-/* PSPP - RANK. -*-c-*-
-Copyright (C) 2005, 2006, 2007 Free Software Foundation, Inc.
+/* PSPP - a program for statistical analysis.
+ Copyright (C) 2005, 2006, 2007, 2009 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. */
+ You should have received a copy of the GNU General Public License
+ along with this program. If not, see <http://www.gnu.org/licenses/>. */
#include <config.h>
-#include "sort-criteria.h"
+#include <limits.h>
+#include <math.h>
+#include <data/case.h>
+#include <data/casegrouper.h>
+#include <data/casereader.h>
+#include <data/casewriter.h>
#include <data/dictionary.h>
#include <data/format.h>
#include <data/missing-values.h>
#include <data/procedure.h>
+#include <data/short-names.h>
+#include <data/subcase.h>
#include <data/variable.h>
-#include <data/case.h>
-#include <data/casefile.h>
-#include <data/fastfile.h>
-#include <data/storage-stream.h>
#include <language/command.h>
#include <language/stats/sort-criteria.h>
-#include <limits.h>
#include <libpspp/compiler.h>
+#include <libpspp/taint.h>
#include <math/sort.h>
-#include <output/table.h>
#include <output/manager.h>
+#include <output/table.h>
#include <gsl/gsl_cdf.h>
-#include <math.h>
#include "gettext.h"
#define _(msgid) gettext (msgid)
static struct rank_spec *rank_specs;
static size_t n_rank_specs;
-static struct sort_criteria *sc;
+static struct subcase sc;
-static struct variable **group_vars;
+static const struct variable **group_vars;
static size_t n_group_vars;
-static struct variable **src_vars;
+static const struct variable **src_vars;
static size_t n_src_vars;
static struct cmd_rank cmd;
-static struct casefile *rank_sorted_casefile (struct casefile *cf,
- const struct sort_criteria *,
- const struct dictionary *,
- const struct rank_spec *rs,
- int n_rank_specs,
- int idx,
- const struct missing_values *miss
- );
+static void rank_sorted_file (struct casereader *,
+ struct casewriter *,
+ const struct dictionary *,
+ const struct rank_spec *rs,
+ int n_rank_specs,
+ int idx,
+ const struct variable *rank_var);
+
static const char *
fraction_name(void)
{
static bool
-rank_cmd (struct dataset *ds, const struct sort_criteria *sc,
+rank_cmd (struct dataset *ds, const struct subcase *sc,
const struct rank_spec *rank_specs, int n_rank_specs)
{
- struct sort_criteria criteria;
- bool result = true;
+ struct dictionary *d = dataset_dict (ds);
+ bool ok = true;
int i;
- const int n_splits = dict_get_split_cnt (dataset_dict (ds));
- criteria.crit_cnt = n_splits + n_group_vars + 1;
- criteria.crits = xnmalloc (criteria.crit_cnt, sizeof *criteria.crits);
- for (i = 0; i < n_splits ; i++)
- {
- struct variable *v = dict_get_split_vars (dataset_dict (ds))[i];
- criteria.crits[i].fv = var_get_case_index (v);
- criteria.crits[i].width = var_get_width (v);
- criteria.crits[i].dir = SRT_ASCEND;
- }
- for (i = 0; i < n_group_vars; i++)
- {
- criteria.crits[i + n_splits].fv = var_get_case_index (group_vars[i]);
- criteria.crits[i + n_splits].width = var_get_width (group_vars[i]);
- criteria.crits[i + n_splits].dir = SRT_ASCEND;
- }
- for (i = 0 ; i < sc->crit_cnt ; ++i )
+ for (i = 0 ; i < subcase_get_n_fields (sc) ; ++i )
{
- struct casefile *out ;
- struct casefile *cf ;
- struct casereader *reader ;
- struct casefile *sorted_cf ;
-
- /* Obtain active file in CF. */
- if (!procedure (ds, NULL, NULL))
- goto error;
-
- cf = proc_capture_output (ds);
-
- /* Sort CF into SORTED_CF. */
- reader = casefile_get_destructive_reader (cf) ;
- criteria.crits[criteria.crit_cnt - 1] = sc->crits[i];
- assert ( sc->crits[i].fv == var_get_case_index (src_vars[i]) );
- sorted_cf = sort_execute (reader, &criteria, NULL);
- casefile_destroy (cf);
-
- out = rank_sorted_casefile (sorted_cf, &criteria,
- dataset_dict (ds),
- rank_specs, n_rank_specs,
- i, var_get_missing_values (src_vars[i]));
- if ( NULL == out )
- {
- result = false ;
- continue ;
- }
+ /* Rank variable at index I in SC. */
+ struct casegrouper *split_grouper;
+ struct casereader *split_group;
+ struct casewriter *output;
- proc_set_source (ds, storage_source_create (out));
- }
+ proc_discard_output (ds);
+ split_grouper = casegrouper_create_splits (proc_open (ds), d);
+ output = autopaging_writer_create (dict_get_proto (d));
- free (criteria.crits);
- return result ;
+ while (casegrouper_get_next_group (split_grouper, &split_group))
+ {
+ struct subcase ordering;
+ struct casereader *ordered;
+ struct casegrouper *by_grouper;
+ struct casereader *by_group;
+
+ /* Sort this split group by the BY variables as primary
+ keys and the rank variable as secondary key. */
+ subcase_init_vars (&ordering, group_vars, n_group_vars);
+ subcase_add_var (&ordering, src_vars[i],
+ subcase_get_direction (sc, i));
+ ordered = sort_execute (split_group, &ordering);
+ subcase_destroy (&ordering);
+
+ /* Rank the rank variable within this split group. */
+ by_grouper = casegrouper_create_vars (ordered,
+ group_vars, n_group_vars);
+ while (casegrouper_get_next_group (by_grouper, &by_group))
+ {
+ /* Rank the rank variable within this BY group
+ within the split group. */
-error:
- free (criteria.crits);
- return false ;
+ rank_sorted_file (by_group, output, d, rank_specs, n_rank_specs,
+ i, src_vars[i]);
+ }
+ ok = casegrouper_destroy (by_grouper) && ok;
+ }
+ ok = casegrouper_destroy (split_grouper);
+ ok = proc_commit (ds) && ok;
+ ok = (proc_set_active_file_data (ds, casewriter_make_reader (output))
+ && ok);
+ if (!ok)
+ break;
+ }
+
+ return ok;
}
/* Hardly a rank function !! */
int i, double w UNUSED)
{
double rank;
+
if ( c >= 1.0 )
{
switch (cmd.ties)
NOT_REACHED();
}
-
-/* Rank the casefile belonging to CR, starting from the current
- postition of CR continuing up to and including the ENDth case.
-
- RS points to an array containing the rank specifications to
- use. N_RANK_SPECS is the number of elements of RS.
-
-
- DEST_VAR_INDEX is the index into the rank_spec destvar element
- to be used for this ranking.
-
- Prerequisites: 1. The casefile must be sorted according to CRITERION.
- 2. W is the sum of the non-missing caseweights for this
- range of the casefile.
-*/
static void
-rank_cases (struct casereader *cr,
- unsigned long end,
- const struct dictionary *dict,
- const struct sort_criterion *criterion,
- const struct missing_values *mv,
- double w,
- const struct rank_spec *rs,
- int n_rank_specs,
- int dest_var_index,
- struct casefile *dest)
+rank_sorted_file (struct casereader *input,
+ struct casewriter *output,
+ const struct dictionary *dict,
+ const struct rank_spec *rs,
+ int n_rank_specs,
+ int dest_idx,
+ const struct variable *rank_var)
{
- bool warn = true;
+ struct casereader *pass1, *pass2, *pass2_1;
+ struct casegrouper *tie_grouper;
+ struct ccase *c;
+ double w = 0.0;
double cc = 0.0;
- double cc_1;
- int iter = 1;
+ int tie_group = 1;
- const int fv = criterion->fv;
- const int width = criterion->width;
- while (casereader_cnum (cr) < end)
- {
- struct casereader *lookahead;
- const union value *this_value;
- struct ccase this_case, lookahead_case;
- double c;
- int i;
- size_t n = 0;
+ input = casereader_create_filter_missing (input, &rank_var, 1,
+ exclude_values, NULL, output);
+ input = casereader_create_filter_weight (input, dict, NULL, output);
- if (!casereader_read_xfer (cr, &this_case))
- break;
+ casereader_split (input, &pass1, &pass2);
- this_value = case_data_idx (&this_case, fv);
- c = dict_get_case_weight (dict, &this_case, &warn);
-
- lookahead = casereader_clone (cr);
- n = 0;
- while (casereader_cnum (lookahead) < end
- && casereader_read_xfer (lookahead, &lookahead_case))
- {
- const union value *lookahead_value = case_data_idx (&lookahead_case, fv);
- int diff = compare_values (this_value, lookahead_value, width);
-
- if (diff != 0)
- {
- /* Make sure the casefile was sorted */
- assert ( diff == ((criterion->dir == SRT_ASCEND) ? -1 :1));
-
- case_destroy (&lookahead_case);
- break;
- }
+ /* Pass 1: Get total group weight. */
+ for (; (c = casereader_read (pass1)) != NULL; case_unref (c))
+ w += dict_get_case_weight (dict, c, NULL);
+ casereader_destroy (pass1);
- c += dict_get_case_weight (dict, &lookahead_case, &warn);
- case_destroy (&lookahead_case);
- n++;
- }
- casereader_destroy (lookahead);
-
- cc_1 = cc;
- if ( !mv_is_value_missing (mv, this_value, exclude_values) )
- cc += c;
-
- do
- {
- for (i = 0; i < n_rank_specs; ++i)
- {
- const struct variable *dst_var = rs[i].destvars[dest_var_index];
-
- if ( mv_is_value_missing (mv, this_value, exclude_values) )
- case_data_rw (&this_case, dst_var)->f = SYSMIS;
- else
- case_data_rw (&this_case, dst_var)->f =
- rank_func[rs[i].rfunc](c, cc, cc_1, iter, w);
- }
- casefile_append_xfer (dest, &this_case);
- }
- while (n-- > 0 && casereader_read_xfer (cr, &this_case));
-
- if ( !mv_is_value_missing (mv, this_value, exclude_values) )
- iter++;
- }
-
- /* If this isn't true, then all the results will be wrong */
- assert ( w == cc );
-}
-
-static bool
-same_group (const struct ccase *a, const struct ccase *b,
- const struct sort_criteria *crit)
-{
- size_t i;
-
- for (i = 0; i < crit->crit_cnt - 1; i++)
+ /* Pass 2: Do ranking. */
+ tie_grouper = casegrouper_create_vars (pass2, &rank_var, 1);
+ while (casegrouper_get_next_group (tie_grouper, &pass2_1))
{
- struct sort_criterion *c = &crit->crits[i];
- if (compare_values (case_data_idx (a, c->fv),
- case_data_idx (b, c->fv), c->width) != 0)
- return false;
- }
-
- return true;
-}
-
-static struct casefile *
-rank_sorted_casefile (struct casefile *cf,
- const struct sort_criteria *crit,
- const struct dictionary *dict,
- const struct rank_spec *rs,
- int n_rank_specs,
- int dest_idx,
- const struct missing_values *mv)
-{
- struct casefile *dest = fastfile_create (casefile_get_value_cnt (cf));
- struct casereader *lookahead = casefile_get_reader (cf, NULL);
- struct casereader *pos = casereader_clone (lookahead);
- struct ccase group_case;
- bool warn = true;
-
- struct sort_criterion *ultimate_crit = &crit->crits[crit->crit_cnt - 1];
+ struct casereader *pass2_2;
+ double cc_1 = cc;
+ double tw = 0.0;
+ int i;
- if (casereader_read (lookahead, &group_case))
- {
- struct ccase this_case;
- const union value *this_value ;
- double w = 0.0;
- this_value = case_data_idx( &group_case, ultimate_crit->fv);
+ pass2_2 = casereader_clone (pass2_1);
+ taint_propagate (casereader_get_taint (pass2_2),
+ casewriter_get_taint (output));
- if ( !mv_is_value_missing (mv, this_value, exclude_values) )
- w = dict_get_case_weight (dict, &group_case, &warn);
+ /* Pass 2.1: Sum up weight for tied cases. */
+ for (; (c = casereader_read (pass2_1)) != NULL; case_unref (c))
+ tw += dict_get_case_weight (dict, c, NULL);
+ cc += tw;
+ casereader_destroy (pass2_1);
- while (casereader_read (lookahead, &this_case))
+ /* Pass 2.2: Rank tied cases. */
+ while ((c = casereader_read (pass2_2)) != NULL)
{
- const union value *this_value =
- case_data_idx(&this_case, ultimate_crit->fv);
- double c = dict_get_case_weight (dict, &this_case, &warn);
- if (!same_group (&group_case, &this_case, crit))
+ c = case_unshare (c);
+ for (i = 0; i < n_rank_specs; ++i)
{
- rank_cases (pos, casereader_cnum (lookahead) - 1,
- dict,
- ultimate_crit,
- mv, w,
- rs, n_rank_specs,
- dest_idx, dest);
-
- w = 0.0;
- case_destroy (&group_case);
- case_move (&group_case, &this_case);
+ const struct variable *dst_var = rs[i].destvars[dest_idx];
+ double *dst_value = &case_data_rw (c, dst_var)->f;
+ *dst_value = rank_func[rs[i].rfunc] (tw, cc, cc_1, tie_group, w);
}
- if ( !mv_is_value_missing (mv, this_value, exclude_values) )
- w += c;
- case_destroy (&this_case);
+ casewriter_write (output, c);
}
- case_destroy (&group_case);
- rank_cases (pos, ULONG_MAX, dict, ultimate_crit, mv, w,
- rs, n_rank_specs, dest_idx, dest);
- }
+ casereader_destroy (pass2_2);
- if (casefile_error (dest))
- {
- casefile_destroy (dest);
- dest = NULL;
+ tie_group++;
}
-
- casefile_destroy (cf);
- return dest;
+ casegrouper_destroy (tie_grouper);
}
-
/* Transformation function to enumerate all the cases */
static int
-create_resort_key (void *key_var_, struct ccase *cc, casenumber case_num)
+create_resort_key (void *key_var_, struct ccase **cc, casenumber case_num)
{
struct variable *key_var = key_var_;
- case_data_rw(cc, key_var)->f = case_num;
+ *cc = case_unshare (*cc);
+ case_data_rw (*cc, key_var)->f = case_num;
return TRNS_CONTINUE;
}
rank_specs = NULL;
n_rank_specs = 0;
- sort_destroy_criteria (sc);
- sc = NULL;
+ subcase_destroy (&sc);
free (src_vars);
src_vars = NULL;
size_t i;
n_rank_specs = 0;
+ subcase_init_empty (&sc);
if ( !parse_rank (lexer, ds, &cmd, NULL) )
{
rank_cleanup ();
rank_specs = xmalloc (sizeof (*rank_specs));
rank_specs[0].rfunc = RANK;
rank_specs[0].destvars =
- xcalloc (sc->crit_cnt, sizeof (struct variable *));
+ xcalloc (subcase_get_n_fields (&sc), sizeof (struct variable *));
n_rank_specs = 1;
}
- assert ( sc->crit_cnt == n_src_vars);
+ assert ( subcase_get_n_fields (&sc) == n_src_vars);
/* Create variables for all rank destinations which haven't
already been created with INTO.
if ( rank_specs[i].rfunc == NORMAL ||
rank_specs[i].rfunc == PROPORTION )
- tab_output_text (TAT_PRINTF,
- _("%s into %s(%s of %s using %s BY %s)"),
- var_get_name (src_vars[v]),
- var_get_name (rank_specs[i].destvars[v]),
- function_name[rank_specs[i].rfunc],
- var_get_name (src_vars[v]),
- fraction_name(),
- ds_cstr (&varlist)
- );
+ tab_output_text_format (0,
+ _("%s into %s(%s of %s using %s BY %s)"),
+ var_get_name (src_vars[v]),
+ var_get_name (rank_specs[i].destvars[v]),
+ function_name[rank_specs[i].rfunc],
+ var_get_name (src_vars[v]),
+ fraction_name(),
+ ds_cstr (&varlist));
else
- tab_output_text (TAT_PRINTF,
- _("%s into %s(%s of %s BY %s)"),
- var_get_name (src_vars[v]),
- var_get_name (rank_specs[i].destvars[v]),
- function_name[rank_specs[i].rfunc],
- var_get_name (src_vars[v]),
- ds_cstr (&varlist)
- );
+ tab_output_text_format (0,
+ _("%s into %s(%s of %s BY %s)"),
+ var_get_name (src_vars[v]),
+ var_get_name (rank_specs[i].destvars[v]),
+ function_name[rank_specs[i].rfunc],
+ var_get_name (src_vars[v]),
+ ds_cstr (&varlist));
ds_destroy (&varlist);
}
else
{
if ( rank_specs[i].rfunc == NORMAL ||
rank_specs[i].rfunc == PROPORTION )
- tab_output_text (TAT_PRINTF,
- _("%s into %s(%s of %s using %s)"),
- var_get_name (src_vars[v]),
- var_get_name (rank_specs[i].destvars[v]),
- function_name[rank_specs[i].rfunc],
- var_get_name (src_vars[v]),
- fraction_name()
- );
+ tab_output_text_format (0,
+ _("%s into %s(%s of %s using %s)"),
+ var_get_name (src_vars[v]),
+ var_get_name (rank_specs[i].destvars[v]),
+ function_name[rank_specs[i].rfunc],
+ var_get_name (src_vars[v]),
+ fraction_name());
else
- tab_output_text (TAT_PRINTF,
- _("%s into %s(%s of %s)"),
- var_get_name (src_vars[v]),
- var_get_name (rank_specs[i].destvars[v]),
- function_name[rank_specs[i].rfunc],
- var_get_name (src_vars[v])
- );
+ tab_output_text_format (0,
+ _("%s into %s(%s of %s)"),
+ var_get_name (src_vars[v]),
+ var_get_name (rank_specs[i].destvars[v]),
+ function_name[rank_specs[i].rfunc],
+ var_get_name (src_vars[v]));
}
}
}
add_transformation (ds, create_resort_key, 0, order);
/* Do the ranking */
- result = rank_cmd (ds, sc, rank_specs, n_rank_specs);
+ result = rank_cmd (ds, &sc, rank_specs, n_rank_specs);
- /* Put the active file back in its original order */
+ /* Put the active file back in its original order. Delete
+ our sort key, which we don't need anymore. */
{
- struct sort_criteria criteria;
- struct sort_criterion restore_criterion ;
- restore_criterion.fv = var_get_case_index (order);
- restore_criterion.width = 0;
- restore_criterion.dir = SRT_ASCEND;
+ struct casereader *sorted;
- criteria.crits = &restore_criterion;
- criteria.crit_cnt = 1;
+ /* FIXME: loses error conditions. */
- sort_active_file_in_place (ds, &criteria);
- }
+ proc_discard_output (ds);
+ sorted = sort_execute_1var (proc_open (ds), order);
+ result = proc_commit (ds) && result;
- /* ... and we don't need our sort key anymore. So delete it */
- dict_delete_var (dataset_dict (ds), order);
+ dict_delete_var (dataset_dict (ds), order);
+ result = proc_set_active_file_data (ds, sorted) && result;
+ }
rank_cleanup();
static int
rank_custom_variables (struct lexer *lexer, struct dataset *ds, struct cmd_rank *cmd UNUSED, void *aux UNUSED)
{
- static const int terminators[2] = {T_BY, 0};
-
lex_match (lexer, '=');
if ((lex_token (lexer) != T_ID || dict_lookup_var (dataset_dict (ds), lex_tokid (lexer)) == NULL)
&& lex_token (lexer) != T_ALL)
return 2;
- sc = sort_parse_criteria (lexer, dataset_dict (ds),
- &src_vars, &n_src_vars, 0, terminators);
+ if (!parse_sort_criteria (lexer, dataset_dict (ds), &sc, &src_vars, NULL))
+ return 0;
+ n_src_vars = subcase_get_n_fields (&sc);
if ( lex_match (lexer, T_BY) )
{
return 2;
}
- if (!parse_variables (lexer, dataset_dict (ds),
+ if (!parse_variables_const (lexer, dataset_dict (ds),
&group_vars, &n_group_vars,
PV_NO_DUPLICATE | PV_NO_SCRATCH) )
{
rank_specs[n_rank_specs - 1].destvars = NULL;
rank_specs[n_rank_specs - 1].destvars =
- xcalloc (sc->crit_cnt, sizeof (struct variable *));
+ xcalloc (subcase_get_n_fields (&sc), sizeof (struct variable *));
if (lex_match_id (lexer, "INTO"))
{
msg(SE, _("Variable %s already exists."), lex_tokid (lexer));
return 0;
}
- if ( var_count >= sc->crit_cnt )
+ if ( var_count >= subcase_get_n_fields (&sc) )
{
msg(SE, _("Too many variables in INTO clause."));
return 0;
return parse_rank_function (lexer, dict, cmd, NTILES);
}
+
+/*
+ Local Variables:
+ mode: c
+ End:
+*/
projects / pspp-builds.git / blobdiff