1 2007-12-07 Ben Pfaff <blp@gnu.org>
5 * crosstabs.q (precalc): Initialize data structures even if the
6 first case cannot be read.
8 * frequencies.q (precalc): Ditto.
10 2007-11-03 Ben Pfaff <blp@gnu.org>
12 Allow output files to overwrite input files (bug #21280).
14 * aggregate.c (cmd_aggregate): Manage file handle reference
17 * correlations.q (internal_cmd_frequencies): Ditto.
18 (cor_custom_matrix): Ditto.
20 * regression.q (regression_custom_export): Ditto.
21 (cmd_regression): Ditto.
23 2007-10-12 Ben Pfaff <blp@gnu.org>
25 * flip.c (flip_file): No need to conditionally substitute for
26 "fseeko" and "off_t" manually anymore, as gnulib takes care of it
29 2007-09-21 Jason Stover <jhs@wonko.gcsu.edu>
31 * regression.q (run_regression): Partial fix of memory leak, bug
34 2007-09-19 Ben Pfaff <blp@gnu.org>
38 * aggregate.c (cmd_aggregate): Destroy casereader consistently,
39 even if casereader fails.
41 * examine.q (run_examine): Ditto.
43 * glm.q (run_glm): Ditto.
45 * oneway.q (run_oneway): Ditto.
47 * regression.q (run_regression): Ditto.
49 * t-test.q (calculate): Ditto.
51 * descriptives.c (calc_descriptives): Ditto. Also avoid
52 gratuitous casereader_clone.
54 2007-09-13 Jason Stover <jhs@debs.hoobahooba.net>
56 * regression.q (cmd_regression): Move declaration of models in to
57 definition of cmd_regression.
59 * regression.q (run_regression): Free mom to fix memory leak.
61 2007-09-12 Ben Pfaff <blp@gnu.org>
63 * crosstabs.q (postcalc): Free sorted_tab and the structures that
64 it points to, to plug a memory leak. Fixes bug #20910. Thanks to
65 John Darrington for reporting this bug and for review.
67 2007-09-04 Ben Pfaff <blp@gnu.org>
69 * crosstabs.q (cmd_crosstabs): Free xtab and the structures that
70 it points to, to plug a memory leak. Fixes bug #18315.
72 2007-08-15 Jason Stover <jhs@wonko.gcsu.edu>
74 * regression.q (identify_indep_vars): Print an error if dependent
75 and independent variables are the same. Fixes bug 19819.
77 2007-08-12 Ben Pfaff <blp@gnu.org>
79 * flip.c: Drop use of dict_get_compacted_dict_index_to_case_index
80 and just use the ordinary case indexes. There seemed to be no
81 reason for the former method.
83 2007-08-03 Ben Pfaff <blp@gnu.org>
85 * rank.q (rank_cmd): Instead of sorting by SPLIT FILE vars, group
86 by them. Fixes bug #17239.
87 Reviewed by John Darrington.
89 2007-08-01 Ben Pfaff <blp@gnu.org>
91 Clean up handling of median, by treating it almost like any other
92 percentile. Fixes bug #17424. Thanks to John Darrington for
94 * frequencies.q (internal_cmd_frequencies): Fix handling of bit
95 masks for `stats' variable. If median is selected, turn it off
96 and add a 50th percentile.
97 (add_percentile): Simplify code a little.
98 (calc_stats): Drop special casing of median.
99 (dump_statistics): Ditto, except that we label the 50th percentile
100 as "50 (Median)" to make it clear that it's also the median.
102 2007-07-31 Ben Pfaff <blp@gnu.org>
104 Remove integer mode from FREQUENCIES and incidentally fix bug
105 #17421. Reviewed by John Darrington.
106 * frequencies.q (int_pool): Rename data_pool.
107 (gen_pool): Rename syntax_pool.
108 (enum FRQM_*): Removed.
109 (struct freq_tab): Removed `mode', `vector', `min', `max',
110 `out_of_range', `sysmis' members.
111 (calc): Delete support for integer mode.
113 (postprocess_freq_tab): Ditto.
114 (cleanup_freq_tab): Ditto.
115 (frq_custom_variables): Ditto.
117 2007-07-28 John Darrington <john@darrington.wattle.id.au>
119 * t-test.q: Moved the order in which groups are displayed in the
120 independent samples case, where a cut point is given.
122 2007-07-27 Ben Pfaff <blp@gnu.org>
124 * regression.q (run_regression): Move casereader_destroy call so
125 that it always gets called, not just if there was some valid
126 data. Fixes bug #19581.
127 Reviewed by Jason Stover.
129 2007-07-24 Ben Pfaff <blp@gnu.org>
131 * flip.c (struct flip_pgm): Remove `case_size' member (now
133 (cmd_flip): Pass var_cnt as number of cases instead of case_cnt,
134 to fix bug #20494. Don't assign to `case_size' member, which was
135 unused after assignment.
136 (build_dictionary): When NEWNAMES not used, get the number of
137 variables right, to fix bug #20493.
139 2007-07-10 Jason Stover <jhs@math.gcsu.edu>
141 * glm.q: Initial version of the GLM procedure.
143 2007-06-06 Ben Pfaff <blp@gnu.org>
145 Adapt case sources, sinks, and clients of procedure code to the
148 * aggregate.c: Simplify greatly since everything is more uniform
151 * autorecode.c: Adapt to new procedure code.
153 * chisquare.c: Ditto.
154 * crosstabs.q: Ditto.
155 * descriptives.c: Ditto.
157 * npar-summary.c: Ditto.
158 * frequencies.q: Ditto.
161 * regression.q: Ditto.
162 * sort-cases.c: Ditto.
165 * sort-criteria.c: Rewrite to output a struct case_ordering.
167 * flip.c: Rewrite to be a casereader.
169 * rank.q: Simplify greatly since casereaders are much more
170 flexible than what we had before.
172 2007-05-15 Jason Stover <jhs@math.gcsu.edu>
174 * regression.q (run_regression): Tell the user when the data
175 contain no valid cases.
177 2007-05-08 Jason Stover <jhs@math.gcsu.edu>
179 * regression.q: Partial fix of bug which caused a crash if
180 dependent variable and independent variable were the same.
182 2007-04-16 John Darrington <john@darrington.wattle.id.au>
184 * t-test.q: Changed the output width of reported counts and
185 degrees of freedom, to avoid truncating these values. Thanks
186 to Seth Woolley for reporting this problem. A proper fix involves
187 re-thinking the output driver.
189 2007-04-12 Jason Stover <jhs@math.gcsu.edu>
191 * regression.q (run_regression): Added if (n_data >0) to fix bug
194 2007-03-29 Jason Stover <jhs@math.gcsu.edu>
196 * regression.q (prepare_data): New function.
198 * regression.q (compute_moments): New function.
200 2007-03-18 Ben Pfaff <blp@gnu.org>
202 * crosstabs.q (static var write): Rename write_style to avoid
203 conflict with POSIX function of same name.
205 2007-03-16 Jason Stover <jhs@math.gcsu.edu>
207 * regression.q (run_regression): Added support for moments.
209 Sat Feb 17 08:16:00 2007 Ben Pfaff <blp@gnu.org>
211 * flip.c (flip_sink_create): Improve error message when temporary
212 file cannot be created.
214 Tue Feb 6 19:58:03 2007 Ben Pfaff <blp@gnu.org>
216 * flip.c (flip_file): Give better error message on end-of-file.
218 2007-02-04 Jason Stover <jhs@math.gcsu.edu>
220 * regression.q: Fixed p-value computation in the test for
221 individual regression coefficients.
223 Mon Jan 15 11:03:20 2007 Ben Pfaff <blp@gnu.org>
225 Fix bugs found by valgrind when --enable-debug is used with the
226 new case code. These bugs are hidden when the data set is small
227 enough to find in memory; if a bigger data set that would overflow
228 to disk were used, then data corruption would occur.
230 * chisquare.c (create_freq_hash): Pass free_freq_mutable_hash to
231 hsh_create as free function. Make copy of data put into hash.
233 * oneway.q (free_value): New function.
234 (run_oneway): Use free_value as arg to hsh_create. Make copy of
237 * rank.q (rank_cases): Don't access data in case after we've given
240 Tue Jan 9 19:16:11 2007 Ben Pfaff <blp@gnu.org>
244 * aggregate.c (parse_aggregate_functions) Initialize
247 Fri Dec 22 14:04:09 2006 Ben Pfaff <blp@gnu.org>
249 Simplify missing value handling.
251 * aggregate.c (struct agr_var): Remove `bool include_missing', add
252 `enum mv_class exclude'. Remove `int missing', add `bool
253 saw_missing'. Update users.
255 * descriptives.c (struct dsc_trns): Removed `int
256 include_user_missing', add `enum mv_class exclude'. Update users.
257 (struct dsc_proc): Ditto.
259 * examine.q: (static var value_is_missing): Rename
260 `exclude_values', change type to `enum mv_class'. Update users.
264 Fri Dec 22 19:22:18 WST 2006 John Darrington <john@darrington.wattle.id.au>
266 * frequencies.q : Fixed bug #17420, where the table bounds were overun
267 when /FORMAT=nolabels was given.
269 Wed Dec 20 18:45:31 WST 2006 John Darrington <john@darrington.wattle.id.au>
271 * binomial.c binomial.h : New files. Thanks to Jason Stover
272 for assistance with these.
274 * chisquare.c chisquare.h freq.c freq.h npar-summary.c npar-summary.h
275 npar.h npar.q: New files. Implementing NPAR TESTS.
277 * frequencies.q : Moved structure definitions into freq.[ch]
279 Sat Dec 16 22:26:44 2006 Ben Pfaff <blp@gnu.org>
281 Make it possible to pull cases from the active file with a
282 function call, instead of requiring indirection through a callback
285 * aggregate.c (cmd_aggregate): Take advantage of new procedure
287 (agr_to_active_file): Removed.
288 (presorted_agr_to_sysfile): Removed.
290 * autorecode.c (cmd_autorecode): Take advantage of new procedure
292 (autorecode_proc_func): Removed.
294 * flip.c (struct flip_pgm): New members to allow conformance with
295 new case_source_class interface.
296 (cmd_flip): Adapt to new case_source_class interface.
297 (flip_source_read): Ditto.
298 (flip_source_destroy): Ditto.
300 Sat Dec 16 12:54:27 2006 Ben Pfaff <blp@gnu.org>
302 * rank.q (rank_custom_variables): Allow grouping variables to be
303 strings. Fixes bug #18533. Thanks to John Darrington for review.
305 Sat Dec 9 18:47:51 2006 Ben Pfaff <blp@gnu.org>
307 * regression.q (is_depvar): Compare variable pointers instead of
310 Thu Dec 7 15:26:25 WST 2006 John Darrington <john@darrington.wattle.id.au>
312 * examine.q: Allocated the categorical values for the dependent and
313 independent variables, on the heap. Hence they can be of any width.
315 Wed Dec 6 21:14:26 2006 Ben Pfaff <blp@gnu.org>
317 * regression.q (reg_inserted): Compare variable pointers instead
320 Mon Dec 4 22:33:46 2006 Ben Pfaff <blp@gnu.org>
322 * crosstabs.q (insert_summary): Use var_to_string for labeling.
323 (output_pivot_table) Ditto.
326 * frequencies.q (setup_z_trns): Ditto.
328 (dump_condensed) Ditto.
329 (dump_statistics) Ditto.
332 Sun Nov 5 08:31:42 WST 2006 John Darrington <john@darrington.wattle.id.au>
334 * t-test.q, oneway.q: Changed to use the new casefilter structure.
336 Sat Oct 14 16:52:28 2006 Ben Pfaff <blp@gnu.org>
338 * rank.q: (rank_sorted_casefile) Add some missing case_destroy()
339 calls to fix a memory leak.
341 Sun Oct 8 09:45:40 WST 2006 John Darrington <john@darrington.wattle.id.au>
343 * rank.q: Plugged a small memory leak which occurred under error
346 Sat Oct 7 11:06:01 WST 2006 John Darrington <john@darrington.wattle.id.au>
348 * rank.q: Implemented most of the RANK command.
350 2006-07-14 Jason Stover <jhs@math.gcsu.edu>
352 * regression.q (run_regression): New function to move knowledge of
353 pspp_linreg_cache out of math/coefficient.[ch].
355 Sat Jul 1 17:41:46 2006 Ben Pfaff <blp@gnu.org>
357 Fix bug #11612, "q2c documentation does not agree with code".
359 * examine.q: Audit use of q2c "+" prefixes that indicate that a
360 command may appear multiple times.
362 * frequencies.q: Ditto.
366 * regression.q: Ditto.
370 Fri Jun 23 14:18:22 2006 Ben Pfaff <blp@gnu.org>
372 Support long string variables on FREQUENCIES, as
373 an extension when in enhanced algorithms mode. For Greg Hunt
374 <greg@firmansyah.com>.
376 * frequencies.q: (struct freq) Change `v' member from union value
377 to union value *. Update all references.
378 (struct var_freqs) Add width, print members to represent effective
379 variable width and display format.
380 (calc) Copy entire long string value into the hash table.
381 (frq_custom_variables) Set new width, print members.
382 (hash_value_alpha) Get width from var_freqs.
383 (compare_value_alpha_a) Ditto.
384 (compare_freq_alpha_a) Ditto.
385 (compare_freq_alpha_d) Ditto.
386 (dump_full) Get display format from var_freqs.
387 (dump_condensed) Ditto.
389 Mon Jun 19 22:07:13 2006 Ben Pfaff <blp@gnu.org>
391 * frequencies.q: (dump_full) Only put the first MAX_SHORT_STRING
392 bytes of string variables into the output cells, seeing as we only
394 (dump_condensed) Ditto.
396 Mon Jun 19 21:52:05 2006 Ben Pfaff <blp@gnu.org>
398 Fixes a bug reported by Greg Hunt <greg@firmansyah.com>.
400 * frequencies.q: (hsh_hash_bytes) We only copy the first
401 MAX_SHORT_STRING bytes of string variables, so we must only
402 compare that many bytes, even if the string variable is longer.
403 (compare_value_alpha_a) Ditto.
404 (compare_freq_alpha_a) Ditto.
405 (compare_freq_alpha_d) Ditto.
407 2006-05-11 Jason Stover <jhs@math.gcsu.edu>
409 * regression.q: Adjusted code to account for cache->coeff being a
410 pspp_linreg_coeff **.
412 Sun May 7 18:31:25 2006 Ben Pfaff <blp@gnu.org>
416 * aggregate.c (cmd_aggregate): Free default_dict before replacing
419 Sun May 7 17:09:19 2006 Ben Pfaff <blp@gnu.org>
421 * flip.c (flip_file): Check return value of pool_fclose().
423 Sat May 6 16:00:13 2006 Ben Pfaff <blp@gnu.org>
425 Get rid of `char *c' member in union value, for cleanliness.
427 * aggregate.c: (union agr_argument) New union.
428 (struct agr_var) Change element type of arg[] from union value to
430 (parse_aggregate_functions) Change local variable types likewise.
432 * autorecode.c: (union arc_value) New union.
433 (struct arc_item) Change "from" from union value to union
435 (recode) Change local variable from union value to union
437 (autorecode_trns_proc) Ditto.
438 (compare_alpha_value) Ditto.
439 (hash_alpha_value) Ditto.
440 (compare_numeric_value) Ditto.
441 (hash_numeric_value) Ditto.
442 (autorecode_proc_func) Ditto.
444 Sat May 6 10:43:33 2006 Ben Pfaff <blp@gnu.org>
446 Continue reforming procedure execution. In this phase, get rid of
447 the output code for SPLIT FILE groups in procedure.c, which really
448 shouldn't be doing any output. Move it into the individual
449 procedures instead. This also adds some flexibility.
451 * crosstabs.q (precalc): Call output_split_file_values().
453 * descriptives.c (calc_descriptives): Ditto.
455 * examine.q (run_examine): Ditto.
457 * frequencies.q (precalc): Ditto.
459 * oneway.q (run_oneway): Ditto.
461 * regression.q (run_regression): Ditto.
463 * t-test.q (calculate): Ditto.
465 Wed May 3 23:05:31 2006 Ben Pfaff <blp@gnu.org>
467 Continue reforming procedure execution. In this phase, get rid of
468 many global variables, consolidating procedure execution in
469 procedure.c. Encapsulate transformations in new "struct
470 trns_chain". Also, change implementation of N OF CASES, FILTER,
471 and PROCESS IF from special cases to transformations.
473 * aggregate.c (cmd_aggregate) Use discard_variables().
475 2006-04-28 Jason Stover <jhs@math.gcsu.edu>
477 * regression.q (regression_trns_resid_proc): Pass only the
478 variables used in the model to (*model->residual)().
480 * regression.q (regression_trns_pred_proc): Pass only the
481 variables used in the model to (*model->pred)().
483 2006-04-26 Jason Stover <jhs@math.gcsu.edu>
485 * regression.q: Added support for multiple transformations.
487 * regression.q (regression_trns_resid_proc): New function.
489 * regression.q (regression_trns_pred_proc): New function.
491 * regression.q (subcommand_save): Added support for saving
494 * regression.q (regression_trns_free): New function.
496 * regression.q (reg_get_name): New function.
498 * regression.q (reg_save_var): New function.
500 Tue Apr 25 13:18:56 2006 Ben Pfaff <blp@gnu.org>
502 * rank.q (parse_rank_function): Use SE instead of ME for parse
505 Tue Apr 25 13:16:28 2006 Ben Pfaff <blp@gnu.org>
507 * flip.c (flip_sink_write): Use snprintf() to simplify a bit of
510 2006-04-21 Jason Stover <jhs@math.gcsu.edu>
512 * regression.q (try_name): New function. (Partly copied from
513 try_name in descriptives.c.)
515 * regression.q (subcommand_save): Choose residual variable names
518 2006-04-20 Jason Stover <jhs@math.gcsu.edu>
520 * regression.q (cmd_regression): Moved call to subcommand_save()
521 outside multipass_procedure_with_splits().
523 * regression.q (regression_trns_proc): Fixed value counter n_vals
524 before calling *model->residual().
526 2006-04-19 Jason Stover <jhs@math.gcsu.edu>
528 * regression.q (regression_trns_proc): Fixed the look-up of the
531 2006-04-18 Jason Stover <jhs@math.gcsu.edu>
533 * regression.q (regression_trns_proc): Look up the residual
534 variable in the linear regression cache.
536 * regression.q (subcommand_save): Set the residual variable in the
537 linear regression cache.
539 2006-04-17 Jason Stover <jason@wonko.gcsu.edu>
541 * regression.q (regression_trns_proc): Accept case_idx as an int
542 to match the definition of trns_proc_func.
544 2006-04-17 Jason Stover <jhs@math.gcsu.edu>
546 * regression.q (regression_trns_proc): New function.
548 * regression.q (subcommand_save): Create variable residuals and
549 add a transformation to assign values to them. Also free the
550 linreg_cache if the SAVE command was not called. Removed the
551 casereading loop. Placed actual computation of residuals in
552 regression_trns_proc.
554 * regression.q (run_regression): Moved call to free
555 pspp_linreg_cache to subcommand_save.
557 Sat Apr 15 18:01:03 2006 Ben Pfaff <blp@gnu.org>
559 * examine.q (output_examine): Add casts to fix warnings.
561 2006-04-07 Jason Stover <jhs@math.gcsu.edu>
563 * regression.q (subcommand_save): New function.
565 2006-04-04 Jason Stover <jhs@math.gcsu.edu>
567 * regression.q: New function reg_has_categorical () to tell
568 whether a pspp_linreg_struct was made with any variables of type
571 * regression.q: (subcommand_export): Call
572 reg_print_categorical_encoding() only if the model uses any
573 categorical variables.
575 Mon Mar 27 16:00:42 2006 Ben Pfaff <blp@gnu.org>
577 * crosstabs.q: (output_pivot_table) Drop spurious space from
580 2006-03-15 Jason Stover <jhs@math.gcsu.edu>
582 * regression.q: Added custom syntax parser for VARIABLES subcommand
584 * regression.q: Moved most instructions for run_regression ()
585 inside the loop over dependent variables.
587 Thu Mar 2 08:40:33 WST 2006 John Darrington <john@darrington.wattle.id.au>
589 * Moved files from src directory