1 2007-09-19 Ben Pfaff <blp@gnu.org>
5 * aggregate.c (cmd_aggregate): Destroy casereader consistently,
6 even if casereader fails.
8 * examine.q (run_examine): Ditto.
10 * glm.q (run_glm): Ditto.
12 * oneway.q (run_oneway): Ditto.
14 * regression.q (run_regression): Ditto.
16 * t-test.q (calculate): Ditto.
18 * descriptives.c (calc_descriptives): Ditto. Also avoid
19 gratuitous casereader_clone.
21 2007-09-13 Jason Stover <jhs@debs.hoobahooba.net>
23 * regression.q (cmd_regression): Move declaration of models in to
24 definition of cmd_regression.
26 * regression.q (run_regression): Free mom to fix memory leak.
28 2007-09-12 Ben Pfaff <blp@gnu.org>
30 * crosstabs.q (postcalc): Free sorted_tab and the structures that
31 it points to, to plug a memory leak. Fixes bug #20910. Thanks to
32 John Darrington for reporting this bug and for review.
34 2007-09-04 Ben Pfaff <blp@gnu.org>
36 * crosstabs.q (cmd_crosstabs): Free xtab and the structures that
37 it points to, to plug a memory leak. Fixes bug #18315.
39 2007-08-15 Jason Stover <jhs@wonko.gcsu.edu>
41 * regression.q (identify_indep_vars): Print an error if dependent
42 and independent variables are the same. Fixes bug 19819.
44 2007-08-12 Ben Pfaff <blp@gnu.org>
46 * flip.c: Drop use of dict_get_compacted_dict_index_to_case_index
47 and just use the ordinary case indexes. There seemed to be no
48 reason for the former method.
50 2007-08-03 Ben Pfaff <blp@gnu.org>
52 * rank.q (rank_cmd): Instead of sorting by SPLIT FILE vars, group
53 by them. Fixes bug #17239.
54 Reviewed by John Darrington.
56 2007-08-01 Ben Pfaff <blp@gnu.org>
58 Clean up handling of median, by treating it almost like any other
59 percentile. Fixes bug #17424. Thanks to John Darrington for
61 * frequencies.q (internal_cmd_frequencies): Fix handling of bit
62 masks for `stats' variable. If median is selected, turn it off
63 and add a 50th percentile.
64 (add_percentile): Simplify code a little.
65 (calc_stats): Drop special casing of median.
66 (dump_statistics): Ditto, except that we label the 50th percentile
67 as "50 (Median)" to make it clear that it's also the median.
69 2007-07-31 Ben Pfaff <blp@gnu.org>
71 Remove integer mode from FREQUENCIES and incidentally fix bug
72 #17421. Reviewed by John Darrington.
73 * frequencies.q (int_pool): Rename data_pool.
74 (gen_pool): Rename syntax_pool.
75 (enum FRQM_*): Removed.
76 (struct freq_tab): Removed `mode', `vector', `min', `max',
77 `out_of_range', `sysmis' members.
78 (calc): Delete support for integer mode.
80 (postprocess_freq_tab): Ditto.
81 (cleanup_freq_tab): Ditto.
82 (frq_custom_variables): Ditto.
84 2007-07-28 John Darrington <john@darrington.wattle.id.au>
86 * t-test.q: Moved the order in which groups are displayed in the
87 independent samples case, where a cut point is given.
89 2007-07-27 Ben Pfaff <blp@gnu.org>
91 * regression.q (run_regression): Move casereader_destroy call so
92 that it always gets called, not just if there was some valid
93 data. Fixes bug #19581.
94 Reviewed by Jason Stover.
96 2007-07-24 Ben Pfaff <blp@gnu.org>
98 * flip.c (struct flip_pgm): Remove `case_size' member (now
100 (cmd_flip): Pass var_cnt as number of cases instead of case_cnt,
101 to fix bug #20494. Don't assign to `case_size' member, which was
102 unused after assignment.
103 (build_dictionary): When NEWNAMES not used, get the number of
104 variables right, to fix bug #20493.
106 2007-07-10 Jason Stover <jhs@math.gcsu.edu>
108 * glm.q: Initial version of the GLM procedure.
110 2007-06-06 Ben Pfaff <blp@gnu.org>
112 Adapt case sources, sinks, and clients of procedure code to the
115 * aggregate.c: Simplify greatly since everything is more uniform
118 * autorecode.c: Adapt to new procedure code.
120 * chisquare.c: Ditto.
121 * crosstabs.q: Ditto.
122 * descriptives.c: Ditto.
124 * npar-summary.c: Ditto.
125 * frequencies.q: Ditto.
128 * regression.q: Ditto.
129 * sort-cases.c: Ditto.
132 * sort-criteria.c: Rewrite to output a struct case_ordering.
134 * flip.c: Rewrite to be a casereader.
136 * rank.q: Simplify greatly since casereaders are much more
137 flexible than what we had before.
139 2007-05-15 Jason Stover <jhs@math.gcsu.edu>
141 * regression.q (run_regression): Tell the user when the data
142 contain no valid cases.
144 2007-05-08 Jason Stover <jhs@math.gcsu.edu>
146 * regression.q: Partial fix of bug which caused a crash if
147 dependent variable and independent variable were the same.
149 2007-04-16 John Darrington <john@darrington.wattle.id.au>
151 * t-test.q: Changed the output width of reported counts and
152 degrees of freedom, to avoid truncating these values. Thanks
153 to Seth Woolley for reporting this problem. A proper fix involves
154 re-thinking the output driver.
156 2007-04-12 Jason Stover <jhs@math.gcsu.edu>
158 * regression.q (run_regression): Added if (n_data >0) to fix bug
161 2007-03-29 Jason Stover <jhs@math.gcsu.edu>
163 * regression.q (prepare_data): New function.
165 * regression.q (compute_moments): New function.
167 2007-03-18 Ben Pfaff <blp@gnu.org>
169 * crosstabs.q (static var write): Rename write_style to avoid
170 conflict with POSIX function of same name.
172 2007-03-16 Jason Stover <jhs@math.gcsu.edu>
174 * regression.q (run_regression): Added support for moments.
176 Sat Feb 17 08:16:00 2007 Ben Pfaff <blp@gnu.org>
178 * flip.c (flip_sink_create): Improve error message when temporary
179 file cannot be created.
181 Tue Feb 6 19:58:03 2007 Ben Pfaff <blp@gnu.org>
183 * flip.c (flip_file): Give better error message on end-of-file.
185 2007-02-04 Jason Stover <jhs@math.gcsu.edu>
187 * regression.q: Fixed p-value computation in the test for
188 individual regression coefficients.
190 Mon Jan 15 11:03:20 2007 Ben Pfaff <blp@gnu.org>
192 Fix bugs found by valgrind when --enable-debug is used with the
193 new case code. These bugs are hidden when the data set is small
194 enough to find in memory; if a bigger data set that would overflow
195 to disk were used, then data corruption would occur.
197 * chisquare.c (create_freq_hash): Pass free_freq_mutable_hash to
198 hsh_create as free function. Make copy of data put into hash.
200 * oneway.q (free_value): New function.
201 (run_oneway): Use free_value as arg to hsh_create. Make copy of
204 * rank.q (rank_cases): Don't access data in case after we've given
207 Tue Jan 9 19:16:11 2007 Ben Pfaff <blp@gnu.org>
211 * aggregate.c (parse_aggregate_functions) Initialize
214 Fri Dec 22 14:04:09 2006 Ben Pfaff <blp@gnu.org>
216 Simplify missing value handling.
218 * aggregate.c (struct agr_var): Remove `bool include_missing', add
219 `enum mv_class exclude'. Remove `int missing', add `bool
220 saw_missing'. Update users.
222 * descriptives.c (struct dsc_trns): Removed `int
223 include_user_missing', add `enum mv_class exclude'. Update users.
224 (struct dsc_proc): Ditto.
226 * examine.q: (static var value_is_missing): Rename
227 `exclude_values', change type to `enum mv_class'. Update users.
231 Fri Dec 22 19:22:18 WST 2006 John Darrington <john@darrington.wattle.id.au>
233 * frequencies.q : Fixed bug #17420, where the table bounds were overun
234 when /FORMAT=nolabels was given.
236 Wed Dec 20 18:45:31 WST 2006 John Darrington <john@darrington.wattle.id.au>
238 * binomial.c binomial.h : New files. Thanks to Jason Stover
239 for assistance with these.
241 * chisquare.c chisquare.h freq.c freq.h npar-summary.c npar-summary.h
242 npar.h npar.q: New files. Implementing NPAR TESTS.
244 * frequencies.q : Moved structure definitions into freq.[ch]
246 Sat Dec 16 22:26:44 2006 Ben Pfaff <blp@gnu.org>
248 Make it possible to pull cases from the active file with a
249 function call, instead of requiring indirection through a callback
252 * aggregate.c (cmd_aggregate): Take advantage of new procedure
254 (agr_to_active_file): Removed.
255 (presorted_agr_to_sysfile): Removed.
257 * autorecode.c (cmd_autorecode): Take advantage of new procedure
259 (autorecode_proc_func): Removed.
261 * flip.c (struct flip_pgm): New members to allow conformance with
262 new case_source_class interface.
263 (cmd_flip): Adapt to new case_source_class interface.
264 (flip_source_read): Ditto.
265 (flip_source_destroy): Ditto.
267 Sat Dec 16 12:54:27 2006 Ben Pfaff <blp@gnu.org>
269 * rank.q (rank_custom_variables): Allow grouping variables to be
270 strings. Fixes bug #18533. Thanks to John Darrington for review.
272 Sat Dec 9 18:47:51 2006 Ben Pfaff <blp@gnu.org>
274 * regression.q (is_depvar): Compare variable pointers instead of
277 Thu Dec 7 15:26:25 WST 2006 John Darrington <john@darrington.wattle.id.au>
279 * examine.q: Allocated the categorical values for the dependent and
280 independent variables, on the heap. Hence they can be of any width.
282 Wed Dec 6 21:14:26 2006 Ben Pfaff <blp@gnu.org>
284 * regression.q (reg_inserted): Compare variable pointers instead
287 Mon Dec 4 22:33:46 2006 Ben Pfaff <blp@gnu.org>
289 * crosstabs.q (insert_summary): Use var_to_string for labeling.
290 (output_pivot_table) Ditto.
293 * frequencies.q (setup_z_trns): Ditto.
295 (dump_condensed) Ditto.
296 (dump_statistics) Ditto.
299 Sun Nov 5 08:31:42 WST 2006 John Darrington <john@darrington.wattle.id.au>
301 * t-test.q, oneway.q: Changed to use the new casefilter structure.
303 Sat Oct 14 16:52:28 2006 Ben Pfaff <blp@gnu.org>
305 * rank.q: (rank_sorted_casefile) Add some missing case_destroy()
306 calls to fix a memory leak.
308 Sun Oct 8 09:45:40 WST 2006 John Darrington <john@darrington.wattle.id.au>
310 * rank.q: Plugged a small memory leak which occurred under error
313 Sat Oct 7 11:06:01 WST 2006 John Darrington <john@darrington.wattle.id.au>
315 * rank.q: Implemented most of the RANK command.
317 2006-07-14 Jason Stover <jhs@math.gcsu.edu>
319 * regression.q (run_regression): New function to move knowledge of
320 pspp_linreg_cache out of math/coefficient.[ch].
322 Sat Jul 1 17:41:46 2006 Ben Pfaff <blp@gnu.org>
324 Fix bug #11612, "q2c documentation does not agree with code".
326 * examine.q: Audit use of q2c "+" prefixes that indicate that a
327 command may appear multiple times.
329 * frequencies.q: Ditto.
333 * regression.q: Ditto.
337 Fri Jun 23 14:18:22 2006 Ben Pfaff <blp@gnu.org>
339 Support long string variables on FREQUENCIES, as
340 an extension when in enhanced algorithms mode. For Greg Hunt
341 <greg@firmansyah.com>.
343 * frequencies.q: (struct freq) Change `v' member from union value
344 to union value *. Update all references.
345 (struct var_freqs) Add width, print members to represent effective
346 variable width and display format.
347 (calc) Copy entire long string value into the hash table.
348 (frq_custom_variables) Set new width, print members.
349 (hash_value_alpha) Get width from var_freqs.
350 (compare_value_alpha_a) Ditto.
351 (compare_freq_alpha_a) Ditto.
352 (compare_freq_alpha_d) Ditto.
353 (dump_full) Get display format from var_freqs.
354 (dump_condensed) Ditto.
356 Mon Jun 19 22:07:13 2006 Ben Pfaff <blp@gnu.org>
358 * frequencies.q: (dump_full) Only put the first MAX_SHORT_STRING
359 bytes of string variables into the output cells, seeing as we only
361 (dump_condensed) Ditto.
363 Mon Jun 19 21:52:05 2006 Ben Pfaff <blp@gnu.org>
365 Fixes a bug reported by Greg Hunt <greg@firmansyah.com>.
367 * frequencies.q: (hsh_hash_bytes) We only copy the first
368 MAX_SHORT_STRING bytes of string variables, so we must only
369 compare that many bytes, even if the string variable is longer.
370 (compare_value_alpha_a) Ditto.
371 (compare_freq_alpha_a) Ditto.
372 (compare_freq_alpha_d) Ditto.
374 2006-05-11 Jason Stover <jhs@math.gcsu.edu>
376 * regression.q: Adjusted code to account for cache->coeff being a
377 pspp_linreg_coeff **.
379 Sun May 7 18:31:25 2006 Ben Pfaff <blp@gnu.org>
383 * aggregate.c (cmd_aggregate): Free default_dict before replacing
386 Sun May 7 17:09:19 2006 Ben Pfaff <blp@gnu.org>
388 * flip.c (flip_file): Check return value of pool_fclose().
390 Sat May 6 16:00:13 2006 Ben Pfaff <blp@gnu.org>
392 Get rid of `char *c' member in union value, for cleanliness.
394 * aggregate.c: (union agr_argument) New union.
395 (struct agr_var) Change element type of arg[] from union value to
397 (parse_aggregate_functions) Change local variable types likewise.
399 * autorecode.c: (union arc_value) New union.
400 (struct arc_item) Change "from" from union value to union
402 (recode) Change local variable from union value to union
404 (autorecode_trns_proc) Ditto.
405 (compare_alpha_value) Ditto.
406 (hash_alpha_value) Ditto.
407 (compare_numeric_value) Ditto.
408 (hash_numeric_value) Ditto.
409 (autorecode_proc_func) Ditto.
411 Sat May 6 10:43:33 2006 Ben Pfaff <blp@gnu.org>
413 Continue reforming procedure execution. In this phase, get rid of
414 the output code for SPLIT FILE groups in procedure.c, which really
415 shouldn't be doing any output. Move it into the individual
416 procedures instead. This also adds some flexibility.
418 * crosstabs.q (precalc): Call output_split_file_values().
420 * descriptives.c (calc_descriptives): Ditto.
422 * examine.q (run_examine): Ditto.
424 * frequencies.q (precalc): Ditto.
426 * oneway.q (run_oneway): Ditto.
428 * regression.q (run_regression): Ditto.
430 * t-test.q (calculate): Ditto.
432 Wed May 3 23:05:31 2006 Ben Pfaff <blp@gnu.org>
434 Continue reforming procedure execution. In this phase, get rid of
435 many global variables, consolidating procedure execution in
436 procedure.c. Encapsulate transformations in new "struct
437 trns_chain". Also, change implementation of N OF CASES, FILTER,
438 and PROCESS IF from special cases to transformations.
440 * aggregate.c (cmd_aggregate) Use discard_variables().
442 2006-04-28 Jason Stover <jhs@math.gcsu.edu>
444 * regression.q (regression_trns_resid_proc): Pass only the
445 variables used in the model to (*model->residual)().
447 * regression.q (regression_trns_pred_proc): Pass only the
448 variables used in the model to (*model->pred)().
450 2006-04-26 Jason Stover <jhs@math.gcsu.edu>
452 * regression.q: Added support for multiple transformations.
454 * regression.q (regression_trns_resid_proc): New function.
456 * regression.q (regression_trns_pred_proc): New function.
458 * regression.q (subcommand_save): Added support for saving
461 * regression.q (regression_trns_free): New function.
463 * regression.q (reg_get_name): New function.
465 * regression.q (reg_save_var): New function.
467 Tue Apr 25 13:18:56 2006 Ben Pfaff <blp@gnu.org>
469 * rank.q (parse_rank_function): Use SE instead of ME for parse
472 Tue Apr 25 13:16:28 2006 Ben Pfaff <blp@gnu.org>
474 * flip.c (flip_sink_write): Use snprintf() to simplify a bit of
477 2006-04-21 Jason Stover <jhs@math.gcsu.edu>
479 * regression.q (try_name): New function. (Partly copied from
480 try_name in descriptives.c.)
482 * regression.q (subcommand_save): Choose residual variable names
485 2006-04-20 Jason Stover <jhs@math.gcsu.edu>
487 * regression.q (cmd_regression): Moved call to subcommand_save()
488 outside multipass_procedure_with_splits().
490 * regression.q (regression_trns_proc): Fixed value counter n_vals
491 before calling *model->residual().
493 2006-04-19 Jason Stover <jhs@math.gcsu.edu>
495 * regression.q (regression_trns_proc): Fixed the look-up of the
498 2006-04-18 Jason Stover <jhs@math.gcsu.edu>
500 * regression.q (regression_trns_proc): Look up the residual
501 variable in the linear regression cache.
503 * regression.q (subcommand_save): Set the residual variable in the
504 linear regression cache.
506 2006-04-17 Jason Stover <jason@wonko.gcsu.edu>
508 * regression.q (regression_trns_proc): Accept case_idx as an int
509 to match the definition of trns_proc_func.
511 2006-04-17 Jason Stover <jhs@math.gcsu.edu>
513 * regression.q (regression_trns_proc): New function.
515 * regression.q (subcommand_save): Create variable residuals and
516 add a transformation to assign values to them. Also free the
517 linreg_cache if the SAVE command was not called. Removed the
518 casereading loop. Placed actual computation of residuals in
519 regression_trns_proc.
521 * regression.q (run_regression): Moved call to free
522 pspp_linreg_cache to subcommand_save.
524 Sat Apr 15 18:01:03 2006 Ben Pfaff <blp@gnu.org>
526 * examine.q (output_examine): Add casts to fix warnings.
528 2006-04-07 Jason Stover <jhs@math.gcsu.edu>
530 * regression.q (subcommand_save): New function.
532 2006-04-04 Jason Stover <jhs@math.gcsu.edu>
534 * regression.q: New function reg_has_categorical () to tell
535 whether a pspp_linreg_struct was made with any variables of type
538 * regression.q: (subcommand_export): Call
539 reg_print_categorical_encoding() only if the model uses any
540 categorical variables.
542 Mon Mar 27 16:00:42 2006 Ben Pfaff <blp@gnu.org>
544 * crosstabs.q: (output_pivot_table) Drop spurious space from
547 2006-03-15 Jason Stover <jhs@math.gcsu.edu>
549 * regression.q: Added custom syntax parser for VARIABLES subcommand
551 * regression.q: Moved most instructions for run_regression ()
552 inside the loop over dependent variables.
554 Thu Mar 2 08:40:33 WST 2006 John Darrington <john@darrington.wattle.id.au>
556 * Moved files from src directory