { lex_match (lexer, '='); @@ -554,10 +566,13 @@ cmd_factor (struct lexer *lexer, struct dataset *ds) { if (lex_match_id (lexer, "EIGEN")) { + factor.plot |= PLOT_SCREE; } +#if FACTOR_FULLY_IMPLEMENTED else if (lex_match_id (lexer, "ROTATION")) { } +#endif else { lex_error (lexer, NULL); @@ -565,9 +580,7 @@ cmd_factor (struct lexer *lexer, struct dataset *ds) } } } - else -#endif - if (lex_match_id (lexer, "METHOD")) + else if (lex_match_id (lexer, "METHOD")) { lex_match (lexer, '='); while (lex_token (lexer) != '.' && lex_token (lexer) != '/') @@ -911,6 +924,22 @@ communality (struct idata *idata, int n, int n_factors) } +static void +show_scree (const struct cmd_factor *f, struct idata *idata) +{ + struct scree *s; + const char *label ; + + if ( !(f->plot & PLOT_SCREE) ) + return; + + + label = f->extraction == EXTRACTION_PC ? _("Component Number") : _("Factor Number"); + + s = scree_create (idata->eval, label); + + scree_submit (s); +} static void show_communalities (const struct cmd_factor * factor, @@ -934,12 +963,10 @@ show_communalities (const struct cmd_factor * factor, if (nc <= 1) return; - t = tab_create (nc, nr, 0); + t = tab_create (nc, nr); tab_title (t, _("Communalities")); - tab_dim (t, tab_natural_dimensions, NULL); - tab_headers (t, heading_columns, 0, heading_rows, 0); c = 1; @@ -986,7 +1013,7 @@ static void show_factor_matrix (const struct cmd_factor *factor, struct idata *idata, const gsl_matrix *fm) { int i; - const int n_factors = n_extracted_factors (factor, idata); + const int n_factors = idata->n_extractions; const int heading_columns = 1; const int heading_rows = 2; @@ -994,15 +1021,13 @@ show_factor_matrix (const struct cmd_factor *factor, struct idata *idata, const const int nc = heading_columns + n_factors; gsl_permutation *perm; - struct tab_table *t = tab_create (nc, nr, 0); + struct tab_table *t = tab_create (nc, nr); if ( factor->extraction == EXTRACTION_PC ) tab_title (t, _("Component Matrix")); else tab_title (t, _("Factor Matrix")); - tab_dim (t, tab_natural_dimensions, NULL); - tab_headers (t, heading_columns, 0, heading_rows, 0); if ( factor->extraction == EXTRACTION_PC ) @@ -1106,12 +1131,10 @@ show_explained_variance (const struct cmd_factor * factor, struct idata *idata, if ( nc <= heading_columns) return; - t = tab_create (nc, nr, 0); + t = tab_create (nc, nr); tab_title (t, _("Total Variance Explained")); - tab_dim (t, tab_natural_dimensions, NULL); - tab_headers (t, heading_columns, 0, heading_rows, 0); /* Outline the box */ @@ -1161,6 +1184,7 @@ show_explained_variance (const struct cmd_factor * factor, struct idata *idata, for (i = 0; i < (nc - heading_columns) / 3 ; ++i) { tab_text (t, i * 3 + 1, 1, TAB_CENTER | TAT_TITLE, _("Total")); + /* xgettext:no-c-format */ tab_text (t, i * 3 + 2, 1, TAB_CENTER | TAT_TITLE, _("% of Variance")); tab_text (t, i * 3 + 3, 1, TAB_CENTER | TAT_TITLE, _("Cumulative %")); @@ -1205,7 +1229,7 @@ show_explained_variance (const struct cmd_factor * factor, struct idata *idata, if (factor->print & PRINT_EXTRACTION) { - if ( i < n_extracted_factors (factor, idata)) + if (i < idata->n_extractions) { /* Sums of squared loadings */ tab_double (t, c++, i + heading_rows, 0, e_lambda, NULL); @@ -1258,12 +1282,10 @@ show_correlation_matrix (const struct cmd_factor *factor, const struct idata *id if (nr <= heading_rows && suffix_rows == 0) return; - t = tab_create (nc, nr + suffix_rows, 0); + t = tab_create (nc, nr + suffix_rows); tab_title (t, _("Correlation Matrix")); - tab_dim (t, tab_natural_dimensions, NULL); - tab_hline (t, TAL_1, 0, nc - 1, heading_rows); if (nr > heading_rows) @@ -1403,9 +1425,8 @@ do_factor (const struct cmd_factor *factor, struct casereader *r) const int nr = heading_rows + factor->n_vars; - struct tab_table *t = tab_create (nc, nr, 0); + struct tab_table *t = tab_create (nc, nr); tab_title (t, _("Descriptive Statistics")); - tab_dim (t, tab_natural_dimensions, NULL); tab_headers (t, heading_columns, 0, heading_rows, 0); @@ -1457,12 +1478,26 @@ do_factor (const struct cmd_factor *factor, struct casereader *r) gsl_eigen_symmv_sort (idata->eval, idata->evec, GSL_EIGEN_SORT_ABS_DESC); #endif + idata->n_extractions = n_extracted_factors (factor, idata); + + if (idata->n_extractions == 0) + { + msg (MW, _("The FACTOR criteria result in zero factors extracted. Therefore no analysis will be performed.")); + goto finish; + } + + if (idata->n_extractions > factor->n_vars) + { + msg (MW, _("The FACTOR criteria result in more factors than variables, which is not meaningful. No analysis will be performed.")); + goto finish; + } + { const gsl_vector *extracted_eigenvalues = NULL; gsl_vector *initial_communalities = gsl_vector_alloc (factor->n_vars); gsl_vector *extracted_communalities = gsl_vector_alloc (factor->n_vars); size_t i; - struct factor_matrix_workspace *fmw = factor_matrix_workspace_alloc (idata->msr->size, n_extracted_factors (factor, idata)); + struct factor_matrix_workspace *fmw = factor_matrix_workspace_alloc (idata->msr->size, idata->n_extractions); gsl_matrix *factor_matrix = gsl_matrix_calloc (factor->n_vars, fmw->n_factors); if ( factor->extraction == EXTRACTION_PAF) @@ -1501,10 +1536,9 @@ do_factor (const struct cmd_factor *factor, struct casereader *r) } else if (factor->extraction == EXTRACTION_PC) { - for (i = 0 ; i < factor->n_vars; ++i) - { - gsl_vector_set (initial_communalities, i, communality (idata, i, factor->n_vars)); - } + for (i = 0; i < factor->n_vars; ++i) + gsl_vector_set (initial_communalities, i, communality (idata, i, factor->n_vars)); + gsl_vector_memcpy (extracted_communalities, initial_communalities); iterate_factor_matrix (analysis_matrix, extracted_communalities, factor_matrix, fmw); @@ -1517,12 +1551,16 @@ do_factor (const struct cmd_factor *factor, struct casereader *r) factor_matrix_workspace_free (fmw); + show_scree (factor, idata); + show_factor_matrix (factor, idata, factor_matrix); gsl_vector_free (initial_communalities); gsl_vector_free (extracted_communalities); } + finish: + idata_free (idata); casereader_destroy (r);