X-Git-Url: https://pintos-os.org/cgi-bin/gitweb.cgi?a=blobdiff_plain;f=src%2Flanguage%2Fstats%2Ffactor.c;h=cf67ecd8d4de1f8f82528bd1de9cf9662e7e0a16;hb=0200682d517fd4b7cfc4e333378de03bcf74be43;hp=fc3eb5b4d279446d7057559a707a664d3298e2a8;hpb=d4ff0e074d703dbeb8af5aa3ac470ddda5ebe301;p=pspp diff --git a/src/language/stats/factor.c b/src/language/stats/factor.c index fc3eb5b4d2..cf67ecd8d4 100644 --- a/src/language/stats/factor.c +++ b/src/language/stats/factor.c @@ -47,7 +47,7 @@ #include "math/moments.h" #include "output/chart-item.h" #include "output/charts/scree.h" -#include "output/tab.h" +#include "output/pivot-table.h" #include "gettext.h" @@ -107,13 +107,13 @@ enum rotation_type typedef void (*rotation_coefficients) (double *x, double *y, double a, double b, double c, double d, - const gsl_matrix *loadings ); + const gsl_matrix *loadings); static void varimax_coefficients (double *x, double *y, double a, double b, double c, double d, - const gsl_matrix *loadings ) + const gsl_matrix *loadings) { *x = d - 2 * a * b / loadings->size1; *y = c - (a * a - b * b) / loadings->size1; @@ -122,7 +122,7 @@ varimax_coefficients (double *x, double *y, static void equamax_coefficients (double *x, double *y, double a, double b, double c, double d, - const gsl_matrix *loadings ) + const gsl_matrix *loadings) { *x = d - loadings->size2 * a * b / loadings->size1; *y = c - loadings->size2 * (a * a - b * b) / (2 * loadings->size1); @@ -243,6 +243,8 @@ struct idata double detR; /* The determinant of the correlation matrix */ + gsl_matrix *ai_cov; /* The anti-image covariance matrix */ + gsl_matrix *ai_cor; /* The anti-image correlation matrix */ struct covariance *cvm; }; @@ -266,36 +268,32 @@ idata_free (struct idata *id) gsl_vector_free (id->msr); gsl_vector_free (id->eval); gsl_matrix_free (id->evec); + gsl_matrix_free (id->ai_cov); + gsl_matrix_free (id->ai_cor); free (id); } - -static gsl_matrix * -anti_image (const gsl_matrix *m) +/* Return the sum of squares of all the elements in row J excluding column J */ +static double +ssq_row_od_n (const gsl_matrix *m, int j) { - int i, j; - gsl_matrix *a; + int i; + double ss = 0; assert (m->size1 == m->size2); - a = gsl_matrix_alloc (m->size1, m->size2); + assert (j < m->size1); for (i = 0; i < m->size1; ++i) { - for (j = 0; j < m->size2; ++j) - { - double *p = gsl_matrix_ptr (a, i, j); - *p = gsl_matrix_get (m, i, j); - *p /= gsl_matrix_get (m, i, i); - *p /= gsl_matrix_get (m, j, j); - } + if (i == j) continue; + ss += pow2 (gsl_matrix_get (m, i, j)); } - return a; + return ss; } - -/* Return the sum of all the elements excluding row N */ +/* Return the sum of squares of all the elements excluding row N */ static double ssq_od_n (const gsl_matrix *m, int n) { @@ -307,9 +305,9 @@ ssq_od_n (const gsl_matrix *m, int n) for (i = 0; i < m->size1; ++i) { - if (i == n ) continue; for (j = 0; j < m->size2; ++j) { + if (i == j) continue; ss += pow2 (gsl_matrix_get (m, i, j)); } } @@ -318,6 +316,58 @@ ssq_od_n (const gsl_matrix *m, int n) } +static gsl_matrix * +anti_image_corr (const gsl_matrix *m, const struct idata *idata) +{ + int i, j; + gsl_matrix *a; + assert (m->size1 == m->size2); + + a = gsl_matrix_alloc (m->size1, m->size2); + + for (i = 0; i < m->size1; ++i) + { + for (j = 0; j < m->size2; ++j) + { + double *p = gsl_matrix_ptr (a, i, j); + *p = gsl_matrix_get (m, i, j); + *p /= sqrt (gsl_matrix_get (m, i, i) * + gsl_matrix_get (m, j, j)); + } + } + + for (i = 0; i < m->size1; ++i) + { + double r = ssq_row_od_n (idata->mm.corr, i); + double u = ssq_row_od_n (a, i); + gsl_matrix_set (a, i, i, r / (r + u)); + } + + return a; +} + +static gsl_matrix * +anti_image_cov (const gsl_matrix *m) +{ + int i, j; + gsl_matrix *a; + assert (m->size1 == m->size2); + + a = gsl_matrix_alloc (m->size1, m->size2); + + for (i = 0; i < m->size1; ++i) + { + for (j = 0; j < m->size2; ++j) + { + double *p = gsl_matrix_ptr (a, i, j); + *p = gsl_matrix_get (m, i, j); + *p /= gsl_matrix_get (m, i, i); + *p /= gsl_matrix_get (m, j, j); + } + } + + return a; +} #if 0 static void @@ -366,7 +416,7 @@ n_extracted_factors (const struct cmd_factor *factor, struct idata *idata) int i; /* If there is a cached value, then return that. */ - if ( idata->n_extractions != 0) + if (idata->n_extractions != 0) return idata->n_extractions; /* Otherwise, if the number of factors has been explicitly requested, @@ -595,7 +645,7 @@ sort_matrix_indirect (const gsl_matrix *input, gsl_permutation *perm) gsl_vector_view row = gsl_matrix_row (mat, p->data[n - 1 - i]); size_t maxindex = gsl_vector_max_index (&row.vector); - if ( maxindex > column_n ) + if (maxindex > column_n) break; /* All subsequent elements of this row, are of no interest. @@ -613,7 +663,7 @@ sort_matrix_indirect (const gsl_matrix *input, gsl_permutation *perm) gsl_permutation_free (p); gsl_matrix_free (mat); - assert ( 0 == gsl_permutation_valid (perm)); + assert (0 == gsl_permutation_valid (perm)); /* We want the biggest value to be first */ gsl_permutation_reverse (perm); @@ -670,7 +720,7 @@ initial_sv (const gsl_matrix *fm) l4s += lambda_4; l2s += lambda_sq; } - sv += ( fm->size1 * l4s - (l2s * l2s) ) / (fm->size1 * fm->size1 ); + sv += (fm->size1 * l4s - (l2s * l2s)) / (fm->size1 * fm->size1); } return sv; } @@ -755,7 +805,7 @@ rotate (const struct cmd_factor *cf, const gsl_matrix *unrot, phi = atan2 (x, y) / 4.0 ; /* Don't bother rotating if the angle is small */ - if ( fabs (sin (phi) ) <= pow (10.0, -15.0)) + if (fabs (sin (phi)) <= pow (10.0, -15.0)) continue; for (p = 0; p < normalised->size1; ++p) @@ -775,10 +825,10 @@ rotate (const struct cmd_factor *cf, const gsl_matrix *unrot, l2s += lambda_sq; } } - sv += ( normalised->size1 * l4s - (l2s * l2s) ) / (normalised->size1 * normalised->size1 ); + sv += (normalised->size1 * l4s - (l2s * l2s)) / (normalised->size1 * normalised->size1); } - if ( fabs (sv - prev_sv) <= cf->rconverge) + if (fabs (sv - prev_sv) <= cf->rconverge) break; prev_sv = sv; @@ -941,7 +991,7 @@ rotate (const struct cmd_factor *cf, const gsl_matrix *unrot, gsl_vector_set (rotated_loadings, i, ssq); - if ( sum < 0 ) + if (sum < 0) for (j = 0 ; j < result->size1; ++j) { double *lambda = gsl_matrix_ptr (result, j, i); @@ -1220,7 +1270,7 @@ cmd_factor (struct lexer *lexer, struct dataset *ds) { if (lex_match_id (lexer, "FACTORS")) { - if ( lex_force_match (lexer, T_LPAREN) + if (lex_force_match (lexer, T_LPAREN) && lex_force_int (lexer)) { factor.n_factors = lex_integer (lexer); @@ -1231,7 +1281,7 @@ cmd_factor (struct lexer *lexer, struct dataset *ds) } else if (lex_match_id (lexer, "MINEIGEN")) { - if ( lex_force_match (lexer, T_LPAREN) + if (lex_force_match (lexer, T_LPAREN) && lex_force_num (lexer)) { factor.min_eigen = lex_number (lexer); @@ -1242,7 +1292,7 @@ cmd_factor (struct lexer *lexer, struct dataset *ds) } else if (lex_match_id (lexer, "ECONVERGE")) { - if ( lex_force_match (lexer, T_LPAREN) + if (lex_force_match (lexer, T_LPAREN) && lex_force_num (lexer)) { factor.econverge = lex_number (lexer); @@ -1264,7 +1314,7 @@ cmd_factor (struct lexer *lexer, struct dataset *ds) } else if (lex_match_id (lexer, "ITERATE")) { - if ( lex_force_match (lexer, T_LPAREN) + if (lex_force_match (lexer, T_LPAREN) && lex_force_int (lexer)) { n_iterations = lex_integer (lexer); @@ -1326,7 +1376,7 @@ cmd_factor (struct lexer *lexer, struct dataset *ds) } else if (lex_match_id (lexer, "BLANK")) { - if ( lex_force_match (lexer, T_LPAREN) + if (lex_force_match (lexer, T_LPAREN) && lex_force_num (lexer)) { factor.blank = lex_number (lexer); @@ -1365,10 +1415,11 @@ cmd_factor (struct lexer *lexer, struct dataset *ds) else if (lex_match_id (lexer, "INV")) { } +#endif else if (lex_match_id (lexer, "AIC")) { + factor.print |= PRINT_AIC; } -#endif else if (lex_match_id (lexer, "SIG")) { factor.print |= PRINT_SIG; @@ -1461,7 +1512,7 @@ cmd_factor (struct lexer *lexer, struct dataset *ds) } } - if ( factor.rotation == ROT_NONE ) + if (factor.rotation == ROT_NONE) factor.print &= ~PRINT_ROTATION; if (factor.n_vars < 2) @@ -1482,6 +1533,10 @@ cmd_factor (struct lexer *lexer, struct dataset *ds) { do_factor_by_matrix (&factor, id); + gsl_matrix_free (id->ai_cov); + id->ai_cov = NULL; + gsl_matrix_free (id->ai_cor); + id->ai_cor = NULL; gsl_matrix_free (id->mm.corr); id->mm.corr = NULL; gsl_matrix_free (id->mm.cov); @@ -1491,7 +1546,7 @@ cmd_factor (struct lexer *lexer, struct dataset *ds) idata_free (id); } else - if ( ! run_factor (ds, &factor)) + if (! run_factor (ds, &factor)) goto error; @@ -1519,7 +1574,7 @@ run_factor (struct dataset *ds, const struct cmd_factor *factor) while (casegrouper_get_next_group (grouper, &group)) { - if ( factor->missing_type == MISS_LISTWISE ) + if (factor->missing_type == MISS_LISTWISE) group = casereader_create_filter_missing (group, factor->vars, factor->n_vars, factor->exclude, NULL, NULL); @@ -1572,7 +1627,7 @@ show_scree (const struct cmd_factor *f, const struct idata *idata) struct scree *s; const char *label ; - if ( !(f->plot & PLOT_SCREE) ) + if (!(f->plot & PLOT_SCREE)) return; @@ -1587,162 +1642,101 @@ static void show_communalities (const struct cmd_factor * factor, const gsl_vector *initial, const gsl_vector *extracted) { - int i; - int c = 0; - const int heading_columns = 1; - int nc = heading_columns; - const int heading_rows = 1; - const int nr = heading_rows + factor->n_vars; - struct tab_table *t; - - if (factor->print & PRINT_EXTRACTION) - nc++; - - if (factor->print & PRINT_INITIAL) - nc++; - - /* No point having a table with only headings */ - if (nc <= 1) + if (!(factor->print & (PRINT_INITIAL | PRINT_EXTRACTION))) return; - t = tab_create (nc, nr); + struct pivot_table *table = pivot_table_create (N_("Communalities")); - tab_title (t, _("Communalities")); - - tab_headers (t, heading_columns, 0, heading_rows, 0); - - c = 1; + struct pivot_dimension *communalities = pivot_dimension_create ( + table, PIVOT_AXIS_COLUMN, N_("Communalities")); if (factor->print & PRINT_INITIAL) - tab_text (t, c++, 0, TAB_CENTER | TAT_TITLE, _("Initial")); - + pivot_category_create_leaves (communalities->root, N_("Initial")); if (factor->print & PRINT_EXTRACTION) - tab_text (t, c++, 0, TAB_CENTER | TAT_TITLE, _("Extraction")); - - /* Outline the box */ - tab_box (t, - TAL_2, TAL_2, - -1, -1, - 0, 0, - nc - 1, nr - 1); - - /* Vertical lines */ - tab_box (t, - -1, -1, - -1, TAL_1, - heading_columns, 0, - nc - 1, nr - 1); - - tab_hline (t, TAL_1, 0, nc - 1, heading_rows); - tab_vline (t, TAL_2, heading_columns, 0, nr - 1); - - for (i = 0 ; i < factor->n_vars; ++i) + pivot_category_create_leaves (communalities->root, N_("Extraction")); + + struct pivot_dimension *variables = pivot_dimension_create ( + table, PIVOT_AXIS_ROW, N_("Variables")); + + for (size_t i = 0 ; i < factor->n_vars; ++i) { - c = 0; - tab_text (t, c++, i + heading_rows, TAT_TITLE, var_to_string (factor->vars[i])); + int row = pivot_category_create_leaf ( + variables->root, pivot_value_new_variable (factor->vars[i])); + int col = 0; if (factor->print & PRINT_INITIAL) - tab_double (t, c++, i + heading_rows, 0, gsl_vector_get (initial, i), NULL, RC_OTHER); - + pivot_table_put2 (table, col++, row, pivot_value_new_number ( + gsl_vector_get (initial, i))); if (factor->print & PRINT_EXTRACTION) - tab_double (t, c++, i + heading_rows, 0, gsl_vector_get (extracted, i), NULL, RC_OTHER); + pivot_table_put2 (table, col++, row, pivot_value_new_number ( + gsl_vector_get (extracted, i))); } - tab_submit (t); + pivot_table_submit (table); } +static struct pivot_dimension * +create_numeric_dimension (struct pivot_table *table, + enum pivot_axis_type axis_type, const char *name, + size_t n, bool show_label) +{ + struct pivot_dimension *d = pivot_dimension_create (table, axis_type, name); + d->root->show_label = show_label; + for (int i = 0 ; i < n; ++i) + pivot_category_create_leaf (d->root, pivot_value_new_integer (i + 1)); + return d; +} static void show_factor_matrix (const struct cmd_factor *factor, const struct idata *idata, const char *title, const gsl_matrix *fm) { - int i; + struct pivot_table *table = pivot_table_create (title); const int n_factors = idata->n_extractions; + create_numeric_dimension ( + table, PIVOT_AXIS_COLUMN, + factor->extraction == EXTRACTION_PC ? N_("Component") : N_("Factor"), + n_factors, true); - const int heading_columns = 1; - const int heading_rows = 2; - const int nr = heading_rows + factor->n_vars; - const int nc = heading_columns + n_factors; - gsl_permutation *perm; - - 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_title (t, "%s", title); - - tab_headers (t, heading_columns, 0, heading_rows, 0); - - if ( factor->extraction == EXTRACTION_PC ) - tab_joint_text (t, - 1, 0, - nc - 1, 0, - TAB_CENTER | TAT_TITLE, _("Component")); - else - tab_joint_text (t, - 1, 0, - nc - 1, 0, - TAB_CENTER | TAT_TITLE, _("Factor")); - - - tab_hline (t, TAL_1, heading_columns, nc - 1, 1); - - - /* Outline the box */ - tab_box (t, - TAL_2, TAL_2, - -1, -1, - 0, 0, - nc - 1, nr - 1); - - /* Vertical lines */ - tab_box (t, - -1, -1, - -1, TAL_1, - heading_columns, 1, - nc - 1, nr - 1); - - tab_hline (t, TAL_1, 0, nc - 1, heading_rows); - tab_vline (t, TAL_2, heading_columns, 0, nr - 1); - + struct pivot_dimension *variables = pivot_dimension_create ( + table, PIVOT_AXIS_ROW, N_("Variables")); /* Initialise to the identity permutation */ - perm = gsl_permutation_calloc (factor->n_vars); + gsl_permutation *perm = gsl_permutation_calloc (factor->n_vars); - if ( factor->sort) + if (factor->sort) sort_matrix_indirect (fm, perm); - for (i = 0 ; i < n_factors; ++i) - { - tab_text_format (t, heading_columns + i, 1, TAB_CENTER | TAT_TITLE, _("%d"), i + 1); - } - - for (i = 0 ; i < factor->n_vars; ++i) + for (size_t i = 0 ; i < factor->n_vars; ++i) { - int j; const int matrix_row = perm->data[i]; - tab_text (t, 0, i + heading_rows, TAT_TITLE, var_to_string (factor->vars[matrix_row])); - for (j = 0 ; j < n_factors; ++j) + int var_idx = pivot_category_create_leaf ( + variables->root, pivot_value_new_variable (factor->vars[matrix_row])); + + for (size_t j = 0 ; j < n_factors; ++j) { double x = gsl_matrix_get (fm, matrix_row, j); - - if ( fabs (x) < factor->blank) + if (fabs (x) < factor->blank) continue; - tab_double (t, heading_columns + j, heading_rows + i, 0, x, NULL, RC_OTHER); + pivot_table_put2 (table, j, var_idx, pivot_value_new_number (x)); } } gsl_permutation_free (perm); - tab_submit (t); + pivot_table_submit (table); } +static void +put_variance (struct pivot_table *table, int row, int phase_idx, + double lambda, double percent, double cum) +{ + double entries[] = { lambda, percent, cum }; + for (size_t i = 0; i < sizeof entries / sizeof *entries; i++) + pivot_table_put3 (table, i, phase_idx, row, + pivot_value_new_number (entries[i])); +} static void show_explained_variance (const struct cmd_factor * factor, @@ -1751,442 +1745,226 @@ show_explained_variance (const struct cmd_factor * factor, const gsl_vector *extracted_eigenvalues, const gsl_vector *rotated_loadings) { - size_t i; - int c = 0; - const int heading_columns = 1; - const int heading_rows = 2; - const int nr = heading_rows + factor->n_vars; - - struct tab_table *t ; - - double i_total = 0.0; - double i_cum = 0.0; - - double e_total = 0.0; - double e_cum = 0.0; - - double r_cum = 0.0; - - int nc = heading_columns; - - if (factor->print & PRINT_EXTRACTION) - nc += 3; - - if (factor->print & PRINT_INITIAL) - nc += 3; - - if (factor->print & PRINT_ROTATION) - { - nc += factor->rotation == ROT_PROMAX ? 1 : 3; - } - - /* No point having a table with only headings */ - if ( nc <= heading_columns) + if (!(factor->print & (PRINT_INITIAL | PRINT_EXTRACTION | PRINT_ROTATION))) return; - t = tab_create (nc, nr); - - tab_title (t, _("Total Variance Explained")); - - tab_headers (t, heading_columns, 0, heading_rows, 0); - - /* Outline the box */ - tab_box (t, - TAL_2, TAL_2, - -1, -1, - 0, 0, - nc - 1, nr - 1); - - /* Vertical lines */ - tab_box (t, - -1, -1, - -1, TAL_1, - heading_columns, 0, - nc - 1, nr - 1); - - tab_hline (t, TAL_1, 0, nc - 1, heading_rows); - tab_hline (t, TAL_1, 1, nc - 1, 1); + struct pivot_table *table = pivot_table_create ( + N_("Total Variance Explained")); - tab_vline (t, TAL_2, heading_columns, 0, nr - 1); + pivot_dimension_create (table, PIVOT_AXIS_COLUMN, N_("Statistics"), + N_("Total"), PIVOT_RC_OTHER, + /* xgettext:no-c-format */ + N_("% of Variance"), PIVOT_RC_PERCENT, + /* xgettext:no-c-format */ + N_("Cumulative %"), PIVOT_RC_PERCENT); - - if ( factor->extraction == EXTRACTION_PC) - tab_text (t, 0, 1, TAB_LEFT | TAT_TITLE, _("Component")); - else - tab_text (t, 0, 1, TAB_LEFT | TAT_TITLE, _("Factor")); - - c = 1; + struct pivot_dimension *phase = pivot_dimension_create ( + table, PIVOT_AXIS_COLUMN, N_("Phase")); if (factor->print & PRINT_INITIAL) - { - tab_joint_text (t, c, 0, c + 2, 0, TAB_CENTER | TAT_TITLE, _("Initial Eigenvalues")); - c += 3; - } + pivot_category_create_leaves (phase->root, N_("Initial Eigenvalues")); if (factor->print & PRINT_EXTRACTION) - { - tab_joint_text (t, c, 0, c + 2, 0, TAB_CENTER | TAT_TITLE, _("Extraction Sums of Squared Loadings")); - c += 3; - } + pivot_category_create_leaves (phase->root, + N_("Extraction Sums of Squared Loadings")); if (factor->print & PRINT_ROTATION) - { - const int width = factor->rotation == ROT_PROMAX ? 0 : 2; - tab_joint_text (t, c, 0, c + width, 0, TAB_CENTER | TAT_TITLE, _("Rotation Sums of Squared Loadings")); - c += width + 1; - } + pivot_category_create_leaves (phase->root, + N_("Rotation Sums of Squared Loadings")); - for (i = 0; i < (nc - heading_columns + 2) / 3 ; ++i) - { - tab_text (t, i * 3 + 1, 1, TAB_CENTER | TAT_TITLE, _("Total")); - - tab_vline (t, TAL_2, heading_columns + i * 3, 0, nr - 1); - - if (i == 2 && factor->rotation == ROT_PROMAX) - continue; - - /* 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 %")); - } + struct pivot_dimension *components = pivot_dimension_create ( + table, PIVOT_AXIS_ROW, + factor->extraction == EXTRACTION_PC ? N_("Component") : N_("Factor")); - for (i = 0 ; i < initial_eigenvalues->size; ++i) + double i_total = 0.0; + for (size_t i = 0 ; i < initial_eigenvalues->size; ++i) i_total += gsl_vector_get (initial_eigenvalues, i); - if ( factor->extraction == EXTRACTION_PAF) - { - e_total = factor->n_vars; - } - else - { - e_total = i_total; - } + double e_total = (factor->extraction == EXTRACTION_PAF + ? factor->n_vars + : i_total); - for (i = 0 ; i < factor->n_vars; ++i) + double i_cum = 0.0; + double e_cum = 0.0; + double r_cum = 0.0; + for (size_t i = 0 ; i < factor->n_vars; ++i) { const double i_lambda = gsl_vector_get (initial_eigenvalues, i); double i_percent = 100.0 * i_lambda / i_total ; + i_cum += i_percent; const double e_lambda = gsl_vector_get (extracted_eigenvalues, i); double e_percent = 100.0 * e_lambda / e_total ; + e_cum += e_percent; - c = 0; - - tab_text_format (t, c++, i + heading_rows, TAB_LEFT | TAT_TITLE, _("%zu"), i + 1); + int row = pivot_category_create_leaf ( + components->root, pivot_value_new_integer (i + 1)); - i_cum += i_percent; - e_cum += e_percent; + int phase_idx = 0; /* Initial Eigenvalues */ if (factor->print & PRINT_INITIAL) - { - tab_double (t, c++, i + heading_rows, 0, i_lambda, NULL, RC_OTHER); - tab_double (t, c++, i + heading_rows, 0, i_percent, NULL, RC_OTHER); - tab_double (t, c++, i + heading_rows, 0, i_cum, NULL, RC_OTHER); - } - - - if (factor->print & PRINT_EXTRACTION) - { - if (i < idata->n_extractions) - { - /* Sums of squared loadings */ - tab_double (t, c++, i + heading_rows, 0, e_lambda, NULL, RC_OTHER); - tab_double (t, c++, i + heading_rows, 0, e_percent, NULL, RC_OTHER); - tab_double (t, c++, i + heading_rows, 0, e_cum, NULL, RC_OTHER); - } - } + put_variance (table, row, phase_idx++, i_lambda, i_percent, i_cum); - if (rotated_loadings != NULL) + if (i < idata->n_extractions) { - const double r_lambda = gsl_vector_get (rotated_loadings, i); - double r_percent = 100.0 * r_lambda / e_total ; + if (factor->print & PRINT_EXTRACTION) + put_variance (table, row, phase_idx++, e_lambda, e_percent, e_cum); - if (factor->print & PRINT_ROTATION) + if (rotated_loadings != NULL && factor->print & PRINT_ROTATION) { - if (i < idata->n_extractions) - { - r_cum += r_percent; - tab_double (t, c++, i + heading_rows, 0, r_lambda, NULL, RC_OTHER); - if (factor->rotation != ROT_PROMAX) - { - tab_double (t, c++, i + heading_rows, 0, r_percent, NULL, RC_OTHER); - tab_double (t, c++, i + heading_rows, 0, r_cum, NULL, RC_OTHER); - } - } + double r_lambda = gsl_vector_get (rotated_loadings, i); + double r_percent = 100.0 * r_lambda / e_total ; + if (factor->rotation == ROT_PROMAX) + r_lambda = r_percent = SYSMIS; + + r_cum += r_percent; + put_variance (table, row, phase_idx++, r_lambda, r_percent, + r_cum); } } } - tab_submit (t); + pivot_table_submit (table); } - static void show_factor_correlation (const struct cmd_factor * factor, const gsl_matrix *fcm) { - size_t i, j; - const int heading_columns = 1; - const int heading_rows = 1; - const int nr = heading_rows + fcm->size2; - const int nc = heading_columns + fcm->size1; - struct tab_table *t = tab_create (nc, nr); + struct pivot_table *table = pivot_table_create ( + N_("Factor Correlation Matrix")); - tab_title (t, _("Factor Correlation Matrix")); + create_numeric_dimension ( + table, PIVOT_AXIS_ROW, + factor->extraction == EXTRACTION_PC ? N_("Component") : N_("Factor"), + fcm->size2, true); - tab_headers (t, heading_columns, 0, heading_rows, 0); + create_numeric_dimension (table, PIVOT_AXIS_COLUMN, N_("Factor 2"), + fcm->size1, false); - /* Outline the box */ - tab_box (t, - TAL_2, TAL_2, - -1, -1, - 0, 0, - nc - 1, nr - 1); + for (size_t i = 0 ; i < fcm->size1; ++i) + for (size_t j = 0 ; j < fcm->size2; ++j) + pivot_table_put2 (table, j, i, + pivot_value_new_number (gsl_matrix_get (fcm, i, j))); - /* Vertical lines */ - tab_box (t, - -1, -1, - -1, TAL_1, - heading_columns, 0, - nc - 1, nr - 1); - - tab_hline (t, TAL_1, 0, nc - 1, heading_rows); - tab_hline (t, TAL_1, 1, nc - 1, 1); - - tab_vline (t, TAL_2, heading_columns, 0, nr - 1); - - - if ( factor->extraction == EXTRACTION_PC) - tab_text (t, 0, 0, TAB_LEFT | TAT_TITLE, _("Component")); - else - tab_text (t, 0, 0, TAB_LEFT | TAT_TITLE, _("Factor")); - - for (i = 0 ; i < fcm->size1; ++i) - { - tab_text_format (t, heading_columns + i, 0, TAB_CENTER | TAT_TITLE, _("%zu"), i + 1); - } + pivot_table_submit (table); +} - for (i = 0 ; i < fcm->size2; ++i) +static void +add_var_dims (struct pivot_table *table, const struct cmd_factor *factor) +{ + for (int i = 0; i < 2; i++) { - tab_text_format (t, 0, heading_rows + i, TAB_CENTER | TAT_TITLE, _("%zu"), i + 1); - } - + struct pivot_dimension *d = pivot_dimension_create ( + table, i ? PIVOT_AXIS_ROW : PIVOT_AXIS_COLUMN, + N_("Variables")); - for (i = 0 ; i < fcm->size1; ++i) - { - for (j = 0 ; j < fcm->size2; ++j) - tab_double (t, heading_columns + j, heading_rows + i, 0, - gsl_matrix_get (fcm, i, j), NULL, RC_OTHER); + for (size_t j = 0; j < factor->n_vars; j++) + pivot_category_create_leaf ( + d->root, pivot_value_new_variable (factor->vars[j])); } - - tab_submit (t); } - static void -show_correlation_matrix (const struct cmd_factor *factor, const struct idata *idata) +show_aic (const struct cmd_factor *factor, const struct idata *idata) { - struct tab_table *t ; - size_t i, j; - int y_pos_corr = -1; - int y_pos_sig = -1; - int suffix_rows = 0; + if ((factor->print & PRINT_AIC) == 0) + return; - const int heading_rows = 1; - const int heading_columns = 2; + struct pivot_table *table = pivot_table_create (N_("Anti-Image Matrices")); - int nc = heading_columns ; - int nr = heading_rows ; - int n_data_sets = 0; + add_var_dims (table, factor); - if (factor->print & PRINT_CORRELATION) - { - y_pos_corr = n_data_sets; - n_data_sets++; - nc = heading_columns + factor->n_vars; - } + pivot_dimension_create (table, PIVOT_AXIS_ROW, N_("Statistics"), + N_("Anti-image Covariance"), + N_("Anti-image Correlation")); - if (factor->print & PRINT_SIG) - { - y_pos_sig = n_data_sets; - n_data_sets++; - nc = heading_columns + factor->n_vars; - } + for (size_t i = 0; i < factor->n_vars; ++i) + for (size_t j = 0; j < factor->n_vars; ++j) + { + double cov = gsl_matrix_get (idata->ai_cov, i, j); + pivot_table_put3 (table, i, j, 0, pivot_value_new_number (cov)); - nr += n_data_sets * factor->n_vars; + double corr = gsl_matrix_get (idata->ai_cor, i, j); + pivot_table_put3 (table, i, j, 1, pivot_value_new_number (corr)); + } - if (factor->print & PRINT_DETERMINANT) - suffix_rows = 1; + pivot_table_submit (table); +} - /* If the table would contain only headings, don't bother rendering it */ - if (nr <= heading_rows && suffix_rows == 0) +static void +show_correlation_matrix (const struct cmd_factor *factor, const struct idata *idata) +{ + if (!(factor->print & (PRINT_CORRELATION | PRINT_SIG | PRINT_DETERMINANT))) return; - t = tab_create (nc, nr + suffix_rows); - - tab_title (t, _("Correlation Matrix")); - - tab_hline (t, TAL_1, 0, nc - 1, heading_rows); + struct pivot_table *table = pivot_table_create (N_("Correlation Matrix")); - if (nr > heading_rows) + if (factor->print & (PRINT_CORRELATION | PRINT_SIG)) { - tab_headers (t, heading_columns, 0, heading_rows, 0); - - tab_vline (t, TAL_2, 2, 0, nr - 1); - - /* Outline the box */ - tab_box (t, - TAL_2, TAL_2, - -1, -1, - 0, 0, - nc - 1, nr - 1); - - /* Vertical lines */ - tab_box (t, - -1, -1, - -1, TAL_1, - heading_columns, 0, - nc - 1, nr - 1); - - - for (i = 0; i < factor->n_vars; ++i) - tab_text (t, heading_columns + i, 0, TAT_TITLE, var_to_string (factor->vars[i])); - - - for (i = 0 ; i < n_data_sets; ++i) - { - int y = heading_rows + i * factor->n_vars; - size_t v; - for (v = 0; v < factor->n_vars; ++v) - tab_text (t, 1, y + v, TAT_TITLE, var_to_string (factor->vars[v])); - - tab_hline (t, TAL_1, 0, nc - 1, y); - } + add_var_dims (table, factor); + struct pivot_dimension *statistics = pivot_dimension_create ( + table, PIVOT_AXIS_ROW, N_("Statistics")); if (factor->print & PRINT_CORRELATION) - { - const double y = heading_rows + y_pos_corr; - tab_text (t, 0, y, TAT_TITLE, _("Correlations")); - - for (i = 0; i < factor->n_vars; ++i) - { - for (j = 0; j < factor->n_vars; ++j) - tab_double (t, heading_columns + j, y + i, 0, gsl_matrix_get (idata->mm.corr, i, j), NULL, RC_OTHER); - } - } - + pivot_category_create_leaves (statistics->root, N_("Correlation"), + PIVOT_RC_CORRELATION); if (factor->print & PRINT_SIG) - { - const double y = heading_rows + y_pos_sig * factor->n_vars; - tab_text (t, 0, y, TAT_TITLE, _("Sig. (1-tailed)")); + pivot_category_create_leaves (statistics->root, N_("Sig. (1-tailed)"), + PIVOT_RC_SIGNIFICANCE); - for (i = 0; i < factor->n_vars; ++i) - { - for (j = 0; j < factor->n_vars; ++j) - { - double rho = gsl_matrix_get (idata->mm.corr, i, j); - double w = gsl_matrix_get (idata->mm.n, i, j); - - if (i == j) - continue; + int stat_idx = 0; + if (factor->print & PRINT_CORRELATION) + { + for (int i = 0; i < factor->n_vars; ++i) + for (int j = 0; j < factor->n_vars; ++j) + { + double corr = gsl_matrix_get (idata->mm.corr, i, j); + pivot_table_put3 (table, j, i, stat_idx, + pivot_value_new_number (corr)); + } + stat_idx++; + } - tab_double (t, heading_columns + j, y + i, 0, significance_of_correlation (rho, w), NULL, RC_PVALUE); - } - } - } + if (factor->print & PRINT_SIG) + { + for (int i = 0; i < factor->n_vars; ++i) + for (int j = 0; j < factor->n_vars; ++j) + if (i != j) + { + double rho = gsl_matrix_get (idata->mm.corr, i, j); + double w = gsl_matrix_get (idata->mm.n, i, j); + double sig = significance_of_correlation (rho, w); + pivot_table_put3 (table, j, i, stat_idx, + pivot_value_new_number (sig)); + } + stat_idx++; + } } if (factor->print & PRINT_DETERMINANT) - { - tab_text (t, 0, nr, TAB_LEFT | TAT_TITLE, _("Determinant")); + table->caption = pivot_value_new_user_text_nocopy ( + xasprintf ("%s: %.2f", _("Determinant"), idata->detR)); - tab_double (t, 1, nr, 0, idata->detR, NULL, RC_OTHER); - } - - tab_submit (t); + pivot_table_submit (table); } static void show_covariance_matrix (const struct cmd_factor *factor, const struct idata *idata) { - struct tab_table *t ; - size_t i, j; - int y_pos_corr = -1; - int suffix_rows = 0; - - const int heading_rows = 1; - const int heading_columns = 1; - - int nc = heading_columns ; - int nr = heading_rows ; - int n_data_sets = 0; - - if (factor->print & PRINT_COVARIANCE) - { - y_pos_corr = n_data_sets; - n_data_sets++; - nc = heading_columns + factor->n_vars; - } - - nr += n_data_sets * factor->n_vars; - - /* If the table would contain only headings, don't bother rendering it */ - if (nr <= heading_rows && suffix_rows == 0) + if (!(factor->print & PRINT_COVARIANCE)) return; - t = tab_create (nc, nr + suffix_rows); - - tab_title (t, _("Covariance Matrix")); - - tab_hline (t, TAL_1, 0, nc - 1, heading_rows); - - if (nr > heading_rows) - { - tab_headers (t, heading_columns, 0, heading_rows, 0); - - tab_vline (t, TAL_2, heading_columns, 0, nr - 1); - - /* Outline the box */ - tab_box (t, - TAL_2, TAL_2, - -1, -1, - 0, 0, - nc - 1, nr - 1); - - /* Vertical lines */ - tab_box (t, - -1, -1, - -1, TAL_1, - heading_columns, 0, - nc - 1, nr - 1); - + struct pivot_table *table = pivot_table_create (N_("Covariance Matrix")); + add_var_dims (table, factor); - for (i = 0; i < factor->n_vars; ++i) - tab_text (t, heading_columns + i, 0, TAT_TITLE, var_to_string (factor->vars[i])); - - - for (i = 0 ; i < n_data_sets; ++i) - { - int y = heading_rows + i * factor->n_vars; - size_t v; - for (v = 0; v < factor->n_vars; ++v) - tab_text (t, heading_columns -1, y + v, TAT_TITLE, var_to_string (factor->vars[v])); - - tab_hline (t, TAL_1, 0, nc - 1, y); - } - - if (factor->print & PRINT_COVARIANCE) - { - const double y = heading_rows + y_pos_corr; - - for (i = 0; i < factor->n_vars; ++i) - { - for (j = 0; j < factor->n_vars; ++j) - tab_double (t, heading_columns + j, y + i, 0, gsl_matrix_get (idata->mm.cov, i, j), NULL, RC_OTHER); - } - } - } + for (int i = 0; i < factor->n_vars; ++i) + for (int j = 0; j < factor->n_vars; ++j) + { + double cov = gsl_matrix_get (idata->mm.cov, i, j); + pivot_table_put2 (table, j, i, pivot_value_new_number (cov)); + } - tab_submit (t); + pivot_table_submit (table); } @@ -2199,7 +1977,7 @@ do_factor (const struct cmd_factor *factor, struct casereader *r) idata->cvm = covariance_1pass_create (factor->n_vars, factor->vars, factor->wv, factor->exclude, true); - for ( ; (c = casereader_read (r) ); case_unref (c)) + for (; (c = casereader_read (r)); case_unref (c)) { covariance_accumulate (idata->cvm, c); } @@ -2245,6 +2023,25 @@ do_factor_by_matrix (const struct cmd_factor *factor, struct idata *idata) else idata->analysis_matrix = idata->mm.cov; + gsl_matrix *r_inv; + r_inv = clone_matrix (idata->mm.corr); + gsl_linalg_cholesky_decomp (r_inv); + gsl_linalg_cholesky_invert (r_inv); + + idata->ai_cov = anti_image_cov (r_inv); + idata->ai_cor = anti_image_corr (r_inv, idata); + + int i; + double sum_ssq_r = 0; + double sum_ssq_a = 0; + for (i = 0; i < r_inv->size1; ++i) + { + sum_ssq_r += ssq_od_n (idata->mm.corr, i); + sum_ssq_a += ssq_od_n (idata->ai_cor, i); + } + + gsl_matrix_free (r_inv); + if (factor->print & PRINT_DETERMINANT || factor->print & PRINT_KMO) { @@ -2261,135 +2058,75 @@ do_factor_by_matrix (const struct cmd_factor *factor, struct idata *idata) gsl_matrix_free (tmp); } - if ( factor->print & PRINT_UNIVARIATE) + if (factor->print & PRINT_UNIVARIATE) { - const struct fmt_spec *wfmt = factor->wv ? var_get_print_format (factor->wv) : & F_8_0; - const int nc = 4; - int i; - - const int heading_columns = 1; - const int heading_rows = 1; - - const int nr = heading_rows + factor->n_vars; - - struct tab_table *t = tab_create (nc, nr); - tab_set_format (t, RC_WEIGHT, wfmt); - tab_title (t, _("Descriptive Statistics")); + struct pivot_table *table = pivot_table_create ( + N_("Descriptive Statistics")); + pivot_table_set_weight_var (table, factor->wv); - tab_headers (t, heading_columns, 0, heading_rows, 0); + pivot_dimension_create (table, PIVOT_AXIS_COLUMN, N_("Statistics"), + N_("Mean"), PIVOT_RC_OTHER, + N_("Std. Deviation"), PIVOT_RC_OTHER, + N_("Analysis N"), PIVOT_RC_COUNT); - /* Outline the box */ - tab_box (t, - TAL_2, TAL_2, - -1, -1, - 0, 0, - nc - 1, nr - 1); - - /* Vertical lines */ - tab_box (t, - -1, -1, - -1, TAL_1, - heading_columns, 0, - nc - 1, nr - 1); - - tab_hline (t, TAL_1, 0, nc - 1, heading_rows); - tab_vline (t, TAL_2, heading_columns, 0, nr - 1); - - tab_text (t, 1, 0, TAB_CENTER | TAT_TITLE, _("Mean")); - tab_text (t, 2, 0, TAB_CENTER | TAT_TITLE, _("Std. Deviation")); - tab_text (t, 3, 0, TAB_CENTER | TAT_TITLE, _("Analysis N")); + struct pivot_dimension *variables = pivot_dimension_create ( + table, PIVOT_AXIS_ROW, N_("Variables")); for (i = 0 ; i < factor->n_vars; ++i) { const struct variable *v = factor->vars[i]; - tab_text (t, 0, i + heading_rows, TAB_LEFT | TAT_TITLE, var_to_string (v)); - tab_double (t, 1, i + heading_rows, 0, gsl_matrix_get (idata->mm.mean_matrix, i, i), NULL, RC_OTHER); - tab_double (t, 2, i + heading_rows, 0, sqrt (gsl_matrix_get (idata->mm.var_matrix, i, i)), NULL, RC_OTHER); - tab_double (t, 3, i + heading_rows, 0, gsl_matrix_get (idata->mm.n, i, i), NULL, RC_WEIGHT); + int row = pivot_category_create_leaf ( + variables->root, pivot_value_new_variable (v)); + + double entries[] = { + gsl_matrix_get (idata->mm.mean_matrix, i, i), + sqrt (gsl_matrix_get (idata->mm.var_matrix, i, i)), + gsl_matrix_get (idata->mm.n, i, i), + }; + for (size_t j = 0; j < sizeof entries / sizeof *entries; j++) + pivot_table_put2 (table, j, row, + pivot_value_new_number (entries[j])); } - tab_submit (t); + pivot_table_submit (table); } if (factor->print & PRINT_KMO) { - int i; - double sum_ssq_r = 0; - double sum_ssq_a = 0; - - double df = factor->n_vars * (factor->n_vars - 1) / 2; - - double w = 0; - - - double xsq; - - const int heading_columns = 2; - const int heading_rows = 0; - - const int nr = heading_rows + 4; - const int nc = heading_columns + 1; - - gsl_matrix *a, *x; - - struct tab_table *t = tab_create (nc, nr); - tab_title (t, _("KMO and Bartlett's Test")); - - x = clone_matrix (idata->mm.corr); - gsl_linalg_cholesky_decomp (x); - gsl_linalg_cholesky_invert (x); - - a = anti_image (x); - - for (i = 0; i < x->size1; ++i) - { - sum_ssq_r += ssq_od_n (x, i); - sum_ssq_a += ssq_od_n (a, i); - } - - gsl_matrix_free (a); - gsl_matrix_free (x); - - tab_headers (t, heading_columns, 0, heading_rows, 0); - - /* Outline the box */ - tab_box (t, - TAL_2, TAL_2, - -1, -1, - 0, 0, - nc - 1, nr - 1); - - tab_vline (t, TAL_2, heading_columns, 0, nr - 1); - - tab_text (t, 0, 0, TAT_TITLE | TAB_LEFT, _("Kaiser-Meyer-Olkin Measure of Sampling Adequacy")); - - tab_double (t, 2, 0, 0, sum_ssq_r / (sum_ssq_r + sum_ssq_a), NULL, RC_OTHER); - - tab_text (t, 0, 1, TAT_TITLE | TAB_LEFT, _("Bartlett's Test of Sphericity")); - - tab_text (t, 1, 1, TAT_TITLE, _("Approx. Chi-Square")); - tab_text (t, 1, 2, TAT_TITLE, _("df")); - tab_text (t, 1, 3, TAT_TITLE, _("Sig.")); - + struct pivot_table *table = pivot_table_create ( + N_("KMO and Bartlett's Test")); + + struct pivot_dimension *statistics = pivot_dimension_create ( + table, PIVOT_AXIS_ROW, N_("Statistics"), + N_("Kaiser-Meyer-Olkin Measure of Sampling Adequacy"), PIVOT_RC_OTHER); + pivot_category_create_group ( + statistics->root, N_("Bartlett's Test of Sphericity"), + N_("Approx. Chi-Square"), PIVOT_RC_OTHER, + N_("df"), PIVOT_RC_INTEGER, + N_("Sig."), PIVOT_RC_SIGNIFICANCE); /* The literature doesn't say what to do for the value of W when missing values are involved. The best thing I can think of is to take the mean average. */ - w = 0; + double w = 0; for (i = 0; i < idata->mm.n->size1; ++i) w += gsl_matrix_get (idata->mm.n, i, i); w /= idata->mm.n->size1; - xsq = w - 1 - (2 * factor->n_vars + 5) / 6.0; - xsq *= -log (idata->detR); - - tab_double (t, 2, 1, 0, xsq, NULL, RC_OTHER); - tab_double (t, 2, 2, 0, df, NULL, RC_INTEGER); - tab_double (t, 2, 3, 0, gsl_cdf_chisq_Q (xsq, df), NULL, RC_PVALUE); - - - tab_submit (t); + double xsq = ((w - 1 - (2 * factor->n_vars + 5) / 6.0) + * -log (idata->detR)); + double df = factor->n_vars * (factor->n_vars - 1) / 2; + double entries[] = { + sum_ssq_r / (sum_ssq_r + sum_ssq_a), + xsq, + df, + gsl_cdf_chisq_Q (xsq, df) + }; + for (size_t i = 0; i < sizeof entries / sizeof *entries; i++) + pivot_table_put1 (table, i, pivot_value_new_number (entries[i])); + + pivot_table_submit (table); } show_correlation_matrix (factor, idata); @@ -2438,7 +2175,7 @@ do_factor_by_matrix (const struct cmd_factor *factor, struct idata *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) + if (factor->extraction == EXTRACTION_PAF) { gsl_vector *diff = gsl_vector_alloc (idata->msr->size); struct smr_workspace *ws = ws_create (idata->analysis_matrix); @@ -2464,7 +2201,7 @@ do_factor_by_matrix (const struct cmd_factor *factor, struct idata *idata) gsl_vector_minmax (diff, &min, &max); - if ( fabs (min) < factor->econverge && fabs (max) < factor->econverge) + if (fabs (min) < factor->econverge && fabs (max) < factor->econverge) break; } gsl_vector_free (diff); @@ -2488,10 +2225,10 @@ do_factor_by_matrix (const struct cmd_factor *factor, struct idata *idata) } + show_aic (factor, idata); show_communalities (factor, initial_communalities, extracted_communalities); - - if ( factor->rotation != ROT_NONE) + if (factor->rotation != ROT_NONE) { rotated_factors = gsl_matrix_calloc (factor_matrix->size1, factor_matrix->size2); rotated_loadings = gsl_vector_calloc (factor_matrix->size2); @@ -2512,27 +2249,31 @@ do_factor_by_matrix (const struct cmd_factor *factor, struct idata *idata) show_scree (factor, idata); show_factor_matrix (factor, idata, - factor->extraction == EXTRACTION_PC ? _("Component Matrix") : _("Factor Matrix"), + (factor->extraction == EXTRACTION_PC + ? N_("Component Matrix") : N_("Factor Matrix")), factor_matrix); - if ( factor->rotation == ROT_PROMAX) + if (factor->rotation == ROT_PROMAX) { - show_factor_matrix (factor, idata, _("Pattern Matrix"), pattern_matrix); + show_factor_matrix (factor, idata, N_("Pattern Matrix"), + pattern_matrix); gsl_matrix_free (pattern_matrix); } - if ( factor->rotation != ROT_NONE) + if (factor->rotation != ROT_NONE) { show_factor_matrix (factor, idata, - (factor->rotation == ROT_PROMAX) ? _("Structure Matrix") : - (factor->extraction == EXTRACTION_PC ? _("Rotated Component Matrix") : - _("Rotated Factor Matrix")), + (factor->rotation == ROT_PROMAX + ? N_("Structure Matrix") + : factor->extraction == EXTRACTION_PC + ? N_("Rotated Component Matrix") + : N_("Rotated Factor Matrix")), rotated_factors); gsl_matrix_free (rotated_factors); } - if ( factor->rotation == ROT_PROMAX) + if (factor->rotation == ROT_PROMAX) { show_factor_correlation (factor, fcm); gsl_matrix_free (fcm);