1 /* PSPP - computes sample statistics.
2 Copyright (C) 1997-9, 2000 Free Software Foundation, Inc.
3 Written by Ben Pfaff <blp@gnu.org>.
5 This program is free software; you can redistribute it and/or
6 modify it under the terms of the GNU General Public License as
7 published by the Free Software Foundation; either version 2 of the
8 License, or (at your option) any later version.
10 This program is distributed in the hope that it will be useful, but
11 WITHOUT ANY WARRANTY; without even the implied warranty of
12 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
13 General Public License for more details.
15 You should have received a copy of the GNU General Public License
16 along with this program; if not, write to the Free Software
17 Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
21 #include <libpspp/message.h>
23 #include <libpspp/alloc.h>
24 #include <data/case.h>
25 #include <language/command.h>
26 #include <libpspp/compiler.h>
27 #include <data/dictionary.h>
28 #include <libpspp/message.h>
29 #include <libpspp/hash.h>
30 #include <language/lexer/lexer.h>
31 #include <libpspp/pool.h>
32 #include <libpspp/str.h>
33 #include <data/variable.h>
34 #include <procedure.h>
37 #define _(msgid) gettext (msgid)
39 /* FIXME: Implement PRINT subcommand. */
41 /* Explains how to recode one value. `from' must be first element. */
44 union value from; /* Original value. */
45 double to; /* Recoded value. */
48 /* Explains how to recode an AUTORECODE variable. */
51 struct variable *src; /* Source variable. */
52 struct variable *dest; /* Target variable. */
53 struct hsh_table *items; /* Hash table of `freq's. */
56 /* AUTORECODE transformation. */
57 struct autorecode_trns
59 struct pool *pool; /* Contains AUTORECODE specs. */
60 struct arc_spec *specs; /* AUTORECODE specifications. */
61 size_t spec_cnt; /* Number of specifications. */
64 /* Descending or ascending sort order. */
71 /* AUTORECODE data. */
74 struct variable **src_vars; /* Source variables. */
75 char **dst_names; /* Target variable names. */
76 struct variable **dst_vars; /* Target variables. */
77 struct hsh_table **src_values; /* `union value's of source vars. */
78 size_t var_cnt; /* Number of variables. */
79 struct pool *src_values_pool; /* Pool used by src_values. */
80 enum direction direction; /* Sort order. */
81 int print; /* Print mapping table if nonzero. */
84 static trns_proc_func autorecode_trns_proc;
85 static trns_free_func autorecode_trns_free;
86 static bool autorecode_proc_func (struct ccase *, void *);
87 static hsh_compare_func compare_alpha_value, compare_numeric_value;
88 static hsh_hash_func hash_alpha_value, hash_numeric_value;
90 static void recode (const struct autorecode_pgm *);
91 static void arc_free (struct autorecode_pgm *);
93 /* Performs the AUTORECODE procedure. */
97 struct autorecode_pgm arc;
103 arc.dst_names = NULL;
105 arc.src_values = NULL;
107 arc.src_values_pool = NULL;
108 arc.direction = ASCENDING;
112 lex_match_id ("VARIABLES");
114 if (!parse_variables (default_dict, &arc.src_vars, &arc.var_cnt,
117 if (!lex_force_match_id ("INTO"))
120 if (!parse_DATA_LIST_vars (&arc.dst_names, &dst_cnt, PV_NONE))
122 if (dst_cnt != arc.var_cnt)
126 msg (SE, _("Source variable count (%u) does not match "
127 "target variable count (%u)."),
128 (unsigned) arc.var_cnt, (unsigned) dst_cnt);
130 for (i = 0; i < dst_cnt; i++)
131 free (arc.dst_names[i]);
132 free (arc.dst_names);
133 arc.dst_names = NULL;
137 while (lex_match ('/'))
138 if (lex_match_id ("DESCENDING"))
139 arc.direction = DESCENDING;
140 else if (lex_match_id ("PRINT"))
144 lex_error (_("expecting end of command"));
148 for (i = 0; i < arc.var_cnt; i++)
152 if (dict_lookup_var (default_dict, arc.dst_names[i]) != NULL)
154 msg (SE, _("Target variable %s duplicates existing variable %s."),
155 arc.dst_names[i], arc.dst_names[i]);
158 for (j = 0; j < i; j++)
159 if (!strcasecmp (arc.dst_names[i], arc.dst_names[j]))
161 msg (SE, _("Duplicate variable name %s among target variables."),
167 arc.src_values_pool = pool_create ();
168 arc.dst_vars = xnmalloc (arc.var_cnt, sizeof *arc.dst_vars);
169 arc.src_values = xnmalloc (arc.var_cnt, sizeof *arc.src_values);
170 for (i = 0; i < dst_cnt; i++)
171 if (arc.src_vars[i]->type == ALPHA)
172 arc.src_values[i] = hsh_create (10, compare_alpha_value,
173 hash_alpha_value, NULL, arc.src_vars[i]);
175 arc.src_values[i] = hsh_create (10, compare_numeric_value,
176 hash_numeric_value, NULL, NULL);
178 ok = procedure (autorecode_proc_func, &arc);
180 for (i = 0; i < arc.var_cnt; i++)
181 arc.dst_vars[i] = dict_create_var_assert (default_dict,
182 arc.dst_names[i], 0);
186 return ok ? CMD_SUCCESS : CMD_CASCADING_FAILURE;
190 return CMD_CASCADING_FAILURE;
194 arc_free (struct autorecode_pgm *arc)
196 free (arc->src_vars);
197 if (arc->dst_names != NULL)
201 for (i = 0; i < arc->var_cnt; i++)
202 free (arc->dst_names[i]);
203 free (arc->dst_names);
205 free (arc->dst_vars);
206 if (arc->src_values != NULL)
210 for (i = 0; i < arc->var_cnt; i++)
211 hsh_destroy (arc->src_values[i]);
212 free (arc->src_values);
214 pool_destroy (arc->src_values_pool);
218 /* AUTORECODE transformation. */
221 recode (const struct autorecode_pgm *arc)
223 struct autorecode_trns *trns;
226 trns = pool_create_container (struct autorecode_trns, pool);
227 trns->specs = pool_nalloc (trns->pool, arc->var_cnt, sizeof *trns->specs);
228 trns->spec_cnt = arc->var_cnt;
229 for (i = 0; i < arc->var_cnt; i++)
231 struct arc_spec *spec = &trns->specs[i];
232 void *const *p = hsh_sort (arc->src_values[i]);
233 int count = hsh_count (arc->src_values[i]);
236 spec->src = arc->src_vars[i];
237 spec->dest = arc->dst_vars[i];
239 if (arc->src_vars[i]->type == ALPHA)
240 spec->items = hsh_create (2 * count, compare_alpha_value,
241 hash_alpha_value, NULL, arc->src_vars[i]);
243 spec->items = hsh_create (2 * count, compare_numeric_value,
244 hash_numeric_value, NULL, NULL);
246 for (j = 0; *p; p++, j++)
248 struct arc_item *item = pool_alloc (trns->pool, sizeof *item);
249 union value *vp = *p;
251 if (arc->src_vars[i]->type == NUMERIC)
252 item->from.f = vp->f;
254 item->from.c = pool_clone (trns->pool, vp->c,
255 arc->src_vars[i]->width);
256 item->to = arc->direction == ASCENDING ? j + 1 : count - j;
257 hsh_force_insert (spec->items, item);
260 add_transformation (autorecode_trns_proc, autorecode_trns_free, trns);
263 /* Executes an AUTORECODE transformation. */
265 autorecode_trns_proc (void *trns_, struct ccase *c, int case_idx UNUSED)
267 struct autorecode_trns *trns = trns_;
270 for (i = 0; i < trns->spec_cnt; i++)
272 struct arc_spec *spec = &trns->specs[i];
273 struct arc_item *item;
276 if (spec->src->type == NUMERIC)
277 v.f = case_num (c, spec->src->fv);
279 v.c = (char *) case_str (c, spec->src->fv);
280 item = hsh_force_find (spec->items, &v);
282 case_data_rw (c, spec->dest->fv)->f = item->to;
284 return TRNS_CONTINUE;
287 /* Frees an AUTORECODE transformation. */
289 autorecode_trns_free (void *trns_)
291 struct autorecode_trns *trns = trns_;
294 for (i = 0; i < trns->spec_cnt; i++)
295 hsh_destroy (trns->specs[i].items);
296 pool_destroy (trns->pool);
300 /* AUTORECODE procedure. */
303 compare_alpha_value (const void *a_, const void *b_, void *v_)
305 const union value *a = a_;
306 const union value *b = b_;
307 const struct variable *v = v_;
309 return memcmp (a->c, b->c, v->width);
313 hash_alpha_value (const void *a_, void *v_)
315 const union value *a = a_;
316 const struct variable *v = v_;
318 return hsh_hash_bytes (a->c, v->width);
322 compare_numeric_value (const void *a_, const void *b_, void *foo UNUSED)
324 const union value *a = a_;
325 const union value *b = b_;
327 return a->f < b->f ? -1 : a->f > b->f;
331 hash_numeric_value (const void *a_, void *foo UNUSED)
333 const union value *a = a_;
335 return hsh_hash_double (a->f);
339 autorecode_proc_func (struct ccase *c, void *arc_)
341 struct autorecode_pgm *arc = arc_;
344 for (i = 0; i < arc->var_cnt; i++)
346 union value v, *vp, **vpp;
348 if (arc->src_vars[i]->type == NUMERIC)
349 v.f = case_num (c, arc->src_vars[i]->fv);
351 v.c = (char *) case_str (c, arc->src_vars[i]->fv);
353 vpp = (union value **) hsh_probe (arc->src_values[i], &v);
356 vp = pool_alloc (arc->src_values_pool, sizeof *vp);
357 if (arc->src_vars[i]->type == NUMERIC)
360 vp->c = pool_clone (arc->src_values_pool,
361 v.c, arc->src_vars[i]->width);