1 2008-04-08 Jason Stover <jhs@math.gcsu.edu>
3 * regression.q (identify_indep_vars): Don't panic unless
6 2008-03-16 Ben Pfaff <blp@gnu.org>
8 Bug #22037. Thanks to John Darrington for reporting this bug.
10 * crosstabs.q (calc_general): Only the short string prefix of long
11 string variables are tabulated, so we must not copy or zero out
14 2008-03-10 Jason Stover <jhs@debs.hoobahooba.net>
16 * regression.q (run_regression): Removed code for EXPORT
17 subcommand. Remove use of coefficient 0 as the intercept.
19 2008-02-14 John Darrington <john@darrington.wattle.id.au>
21 * examine.q: Fixed counts of missing variables. Thanks to
22 Jason Stover for reporting this problem.
24 2008-01-02 John Darrington <john@darrington.wattle.id.au>
26 * binomial.c chisquare.c examine.q frequencies.q oneway.q regression.q : updated
27 all users of var_get_value_name to use replacement function var_append_value_name.
29 2007-12-07 Ben Pfaff <blp@gnu.org>
33 * crosstabs.q (precalc): Initialize data structures even if the
34 first case cannot be read.
36 * frequencies.q (precalc): Ditto.
38 2007-11-03 Ben Pfaff <blp@gnu.org>
40 Allow output files to overwrite input files (bug #21280).
42 * aggregate.c (cmd_aggregate): Manage file handle reference
45 * correlations.q (internal_cmd_frequencies): Ditto.
46 (cor_custom_matrix): Ditto.
48 * regression.q (regression_custom_export): Ditto.
49 (cmd_regression): Ditto.
51 2007-10-12 Ben Pfaff <blp@gnu.org>
53 * flip.c (flip_file): No need to conditionally substitute for
54 "fseeko" and "off_t" manually anymore, as gnulib takes care of it
57 2007-09-21 Jason Stover <jhs@wonko.gcsu.edu>
59 * regression.q (run_regression): Partial fix of memory leak, bug
62 2007-09-19 Ben Pfaff <blp@gnu.org>
66 * aggregate.c (cmd_aggregate): Destroy casereader consistently,
67 even if casereader fails.
69 * examine.q (run_examine): Ditto.
71 * glm.q (run_glm): Ditto.
73 * oneway.q (run_oneway): Ditto.
75 * regression.q (run_regression): Ditto.
77 * t-test.q (calculate): Ditto.
79 * descriptives.c (calc_descriptives): Ditto. Also avoid
80 gratuitous casereader_clone.
82 2007-09-13 Jason Stover <jhs@debs.hoobahooba.net>
84 * regression.q (cmd_regression): Move declaration of models in to
85 definition of cmd_regression.
87 * regression.q (run_regression): Free mom to fix memory leak.
89 2007-09-12 Ben Pfaff <blp@gnu.org>
91 * crosstabs.q (postcalc): Free sorted_tab and the structures that
92 it points to, to plug a memory leak. Fixes bug #20910. Thanks to
93 John Darrington for reporting this bug and for review.
95 2007-09-04 Ben Pfaff <blp@gnu.org>
97 * crosstabs.q (cmd_crosstabs): Free xtab and the structures that
98 it points to, to plug a memory leak. Fixes bug #18315.
100 2007-08-15 Jason Stover <jhs@wonko.gcsu.edu>
102 * regression.q (identify_indep_vars): Print an error if dependent
103 and independent variables are the same. Fixes bug 19819.
105 2007-08-12 Ben Pfaff <blp@gnu.org>
107 * flip.c: Drop use of dict_get_compacted_dict_index_to_case_index
108 and just use the ordinary case indexes. There seemed to be no
109 reason for the former method.
111 2007-08-03 Ben Pfaff <blp@gnu.org>
113 * rank.q (rank_cmd): Instead of sorting by SPLIT FILE vars, group
114 by them. Fixes bug #17239.
115 Reviewed by John Darrington.
117 2007-08-01 Ben Pfaff <blp@gnu.org>
119 Clean up handling of median, by treating it almost like any other
120 percentile. Fixes bug #17424. Thanks to John Darrington for
122 * frequencies.q (internal_cmd_frequencies): Fix handling of bit
123 masks for `stats' variable. If median is selected, turn it off
124 and add a 50th percentile.
125 (add_percentile): Simplify code a little.
126 (calc_stats): Drop special casing of median.
127 (dump_statistics): Ditto, except that we label the 50th percentile
128 as "50 (Median)" to make it clear that it's also the median.
130 2007-07-31 Ben Pfaff <blp@gnu.org>
132 Remove integer mode from FREQUENCIES and incidentally fix bug
133 #17421. Reviewed by John Darrington.
134 * frequencies.q (int_pool): Rename data_pool.
135 (gen_pool): Rename syntax_pool.
136 (enum FRQM_*): Removed.
137 (struct freq_tab): Removed `mode', `vector', `min', `max',
138 `out_of_range', `sysmis' members.
139 (calc): Delete support for integer mode.
141 (postprocess_freq_tab): Ditto.
142 (cleanup_freq_tab): Ditto.
143 (frq_custom_variables): Ditto.
145 2007-07-28 John Darrington <john@darrington.wattle.id.au>
147 * t-test.q: Moved the order in which groups are displayed in the
148 independent samples case, where a cut point is given.
150 2007-07-27 Ben Pfaff <blp@gnu.org>
152 * regression.q (run_regression): Move casereader_destroy call so
153 that it always gets called, not just if there was some valid
154 data. Fixes bug #19581.
155 Reviewed by Jason Stover.
157 2007-07-24 Ben Pfaff <blp@gnu.org>
159 * flip.c (struct flip_pgm): Remove `case_size' member (now
161 (cmd_flip): Pass var_cnt as number of cases instead of case_cnt,
162 to fix bug #20494. Don't assign to `case_size' member, which was
163 unused after assignment.
164 (build_dictionary): When NEWNAMES not used, get the number of
165 variables right, to fix bug #20493.
167 2007-07-10 Jason Stover <jhs@math.gcsu.edu>
169 * glm.q: Initial version of the GLM procedure.
171 2007-06-06 Ben Pfaff <blp@gnu.org>
173 Adapt case sources, sinks, and clients of procedure code to the
176 * aggregate.c: Simplify greatly since everything is more uniform
179 * autorecode.c: Adapt to new procedure code.
181 * chisquare.c: Ditto.
182 * crosstabs.q: Ditto.
183 * descriptives.c: Ditto.
185 * npar-summary.c: Ditto.
186 * frequencies.q: Ditto.
189 * regression.q: Ditto.
190 * sort-cases.c: Ditto.
193 * sort-criteria.c: Rewrite to output a struct case_ordering.
195 * flip.c: Rewrite to be a casereader.
197 * rank.q: Simplify greatly since casereaders are much more
198 flexible than what we had before.
200 2007-05-15 Jason Stover <jhs@math.gcsu.edu>
202 * regression.q (run_regression): Tell the user when the data
203 contain no valid cases.
205 2007-05-08 Jason Stover <jhs@math.gcsu.edu>
207 * regression.q: Partial fix of bug which caused a crash if
208 dependent variable and independent variable were the same.
210 2007-04-16 John Darrington <john@darrington.wattle.id.au>
212 * t-test.q: Changed the output width of reported counts and
213 degrees of freedom, to avoid truncating these values. Thanks
214 to Seth Woolley for reporting this problem. A proper fix involves
215 re-thinking the output driver.
217 2007-04-12 Jason Stover <jhs@math.gcsu.edu>
219 * regression.q (run_regression): Added if (n_data >0) to fix bug
222 2007-03-29 Jason Stover <jhs@math.gcsu.edu>
224 * regression.q (prepare_data): New function.
226 * regression.q (compute_moments): New function.
228 2007-03-18 Ben Pfaff <blp@gnu.org>
230 * crosstabs.q (static var write): Rename write_style to avoid
231 conflict with POSIX function of same name.
233 2007-03-16 Jason Stover <jhs@math.gcsu.edu>
235 * regression.q (run_regression): Added support for moments.
237 Sat Feb 17 08:16:00 2007 Ben Pfaff <blp@gnu.org>
239 * flip.c (flip_sink_create): Improve error message when temporary
240 file cannot be created.
242 Tue Feb 6 19:58:03 2007 Ben Pfaff <blp@gnu.org>
244 * flip.c (flip_file): Give better error message on end-of-file.
246 2007-02-04 Jason Stover <jhs@math.gcsu.edu>
248 * regression.q: Fixed p-value computation in the test for
249 individual regression coefficients.
251 Mon Jan 15 11:03:20 2007 Ben Pfaff <blp@gnu.org>
253 Fix bugs found by valgrind when --enable-debug is used with the
254 new case code. These bugs are hidden when the data set is small
255 enough to find in memory; if a bigger data set that would overflow
256 to disk were used, then data corruption would occur.
258 * chisquare.c (create_freq_hash): Pass free_freq_mutable_hash to
259 hsh_create as free function. Make copy of data put into hash.
261 * oneway.q (free_value): New function.
262 (run_oneway): Use free_value as arg to hsh_create. Make copy of
265 * rank.q (rank_cases): Don't access data in case after we've given
268 Tue Jan 9 19:16:11 2007 Ben Pfaff <blp@gnu.org>
272 * aggregate.c (parse_aggregate_functions) Initialize
275 Fri Dec 22 14:04:09 2006 Ben Pfaff <blp@gnu.org>
277 Simplify missing value handling.
279 * aggregate.c (struct agr_var): Remove `bool include_missing', add
280 `enum mv_class exclude'. Remove `int missing', add `bool
281 saw_missing'. Update users.
283 * descriptives.c (struct dsc_trns): Removed `int
284 include_user_missing', add `enum mv_class exclude'. Update users.
285 (struct dsc_proc): Ditto.
287 * examine.q: (static var value_is_missing): Rename
288 `exclude_values', change type to `enum mv_class'. Update users.
292 Fri Dec 22 19:22:18 WST 2006 John Darrington <john@darrington.wattle.id.au>
294 * frequencies.q : Fixed bug #17420, where the table bounds were overun
295 when /FORMAT=nolabels was given.
297 Wed Dec 20 18:45:31 WST 2006 John Darrington <john@darrington.wattle.id.au>
299 * binomial.c binomial.h : New files. Thanks to Jason Stover
300 for assistance with these.
302 * chisquare.c chisquare.h freq.c freq.h npar-summary.c npar-summary.h
303 npar.h npar.q: New files. Implementing NPAR TESTS.
305 * frequencies.q : Moved structure definitions into freq.[ch]
307 Sat Dec 16 22:26:44 2006 Ben Pfaff <blp@gnu.org>
309 Make it possible to pull cases from the active file with a
310 function call, instead of requiring indirection through a callback
313 * aggregate.c (cmd_aggregate): Take advantage of new procedure
315 (agr_to_active_file): Removed.
316 (presorted_agr_to_sysfile): Removed.
318 * autorecode.c (cmd_autorecode): Take advantage of new procedure
320 (autorecode_proc_func): Removed.
322 * flip.c (struct flip_pgm): New members to allow conformance with
323 new case_source_class interface.
324 (cmd_flip): Adapt to new case_source_class interface.
325 (flip_source_read): Ditto.
326 (flip_source_destroy): Ditto.
328 Sat Dec 16 12:54:27 2006 Ben Pfaff <blp@gnu.org>
330 * rank.q (rank_custom_variables): Allow grouping variables to be
331 strings. Fixes bug #18533. Thanks to John Darrington for review.
333 Sat Dec 9 18:47:51 2006 Ben Pfaff <blp@gnu.org>
335 * regression.q (is_depvar): Compare variable pointers instead of
338 Thu Dec 7 15:26:25 WST 2006 John Darrington <john@darrington.wattle.id.au>
340 * examine.q: Allocated the categorical values for the dependent and
341 independent variables, on the heap. Hence they can be of any width.
343 Wed Dec 6 21:14:26 2006 Ben Pfaff <blp@gnu.org>
345 * regression.q (reg_inserted): Compare variable pointers instead
348 Mon Dec 4 22:33:46 2006 Ben Pfaff <blp@gnu.org>
350 * crosstabs.q (insert_summary): Use var_to_string for labeling.
351 (output_pivot_table) Ditto.
354 * frequencies.q (setup_z_trns): Ditto.
356 (dump_condensed) Ditto.
357 (dump_statistics) Ditto.
360 Sun Nov 5 08:31:42 WST 2006 John Darrington <john@darrington.wattle.id.au>
362 * t-test.q, oneway.q: Changed to use the new casefilter structure.
364 Sat Oct 14 16:52:28 2006 Ben Pfaff <blp@gnu.org>
366 * rank.q: (rank_sorted_casefile) Add some missing case_destroy()
367 calls to fix a memory leak.
369 Sun Oct 8 09:45:40 WST 2006 John Darrington <john@darrington.wattle.id.au>
371 * rank.q: Plugged a small memory leak which occurred under error
374 Sat Oct 7 11:06:01 WST 2006 John Darrington <john@darrington.wattle.id.au>
376 * rank.q: Implemented most of the RANK command.
378 2006-07-14 Jason Stover <jhs@math.gcsu.edu>
380 * regression.q (run_regression): New function to move knowledge of
381 pspp_linreg_cache out of math/coefficient.[ch].
383 Sat Jul 1 17:41:46 2006 Ben Pfaff <blp@gnu.org>
385 Fix bug #11612, "q2c documentation does not agree with code".
387 * examine.q: Audit use of q2c "+" prefixes that indicate that a
388 command may appear multiple times.
390 * frequencies.q: Ditto.
394 * regression.q: Ditto.
398 Fri Jun 23 14:18:22 2006 Ben Pfaff <blp@gnu.org>
400 Support long string variables on FREQUENCIES, as
401 an extension when in enhanced algorithms mode. For Greg Hunt
402 <greg@firmansyah.com>.
404 * frequencies.q: (struct freq) Change `v' member from union value
405 to union value *. Update all references.
406 (struct var_freqs) Add width, print members to represent effective
407 variable width and display format.
408 (calc) Copy entire long string value into the hash table.
409 (frq_custom_variables) Set new width, print members.
410 (hash_value_alpha) Get width from var_freqs.
411 (compare_value_alpha_a) Ditto.
412 (compare_freq_alpha_a) Ditto.
413 (compare_freq_alpha_d) Ditto.
414 (dump_full) Get display format from var_freqs.
415 (dump_condensed) Ditto.
417 Mon Jun 19 22:07:13 2006 Ben Pfaff <blp@gnu.org>
419 * frequencies.q: (dump_full) Only put the first MAX_SHORT_STRING
420 bytes of string variables into the output cells, seeing as we only
422 (dump_condensed) Ditto.
424 Mon Jun 19 21:52:05 2006 Ben Pfaff <blp@gnu.org>
426 Fixes a bug reported by Greg Hunt <greg@firmansyah.com>.
428 * frequencies.q: (hsh_hash_bytes) We only copy the first
429 MAX_SHORT_STRING bytes of string variables, so we must only
430 compare that many bytes, even if the string variable is longer.
431 (compare_value_alpha_a) Ditto.
432 (compare_freq_alpha_a) Ditto.
433 (compare_freq_alpha_d) Ditto.
435 2006-05-11 Jason Stover <jhs@math.gcsu.edu>
437 * regression.q: Adjusted code to account for cache->coeff being a
438 pspp_linreg_coeff **.
440 Sun May 7 18:31:25 2006 Ben Pfaff <blp@gnu.org>
444 * aggregate.c (cmd_aggregate): Free default_dict before replacing
447 Sun May 7 17:09:19 2006 Ben Pfaff <blp@gnu.org>
449 * flip.c (flip_file): Check return value of pool_fclose().
451 Sat May 6 16:00:13 2006 Ben Pfaff <blp@gnu.org>
453 Get rid of `char *c' member in union value, for cleanliness.
455 * aggregate.c: (union agr_argument) New union.
456 (struct agr_var) Change element type of arg[] from union value to
458 (parse_aggregate_functions) Change local variable types likewise.
460 * autorecode.c: (union arc_value) New union.
461 (struct arc_item) Change "from" from union value to union
463 (recode) Change local variable from union value to union
465 (autorecode_trns_proc) Ditto.
466 (compare_alpha_value) Ditto.
467 (hash_alpha_value) Ditto.
468 (compare_numeric_value) Ditto.
469 (hash_numeric_value) Ditto.
470 (autorecode_proc_func) Ditto.
472 Sat May 6 10:43:33 2006 Ben Pfaff <blp@gnu.org>
474 Continue reforming procedure execution. In this phase, get rid of
475 the output code for SPLIT FILE groups in procedure.c, which really
476 shouldn't be doing any output. Move it into the individual
477 procedures instead. This also adds some flexibility.
479 * crosstabs.q (precalc): Call output_split_file_values().
481 * descriptives.c (calc_descriptives): Ditto.
483 * examine.q (run_examine): Ditto.
485 * frequencies.q (precalc): Ditto.
487 * oneway.q (run_oneway): Ditto.
489 * regression.q (run_regression): Ditto.
491 * t-test.q (calculate): Ditto.
493 Wed May 3 23:05:31 2006 Ben Pfaff <blp@gnu.org>
495 Continue reforming procedure execution. In this phase, get rid of
496 many global variables, consolidating procedure execution in
497 procedure.c. Encapsulate transformations in new "struct
498 trns_chain". Also, change implementation of N OF CASES, FILTER,
499 and PROCESS IF from special cases to transformations.
501 * aggregate.c (cmd_aggregate) Use discard_variables().
503 2006-04-28 Jason Stover <jhs@math.gcsu.edu>
505 * regression.q (regression_trns_resid_proc): Pass only the
506 variables used in the model to (*model->residual)().
508 * regression.q (regression_trns_pred_proc): Pass only the
509 variables used in the model to (*model->pred)().
511 2006-04-26 Jason Stover <jhs@math.gcsu.edu>
513 * regression.q: Added support for multiple transformations.
515 * regression.q (regression_trns_resid_proc): New function.
517 * regression.q (regression_trns_pred_proc): New function.
519 * regression.q (subcommand_save): Added support for saving
522 * regression.q (regression_trns_free): New function.
524 * regression.q (reg_get_name): New function.
526 * regression.q (reg_save_var): New function.
528 Tue Apr 25 13:18:56 2006 Ben Pfaff <blp@gnu.org>
530 * rank.q (parse_rank_function): Use SE instead of ME for parse
533 Tue Apr 25 13:16:28 2006 Ben Pfaff <blp@gnu.org>
535 * flip.c (flip_sink_write): Use snprintf() to simplify a bit of
538 2006-04-21 Jason Stover <jhs@math.gcsu.edu>
540 * regression.q (try_name): New function. (Partly copied from
541 try_name in descriptives.c.)
543 * regression.q (subcommand_save): Choose residual variable names
546 2006-04-20 Jason Stover <jhs@math.gcsu.edu>
548 * regression.q (cmd_regression): Moved call to subcommand_save()
549 outside multipass_procedure_with_splits().
551 * regression.q (regression_trns_proc): Fixed value counter n_vals
552 before calling *model->residual().
554 2006-04-19 Jason Stover <jhs@math.gcsu.edu>
556 * regression.q (regression_trns_proc): Fixed the look-up of the
559 2006-04-18 Jason Stover <jhs@math.gcsu.edu>
561 * regression.q (regression_trns_proc): Look up the residual
562 variable in the linear regression cache.
564 * regression.q (subcommand_save): Set the residual variable in the
565 linear regression cache.
567 2006-04-17 Jason Stover <jason@wonko.gcsu.edu>
569 * regression.q (regression_trns_proc): Accept case_idx as an int
570 to match the definition of trns_proc_func.
572 2006-04-17 Jason Stover <jhs@math.gcsu.edu>
574 * regression.q (regression_trns_proc): New function.
576 * regression.q (subcommand_save): Create variable residuals and
577 add a transformation to assign values to them. Also free the
578 linreg_cache if the SAVE command was not called. Removed the
579 casereading loop. Placed actual computation of residuals in
580 regression_trns_proc.
582 * regression.q (run_regression): Moved call to free
583 pspp_linreg_cache to subcommand_save.
585 Sat Apr 15 18:01:03 2006 Ben Pfaff <blp@gnu.org>
587 * examine.q (output_examine): Add casts to fix warnings.
589 2006-04-07 Jason Stover <jhs@math.gcsu.edu>
591 * regression.q (subcommand_save): New function.
593 2006-04-04 Jason Stover <jhs@math.gcsu.edu>
595 * regression.q: New function reg_has_categorical () to tell
596 whether a pspp_linreg_struct was made with any variables of type
599 * regression.q: (subcommand_export): Call
600 reg_print_categorical_encoding() only if the model uses any
601 categorical variables.
603 Mon Mar 27 16:00:42 2006 Ben Pfaff <blp@gnu.org>
605 * crosstabs.q: (output_pivot_table) Drop spurious space from
608 2006-03-15 Jason Stover <jhs@math.gcsu.edu>
610 * regression.q: Added custom syntax parser for VARIABLES subcommand
612 * regression.q: Moved most instructions for run_regression ()
613 inside the loop over dependent variables.
615 Thu Mar 2 08:40:33 WST 2006 John Darrington <john@darrington.wattle.id.au>
617 * Moved files from src directory