1 2008-03-16 Ben Pfaff <blp@gnu.org>
3 Bug #22037. Thanks to John Darrington for reporting this bug.
5 * crosstabs.q (calc_general): Only the short string prefix of long
6 string variables are tabulated, so we must not copy or zero out
9 2008-03-10 Jason Stover <jhs@debs.hoobahooba.net>
11 * regression.q (run_regression): Removed code for EXPORT
12 subcommand. Remove use of coefficient 0 as the intercept.
14 2008-02-14 John Darrington <john@darrington.wattle.id.au>
16 * examine.q: Fixed counts of missing variables. Thanks to
17 Jason Stover for reporting this problem.
19 2008-01-02 John Darrington <john@darrington.wattle.id.au>
21 * binomial.c chisquare.c examine.q frequencies.q oneway.q regression.q : updated
22 all users of var_get_value_name to use replacement function var_append_value_name.
24 2007-12-07 Ben Pfaff <blp@gnu.org>
28 * crosstabs.q (precalc): Initialize data structures even if the
29 first case cannot be read.
31 * frequencies.q (precalc): Ditto.
33 2007-11-03 Ben Pfaff <blp@gnu.org>
35 Allow output files to overwrite input files (bug #21280).
37 * aggregate.c (cmd_aggregate): Manage file handle reference
40 * correlations.q (internal_cmd_frequencies): Ditto.
41 (cor_custom_matrix): Ditto.
43 * regression.q (regression_custom_export): Ditto.
44 (cmd_regression): Ditto.
46 2007-10-12 Ben Pfaff <blp@gnu.org>
48 * flip.c (flip_file): No need to conditionally substitute for
49 "fseeko" and "off_t" manually anymore, as gnulib takes care of it
52 2007-09-21 Jason Stover <jhs@wonko.gcsu.edu>
54 * regression.q (run_regression): Partial fix of memory leak, bug
57 2007-09-19 Ben Pfaff <blp@gnu.org>
61 * aggregate.c (cmd_aggregate): Destroy casereader consistently,
62 even if casereader fails.
64 * examine.q (run_examine): Ditto.
66 * glm.q (run_glm): Ditto.
68 * oneway.q (run_oneway): Ditto.
70 * regression.q (run_regression): Ditto.
72 * t-test.q (calculate): Ditto.
74 * descriptives.c (calc_descriptives): Ditto. Also avoid
75 gratuitous casereader_clone.
77 2007-09-13 Jason Stover <jhs@debs.hoobahooba.net>
79 * regression.q (cmd_regression): Move declaration of models in to
80 definition of cmd_regression.
82 * regression.q (run_regression): Free mom to fix memory leak.
84 2007-09-12 Ben Pfaff <blp@gnu.org>
86 * crosstabs.q (postcalc): Free sorted_tab and the structures that
87 it points to, to plug a memory leak. Fixes bug #20910. Thanks to
88 John Darrington for reporting this bug and for review.
90 2007-09-04 Ben Pfaff <blp@gnu.org>
92 * crosstabs.q (cmd_crosstabs): Free xtab and the structures that
93 it points to, to plug a memory leak. Fixes bug #18315.
95 2007-08-15 Jason Stover <jhs@wonko.gcsu.edu>
97 * regression.q (identify_indep_vars): Print an error if dependent
98 and independent variables are the same. Fixes bug 19819.
100 2007-08-12 Ben Pfaff <blp@gnu.org>
102 * flip.c: Drop use of dict_get_compacted_dict_index_to_case_index
103 and just use the ordinary case indexes. There seemed to be no
104 reason for the former method.
106 2007-08-03 Ben Pfaff <blp@gnu.org>
108 * rank.q (rank_cmd): Instead of sorting by SPLIT FILE vars, group
109 by them. Fixes bug #17239.
110 Reviewed by John Darrington.
112 2007-08-01 Ben Pfaff <blp@gnu.org>
114 Clean up handling of median, by treating it almost like any other
115 percentile. Fixes bug #17424. Thanks to John Darrington for
117 * frequencies.q (internal_cmd_frequencies): Fix handling of bit
118 masks for `stats' variable. If median is selected, turn it off
119 and add a 50th percentile.
120 (add_percentile): Simplify code a little.
121 (calc_stats): Drop special casing of median.
122 (dump_statistics): Ditto, except that we label the 50th percentile
123 as "50 (Median)" to make it clear that it's also the median.
125 2007-07-31 Ben Pfaff <blp@gnu.org>
127 Remove integer mode from FREQUENCIES and incidentally fix bug
128 #17421. Reviewed by John Darrington.
129 * frequencies.q (int_pool): Rename data_pool.
130 (gen_pool): Rename syntax_pool.
131 (enum FRQM_*): Removed.
132 (struct freq_tab): Removed `mode', `vector', `min', `max',
133 `out_of_range', `sysmis' members.
134 (calc): Delete support for integer mode.
136 (postprocess_freq_tab): Ditto.
137 (cleanup_freq_tab): Ditto.
138 (frq_custom_variables): Ditto.
140 2007-07-28 John Darrington <john@darrington.wattle.id.au>
142 * t-test.q: Moved the order in which groups are displayed in the
143 independent samples case, where a cut point is given.
145 2007-07-27 Ben Pfaff <blp@gnu.org>
147 * regression.q (run_regression): Move casereader_destroy call so
148 that it always gets called, not just if there was some valid
149 data. Fixes bug #19581.
150 Reviewed by Jason Stover.
152 2007-07-24 Ben Pfaff <blp@gnu.org>
154 * flip.c (struct flip_pgm): Remove `case_size' member (now
156 (cmd_flip): Pass var_cnt as number of cases instead of case_cnt,
157 to fix bug #20494. Don't assign to `case_size' member, which was
158 unused after assignment.
159 (build_dictionary): When NEWNAMES not used, get the number of
160 variables right, to fix bug #20493.
162 2007-07-10 Jason Stover <jhs@math.gcsu.edu>
164 * glm.q: Initial version of the GLM procedure.
166 2007-06-06 Ben Pfaff <blp@gnu.org>
168 Adapt case sources, sinks, and clients of procedure code to the
171 * aggregate.c: Simplify greatly since everything is more uniform
174 * autorecode.c: Adapt to new procedure code.
176 * chisquare.c: Ditto.
177 * crosstabs.q: Ditto.
178 * descriptives.c: Ditto.
180 * npar-summary.c: Ditto.
181 * frequencies.q: Ditto.
184 * regression.q: Ditto.
185 * sort-cases.c: Ditto.
188 * sort-criteria.c: Rewrite to output a struct case_ordering.
190 * flip.c: Rewrite to be a casereader.
192 * rank.q: Simplify greatly since casereaders are much more
193 flexible than what we had before.
195 2007-05-15 Jason Stover <jhs@math.gcsu.edu>
197 * regression.q (run_regression): Tell the user when the data
198 contain no valid cases.
200 2007-05-08 Jason Stover <jhs@math.gcsu.edu>
202 * regression.q: Partial fix of bug which caused a crash if
203 dependent variable and independent variable were the same.
205 2007-04-16 John Darrington <john@darrington.wattle.id.au>
207 * t-test.q: Changed the output width of reported counts and
208 degrees of freedom, to avoid truncating these values. Thanks
209 to Seth Woolley for reporting this problem. A proper fix involves
210 re-thinking the output driver.
212 2007-04-12 Jason Stover <jhs@math.gcsu.edu>
214 * regression.q (run_regression): Added if (n_data >0) to fix bug
217 2007-03-29 Jason Stover <jhs@math.gcsu.edu>
219 * regression.q (prepare_data): New function.
221 * regression.q (compute_moments): New function.
223 2007-03-18 Ben Pfaff <blp@gnu.org>
225 * crosstabs.q (static var write): Rename write_style to avoid
226 conflict with POSIX function of same name.
228 2007-03-16 Jason Stover <jhs@math.gcsu.edu>
230 * regression.q (run_regression): Added support for moments.
232 Sat Feb 17 08:16:00 2007 Ben Pfaff <blp@gnu.org>
234 * flip.c (flip_sink_create): Improve error message when temporary
235 file cannot be created.
237 Tue Feb 6 19:58:03 2007 Ben Pfaff <blp@gnu.org>
239 * flip.c (flip_file): Give better error message on end-of-file.
241 2007-02-04 Jason Stover <jhs@math.gcsu.edu>
243 * regression.q: Fixed p-value computation in the test for
244 individual regression coefficients.
246 Mon Jan 15 11:03:20 2007 Ben Pfaff <blp@gnu.org>
248 Fix bugs found by valgrind when --enable-debug is used with the
249 new case code. These bugs are hidden when the data set is small
250 enough to find in memory; if a bigger data set that would overflow
251 to disk were used, then data corruption would occur.
253 * chisquare.c (create_freq_hash): Pass free_freq_mutable_hash to
254 hsh_create as free function. Make copy of data put into hash.
256 * oneway.q (free_value): New function.
257 (run_oneway): Use free_value as arg to hsh_create. Make copy of
260 * rank.q (rank_cases): Don't access data in case after we've given
263 Tue Jan 9 19:16:11 2007 Ben Pfaff <blp@gnu.org>
267 * aggregate.c (parse_aggregate_functions) Initialize
270 Fri Dec 22 14:04:09 2006 Ben Pfaff <blp@gnu.org>
272 Simplify missing value handling.
274 * aggregate.c (struct agr_var): Remove `bool include_missing', add
275 `enum mv_class exclude'. Remove `int missing', add `bool
276 saw_missing'. Update users.
278 * descriptives.c (struct dsc_trns): Removed `int
279 include_user_missing', add `enum mv_class exclude'. Update users.
280 (struct dsc_proc): Ditto.
282 * examine.q: (static var value_is_missing): Rename
283 `exclude_values', change type to `enum mv_class'. Update users.
287 Fri Dec 22 19:22:18 WST 2006 John Darrington <john@darrington.wattle.id.au>
289 * frequencies.q : Fixed bug #17420, where the table bounds were overun
290 when /FORMAT=nolabels was given.
292 Wed Dec 20 18:45:31 WST 2006 John Darrington <john@darrington.wattle.id.au>
294 * binomial.c binomial.h : New files. Thanks to Jason Stover
295 for assistance with these.
297 * chisquare.c chisquare.h freq.c freq.h npar-summary.c npar-summary.h
298 npar.h npar.q: New files. Implementing NPAR TESTS.
300 * frequencies.q : Moved structure definitions into freq.[ch]
302 Sat Dec 16 22:26:44 2006 Ben Pfaff <blp@gnu.org>
304 Make it possible to pull cases from the active file with a
305 function call, instead of requiring indirection through a callback
308 * aggregate.c (cmd_aggregate): Take advantage of new procedure
310 (agr_to_active_file): Removed.
311 (presorted_agr_to_sysfile): Removed.
313 * autorecode.c (cmd_autorecode): Take advantage of new procedure
315 (autorecode_proc_func): Removed.
317 * flip.c (struct flip_pgm): New members to allow conformance with
318 new case_source_class interface.
319 (cmd_flip): Adapt to new case_source_class interface.
320 (flip_source_read): Ditto.
321 (flip_source_destroy): Ditto.
323 Sat Dec 16 12:54:27 2006 Ben Pfaff <blp@gnu.org>
325 * rank.q (rank_custom_variables): Allow grouping variables to be
326 strings. Fixes bug #18533. Thanks to John Darrington for review.
328 Sat Dec 9 18:47:51 2006 Ben Pfaff <blp@gnu.org>
330 * regression.q (is_depvar): Compare variable pointers instead of
333 Thu Dec 7 15:26:25 WST 2006 John Darrington <john@darrington.wattle.id.au>
335 * examine.q: Allocated the categorical values for the dependent and
336 independent variables, on the heap. Hence they can be of any width.
338 Wed Dec 6 21:14:26 2006 Ben Pfaff <blp@gnu.org>
340 * regression.q (reg_inserted): Compare variable pointers instead
343 Mon Dec 4 22:33:46 2006 Ben Pfaff <blp@gnu.org>
345 * crosstabs.q (insert_summary): Use var_to_string for labeling.
346 (output_pivot_table) Ditto.
349 * frequencies.q (setup_z_trns): Ditto.
351 (dump_condensed) Ditto.
352 (dump_statistics) Ditto.
355 Sun Nov 5 08:31:42 WST 2006 John Darrington <john@darrington.wattle.id.au>
357 * t-test.q, oneway.q: Changed to use the new casefilter structure.
359 Sat Oct 14 16:52:28 2006 Ben Pfaff <blp@gnu.org>
361 * rank.q: (rank_sorted_casefile) Add some missing case_destroy()
362 calls to fix a memory leak.
364 Sun Oct 8 09:45:40 WST 2006 John Darrington <john@darrington.wattle.id.au>
366 * rank.q: Plugged a small memory leak which occurred under error
369 Sat Oct 7 11:06:01 WST 2006 John Darrington <john@darrington.wattle.id.au>
371 * rank.q: Implemented most of the RANK command.
373 2006-07-14 Jason Stover <jhs@math.gcsu.edu>
375 * regression.q (run_regression): New function to move knowledge of
376 pspp_linreg_cache out of math/coefficient.[ch].
378 Sat Jul 1 17:41:46 2006 Ben Pfaff <blp@gnu.org>
380 Fix bug #11612, "q2c documentation does not agree with code".
382 * examine.q: Audit use of q2c "+" prefixes that indicate that a
383 command may appear multiple times.
385 * frequencies.q: Ditto.
389 * regression.q: Ditto.
393 Fri Jun 23 14:18:22 2006 Ben Pfaff <blp@gnu.org>
395 Support long string variables on FREQUENCIES, as
396 an extension when in enhanced algorithms mode. For Greg Hunt
397 <greg@firmansyah.com>.
399 * frequencies.q: (struct freq) Change `v' member from union value
400 to union value *. Update all references.
401 (struct var_freqs) Add width, print members to represent effective
402 variable width and display format.
403 (calc) Copy entire long string value into the hash table.
404 (frq_custom_variables) Set new width, print members.
405 (hash_value_alpha) Get width from var_freqs.
406 (compare_value_alpha_a) Ditto.
407 (compare_freq_alpha_a) Ditto.
408 (compare_freq_alpha_d) Ditto.
409 (dump_full) Get display format from var_freqs.
410 (dump_condensed) Ditto.
412 Mon Jun 19 22:07:13 2006 Ben Pfaff <blp@gnu.org>
414 * frequencies.q: (dump_full) Only put the first MAX_SHORT_STRING
415 bytes of string variables into the output cells, seeing as we only
417 (dump_condensed) Ditto.
419 Mon Jun 19 21:52:05 2006 Ben Pfaff <blp@gnu.org>
421 Fixes a bug reported by Greg Hunt <greg@firmansyah.com>.
423 * frequencies.q: (hsh_hash_bytes) We only copy the first
424 MAX_SHORT_STRING bytes of string variables, so we must only
425 compare that many bytes, even if the string variable is longer.
426 (compare_value_alpha_a) Ditto.
427 (compare_freq_alpha_a) Ditto.
428 (compare_freq_alpha_d) Ditto.
430 2006-05-11 Jason Stover <jhs@math.gcsu.edu>
432 * regression.q: Adjusted code to account for cache->coeff being a
433 pspp_linreg_coeff **.
435 Sun May 7 18:31:25 2006 Ben Pfaff <blp@gnu.org>
439 * aggregate.c (cmd_aggregate): Free default_dict before replacing
442 Sun May 7 17:09:19 2006 Ben Pfaff <blp@gnu.org>
444 * flip.c (flip_file): Check return value of pool_fclose().
446 Sat May 6 16:00:13 2006 Ben Pfaff <blp@gnu.org>
448 Get rid of `char *c' member in union value, for cleanliness.
450 * aggregate.c: (union agr_argument) New union.
451 (struct agr_var) Change element type of arg[] from union value to
453 (parse_aggregate_functions) Change local variable types likewise.
455 * autorecode.c: (union arc_value) New union.
456 (struct arc_item) Change "from" from union value to union
458 (recode) Change local variable from union value to union
460 (autorecode_trns_proc) Ditto.
461 (compare_alpha_value) Ditto.
462 (hash_alpha_value) Ditto.
463 (compare_numeric_value) Ditto.
464 (hash_numeric_value) Ditto.
465 (autorecode_proc_func) Ditto.
467 Sat May 6 10:43:33 2006 Ben Pfaff <blp@gnu.org>
469 Continue reforming procedure execution. In this phase, get rid of
470 the output code for SPLIT FILE groups in procedure.c, which really
471 shouldn't be doing any output. Move it into the individual
472 procedures instead. This also adds some flexibility.
474 * crosstabs.q (precalc): Call output_split_file_values().
476 * descriptives.c (calc_descriptives): Ditto.
478 * examine.q (run_examine): Ditto.
480 * frequencies.q (precalc): Ditto.
482 * oneway.q (run_oneway): Ditto.
484 * regression.q (run_regression): Ditto.
486 * t-test.q (calculate): Ditto.
488 Wed May 3 23:05:31 2006 Ben Pfaff <blp@gnu.org>
490 Continue reforming procedure execution. In this phase, get rid of
491 many global variables, consolidating procedure execution in
492 procedure.c. Encapsulate transformations in new "struct
493 trns_chain". Also, change implementation of N OF CASES, FILTER,
494 and PROCESS IF from special cases to transformations.
496 * aggregate.c (cmd_aggregate) Use discard_variables().
498 2006-04-28 Jason Stover <jhs@math.gcsu.edu>
500 * regression.q (regression_trns_resid_proc): Pass only the
501 variables used in the model to (*model->residual)().
503 * regression.q (regression_trns_pred_proc): Pass only the
504 variables used in the model to (*model->pred)().
506 2006-04-26 Jason Stover <jhs@math.gcsu.edu>
508 * regression.q: Added support for multiple transformations.
510 * regression.q (regression_trns_resid_proc): New function.
512 * regression.q (regression_trns_pred_proc): New function.
514 * regression.q (subcommand_save): Added support for saving
517 * regression.q (regression_trns_free): New function.
519 * regression.q (reg_get_name): New function.
521 * regression.q (reg_save_var): New function.
523 Tue Apr 25 13:18:56 2006 Ben Pfaff <blp@gnu.org>
525 * rank.q (parse_rank_function): Use SE instead of ME for parse
528 Tue Apr 25 13:16:28 2006 Ben Pfaff <blp@gnu.org>
530 * flip.c (flip_sink_write): Use snprintf() to simplify a bit of
533 2006-04-21 Jason Stover <jhs@math.gcsu.edu>
535 * regression.q (try_name): New function. (Partly copied from
536 try_name in descriptives.c.)
538 * regression.q (subcommand_save): Choose residual variable names
541 2006-04-20 Jason Stover <jhs@math.gcsu.edu>
543 * regression.q (cmd_regression): Moved call to subcommand_save()
544 outside multipass_procedure_with_splits().
546 * regression.q (regression_trns_proc): Fixed value counter n_vals
547 before calling *model->residual().
549 2006-04-19 Jason Stover <jhs@math.gcsu.edu>
551 * regression.q (regression_trns_proc): Fixed the look-up of the
554 2006-04-18 Jason Stover <jhs@math.gcsu.edu>
556 * regression.q (regression_trns_proc): Look up the residual
557 variable in the linear regression cache.
559 * regression.q (subcommand_save): Set the residual variable in the
560 linear regression cache.
562 2006-04-17 Jason Stover <jason@wonko.gcsu.edu>
564 * regression.q (regression_trns_proc): Accept case_idx as an int
565 to match the definition of trns_proc_func.
567 2006-04-17 Jason Stover <jhs@math.gcsu.edu>
569 * regression.q (regression_trns_proc): New function.
571 * regression.q (subcommand_save): Create variable residuals and
572 add a transformation to assign values to them. Also free the
573 linreg_cache if the SAVE command was not called. Removed the
574 casereading loop. Placed actual computation of residuals in
575 regression_trns_proc.
577 * regression.q (run_regression): Moved call to free
578 pspp_linreg_cache to subcommand_save.
580 Sat Apr 15 18:01:03 2006 Ben Pfaff <blp@gnu.org>
582 * examine.q (output_examine): Add casts to fix warnings.
584 2006-04-07 Jason Stover <jhs@math.gcsu.edu>
586 * regression.q (subcommand_save): New function.
588 2006-04-04 Jason Stover <jhs@math.gcsu.edu>
590 * regression.q: New function reg_has_categorical () to tell
591 whether a pspp_linreg_struct was made with any variables of type
594 * regression.q: (subcommand_export): Call
595 reg_print_categorical_encoding() only if the model uses any
596 categorical variables.
598 Mon Mar 27 16:00:42 2006 Ben Pfaff <blp@gnu.org>
600 * crosstabs.q: (output_pivot_table) Drop spurious space from
603 2006-03-15 Jason Stover <jhs@math.gcsu.edu>
605 * regression.q: Added custom syntax parser for VARIABLES subcommand
607 * regression.q: Moved most instructions for run_regression ()
608 inside the loop over dependent variables.
610 Thu Mar 2 08:40:33 WST 2006 John Darrington <john@darrington.wattle.id.au>
612 * Moved files from src directory