1 2008-02-14 John Darrington <john@darrington.wattle.id.au>
3 * examine.q: Fixed counts of missing variables. Thanks to
4 Jason Stover for reporting this problem.
6 2008-01-02 John Darrington <john@darrington.wattle.id.au>
8 * binomial.c chisquare.c examine.q frequencies.q oneway.q regression.q : updated
9 all users of var_get_value_name to use replacement function var_append_value_name.
11 2007-12-07 Ben Pfaff <blp@gnu.org>
15 * crosstabs.q (precalc): Initialize data structures even if the
16 first case cannot be read.
18 * frequencies.q (precalc): Ditto.
20 2007-11-03 Ben Pfaff <blp@gnu.org>
22 Allow output files to overwrite input files (bug #21280).
24 * aggregate.c (cmd_aggregate): Manage file handle reference
27 * correlations.q (internal_cmd_frequencies): Ditto.
28 (cor_custom_matrix): Ditto.
30 * regression.q (regression_custom_export): Ditto.
31 (cmd_regression): Ditto.
33 2007-10-12 Ben Pfaff <blp@gnu.org>
35 * flip.c (flip_file): No need to conditionally substitute for
36 "fseeko" and "off_t" manually anymore, as gnulib takes care of it
39 2007-09-21 Jason Stover <jhs@wonko.gcsu.edu>
41 * regression.q (run_regression): Partial fix of memory leak, bug
44 2007-09-19 Ben Pfaff <blp@gnu.org>
48 * aggregate.c (cmd_aggregate): Destroy casereader consistently,
49 even if casereader fails.
51 * examine.q (run_examine): Ditto.
53 * glm.q (run_glm): Ditto.
55 * oneway.q (run_oneway): Ditto.
57 * regression.q (run_regression): Ditto.
59 * t-test.q (calculate): Ditto.
61 * descriptives.c (calc_descriptives): Ditto. Also avoid
62 gratuitous casereader_clone.
64 2007-09-13 Jason Stover <jhs@debs.hoobahooba.net>
66 * regression.q (cmd_regression): Move declaration of models in to
67 definition of cmd_regression.
69 * regression.q (run_regression): Free mom to fix memory leak.
71 2007-09-12 Ben Pfaff <blp@gnu.org>
73 * crosstabs.q (postcalc): Free sorted_tab and the structures that
74 it points to, to plug a memory leak. Fixes bug #20910. Thanks to
75 John Darrington for reporting this bug and for review.
77 2007-09-04 Ben Pfaff <blp@gnu.org>
79 * crosstabs.q (cmd_crosstabs): Free xtab and the structures that
80 it points to, to plug a memory leak. Fixes bug #18315.
82 2007-08-15 Jason Stover <jhs@wonko.gcsu.edu>
84 * regression.q (identify_indep_vars): Print an error if dependent
85 and independent variables are the same. Fixes bug 19819.
87 2007-08-12 Ben Pfaff <blp@gnu.org>
89 * flip.c: Drop use of dict_get_compacted_dict_index_to_case_index
90 and just use the ordinary case indexes. There seemed to be no
91 reason for the former method.
93 2007-08-03 Ben Pfaff <blp@gnu.org>
95 * rank.q (rank_cmd): Instead of sorting by SPLIT FILE vars, group
96 by them. Fixes bug #17239.
97 Reviewed by John Darrington.
99 2007-08-01 Ben Pfaff <blp@gnu.org>
101 Clean up handling of median, by treating it almost like any other
102 percentile. Fixes bug #17424. Thanks to John Darrington for
104 * frequencies.q (internal_cmd_frequencies): Fix handling of bit
105 masks for `stats' variable. If median is selected, turn it off
106 and add a 50th percentile.
107 (add_percentile): Simplify code a little.
108 (calc_stats): Drop special casing of median.
109 (dump_statistics): Ditto, except that we label the 50th percentile
110 as "50 (Median)" to make it clear that it's also the median.
112 2007-07-31 Ben Pfaff <blp@gnu.org>
114 Remove integer mode from FREQUENCIES and incidentally fix bug
115 #17421. Reviewed by John Darrington.
116 * frequencies.q (int_pool): Rename data_pool.
117 (gen_pool): Rename syntax_pool.
118 (enum FRQM_*): Removed.
119 (struct freq_tab): Removed `mode', `vector', `min', `max',
120 `out_of_range', `sysmis' members.
121 (calc): Delete support for integer mode.
123 (postprocess_freq_tab): Ditto.
124 (cleanup_freq_tab): Ditto.
125 (frq_custom_variables): Ditto.
127 2007-07-28 John Darrington <john@darrington.wattle.id.au>
129 * t-test.q: Moved the order in which groups are displayed in the
130 independent samples case, where a cut point is given.
132 2007-07-27 Ben Pfaff <blp@gnu.org>
134 * regression.q (run_regression): Move casereader_destroy call so
135 that it always gets called, not just if there was some valid
136 data. Fixes bug #19581.
137 Reviewed by Jason Stover.
139 2007-07-24 Ben Pfaff <blp@gnu.org>
141 * flip.c (struct flip_pgm): Remove `case_size' member (now
143 (cmd_flip): Pass var_cnt as number of cases instead of case_cnt,
144 to fix bug #20494. Don't assign to `case_size' member, which was
145 unused after assignment.
146 (build_dictionary): When NEWNAMES not used, get the number of
147 variables right, to fix bug #20493.
149 2007-07-10 Jason Stover <jhs@math.gcsu.edu>
151 * glm.q: Initial version of the GLM procedure.
153 2007-06-06 Ben Pfaff <blp@gnu.org>
155 Adapt case sources, sinks, and clients of procedure code to the
158 * aggregate.c: Simplify greatly since everything is more uniform
161 * autorecode.c: Adapt to new procedure code.
163 * chisquare.c: Ditto.
164 * crosstabs.q: Ditto.
165 * descriptives.c: Ditto.
167 * npar-summary.c: Ditto.
168 * frequencies.q: Ditto.
171 * regression.q: Ditto.
172 * sort-cases.c: Ditto.
175 * sort-criteria.c: Rewrite to output a struct case_ordering.
177 * flip.c: Rewrite to be a casereader.
179 * rank.q: Simplify greatly since casereaders are much more
180 flexible than what we had before.
182 2007-05-15 Jason Stover <jhs@math.gcsu.edu>
184 * regression.q (run_regression): Tell the user when the data
185 contain no valid cases.
187 2007-05-08 Jason Stover <jhs@math.gcsu.edu>
189 * regression.q: Partial fix of bug which caused a crash if
190 dependent variable and independent variable were the same.
192 2007-04-16 John Darrington <john@darrington.wattle.id.au>
194 * t-test.q: Changed the output width of reported counts and
195 degrees of freedom, to avoid truncating these values. Thanks
196 to Seth Woolley for reporting this problem. A proper fix involves
197 re-thinking the output driver.
199 2007-04-12 Jason Stover <jhs@math.gcsu.edu>
201 * regression.q (run_regression): Added if (n_data >0) to fix bug
204 2007-03-29 Jason Stover <jhs@math.gcsu.edu>
206 * regression.q (prepare_data): New function.
208 * regression.q (compute_moments): New function.
210 2007-03-18 Ben Pfaff <blp@gnu.org>
212 * crosstabs.q (static var write): Rename write_style to avoid
213 conflict with POSIX function of same name.
215 2007-03-16 Jason Stover <jhs@math.gcsu.edu>
217 * regression.q (run_regression): Added support for moments.
219 Sat Feb 17 08:16:00 2007 Ben Pfaff <blp@gnu.org>
221 * flip.c (flip_sink_create): Improve error message when temporary
222 file cannot be created.
224 Tue Feb 6 19:58:03 2007 Ben Pfaff <blp@gnu.org>
226 * flip.c (flip_file): Give better error message on end-of-file.
228 2007-02-04 Jason Stover <jhs@math.gcsu.edu>
230 * regression.q: Fixed p-value computation in the test for
231 individual regression coefficients.
233 Mon Jan 15 11:03:20 2007 Ben Pfaff <blp@gnu.org>
235 Fix bugs found by valgrind when --enable-debug is used with the
236 new case code. These bugs are hidden when the data set is small
237 enough to find in memory; if a bigger data set that would overflow
238 to disk were used, then data corruption would occur.
240 * chisquare.c (create_freq_hash): Pass free_freq_mutable_hash to
241 hsh_create as free function. Make copy of data put into hash.
243 * oneway.q (free_value): New function.
244 (run_oneway): Use free_value as arg to hsh_create. Make copy of
247 * rank.q (rank_cases): Don't access data in case after we've given
250 Tue Jan 9 19:16:11 2007 Ben Pfaff <blp@gnu.org>
254 * aggregate.c (parse_aggregate_functions) Initialize
257 Fri Dec 22 14:04:09 2006 Ben Pfaff <blp@gnu.org>
259 Simplify missing value handling.
261 * aggregate.c (struct agr_var): Remove `bool include_missing', add
262 `enum mv_class exclude'. Remove `int missing', add `bool
263 saw_missing'. Update users.
265 * descriptives.c (struct dsc_trns): Removed `int
266 include_user_missing', add `enum mv_class exclude'. Update users.
267 (struct dsc_proc): Ditto.
269 * examine.q: (static var value_is_missing): Rename
270 `exclude_values', change type to `enum mv_class'. Update users.
274 Fri Dec 22 19:22:18 WST 2006 John Darrington <john@darrington.wattle.id.au>
276 * frequencies.q : Fixed bug #17420, where the table bounds were overun
277 when /FORMAT=nolabels was given.
279 Wed Dec 20 18:45:31 WST 2006 John Darrington <john@darrington.wattle.id.au>
281 * binomial.c binomial.h : New files. Thanks to Jason Stover
282 for assistance with these.
284 * chisquare.c chisquare.h freq.c freq.h npar-summary.c npar-summary.h
285 npar.h npar.q: New files. Implementing NPAR TESTS.
287 * frequencies.q : Moved structure definitions into freq.[ch]
289 Sat Dec 16 22:26:44 2006 Ben Pfaff <blp@gnu.org>
291 Make it possible to pull cases from the active file with a
292 function call, instead of requiring indirection through a callback
295 * aggregate.c (cmd_aggregate): Take advantage of new procedure
297 (agr_to_active_file): Removed.
298 (presorted_agr_to_sysfile): Removed.
300 * autorecode.c (cmd_autorecode): Take advantage of new procedure
302 (autorecode_proc_func): Removed.
304 * flip.c (struct flip_pgm): New members to allow conformance with
305 new case_source_class interface.
306 (cmd_flip): Adapt to new case_source_class interface.
307 (flip_source_read): Ditto.
308 (flip_source_destroy): Ditto.
310 Sat Dec 16 12:54:27 2006 Ben Pfaff <blp@gnu.org>
312 * rank.q (rank_custom_variables): Allow grouping variables to be
313 strings. Fixes bug #18533. Thanks to John Darrington for review.
315 Sat Dec 9 18:47:51 2006 Ben Pfaff <blp@gnu.org>
317 * regression.q (is_depvar): Compare variable pointers instead of
320 Thu Dec 7 15:26:25 WST 2006 John Darrington <john@darrington.wattle.id.au>
322 * examine.q: Allocated the categorical values for the dependent and
323 independent variables, on the heap. Hence they can be of any width.
325 Wed Dec 6 21:14:26 2006 Ben Pfaff <blp@gnu.org>
327 * regression.q (reg_inserted): Compare variable pointers instead
330 Mon Dec 4 22:33:46 2006 Ben Pfaff <blp@gnu.org>
332 * crosstabs.q (insert_summary): Use var_to_string for labeling.
333 (output_pivot_table) Ditto.
336 * frequencies.q (setup_z_trns): Ditto.
338 (dump_condensed) Ditto.
339 (dump_statistics) Ditto.
342 Sun Nov 5 08:31:42 WST 2006 John Darrington <john@darrington.wattle.id.au>
344 * t-test.q, oneway.q: Changed to use the new casefilter structure.
346 Sat Oct 14 16:52:28 2006 Ben Pfaff <blp@gnu.org>
348 * rank.q: (rank_sorted_casefile) Add some missing case_destroy()
349 calls to fix a memory leak.
351 Sun Oct 8 09:45:40 WST 2006 John Darrington <john@darrington.wattle.id.au>
353 * rank.q: Plugged a small memory leak which occurred under error
356 Sat Oct 7 11:06:01 WST 2006 John Darrington <john@darrington.wattle.id.au>
358 * rank.q: Implemented most of the RANK command.
360 2006-07-14 Jason Stover <jhs@math.gcsu.edu>
362 * regression.q (run_regression): New function to move knowledge of
363 pspp_linreg_cache out of math/coefficient.[ch].
365 Sat Jul 1 17:41:46 2006 Ben Pfaff <blp@gnu.org>
367 Fix bug #11612, "q2c documentation does not agree with code".
369 * examine.q: Audit use of q2c "+" prefixes that indicate that a
370 command may appear multiple times.
372 * frequencies.q: Ditto.
376 * regression.q: Ditto.
380 Fri Jun 23 14:18:22 2006 Ben Pfaff <blp@gnu.org>
382 Support long string variables on FREQUENCIES, as
383 an extension when in enhanced algorithms mode. For Greg Hunt
384 <greg@firmansyah.com>.
386 * frequencies.q: (struct freq) Change `v' member from union value
387 to union value *. Update all references.
388 (struct var_freqs) Add width, print members to represent effective
389 variable width and display format.
390 (calc) Copy entire long string value into the hash table.
391 (frq_custom_variables) Set new width, print members.
392 (hash_value_alpha) Get width from var_freqs.
393 (compare_value_alpha_a) Ditto.
394 (compare_freq_alpha_a) Ditto.
395 (compare_freq_alpha_d) Ditto.
396 (dump_full) Get display format from var_freqs.
397 (dump_condensed) Ditto.
399 Mon Jun 19 22:07:13 2006 Ben Pfaff <blp@gnu.org>
401 * frequencies.q: (dump_full) Only put the first MAX_SHORT_STRING
402 bytes of string variables into the output cells, seeing as we only
404 (dump_condensed) Ditto.
406 Mon Jun 19 21:52:05 2006 Ben Pfaff <blp@gnu.org>
408 Fixes a bug reported by Greg Hunt <greg@firmansyah.com>.
410 * frequencies.q: (hsh_hash_bytes) We only copy the first
411 MAX_SHORT_STRING bytes of string variables, so we must only
412 compare that many bytes, even if the string variable is longer.
413 (compare_value_alpha_a) Ditto.
414 (compare_freq_alpha_a) Ditto.
415 (compare_freq_alpha_d) Ditto.
417 2006-05-11 Jason Stover <jhs@math.gcsu.edu>
419 * regression.q: Adjusted code to account for cache->coeff being a
420 pspp_linreg_coeff **.
422 Sun May 7 18:31:25 2006 Ben Pfaff <blp@gnu.org>
426 * aggregate.c (cmd_aggregate): Free default_dict before replacing
429 Sun May 7 17:09:19 2006 Ben Pfaff <blp@gnu.org>
431 * flip.c (flip_file): Check return value of pool_fclose().
433 Sat May 6 16:00:13 2006 Ben Pfaff <blp@gnu.org>
435 Get rid of `char *c' member in union value, for cleanliness.
437 * aggregate.c: (union agr_argument) New union.
438 (struct agr_var) Change element type of arg[] from union value to
440 (parse_aggregate_functions) Change local variable types likewise.
442 * autorecode.c: (union arc_value) New union.
443 (struct arc_item) Change "from" from union value to union
445 (recode) Change local variable from union value to union
447 (autorecode_trns_proc) Ditto.
448 (compare_alpha_value) Ditto.
449 (hash_alpha_value) Ditto.
450 (compare_numeric_value) Ditto.
451 (hash_numeric_value) Ditto.
452 (autorecode_proc_func) Ditto.
454 Sat May 6 10:43:33 2006 Ben Pfaff <blp@gnu.org>
456 Continue reforming procedure execution. In this phase, get rid of
457 the output code for SPLIT FILE groups in procedure.c, which really
458 shouldn't be doing any output. Move it into the individual
459 procedures instead. This also adds some flexibility.
461 * crosstabs.q (precalc): Call output_split_file_values().
463 * descriptives.c (calc_descriptives): Ditto.
465 * examine.q (run_examine): Ditto.
467 * frequencies.q (precalc): Ditto.
469 * oneway.q (run_oneway): Ditto.
471 * regression.q (run_regression): Ditto.
473 * t-test.q (calculate): Ditto.
475 Wed May 3 23:05:31 2006 Ben Pfaff <blp@gnu.org>
477 Continue reforming procedure execution. In this phase, get rid of
478 many global variables, consolidating procedure execution in
479 procedure.c. Encapsulate transformations in new "struct
480 trns_chain". Also, change implementation of N OF CASES, FILTER,
481 and PROCESS IF from special cases to transformations.
483 * aggregate.c (cmd_aggregate) Use discard_variables().
485 2006-04-28 Jason Stover <jhs@math.gcsu.edu>
487 * regression.q (regression_trns_resid_proc): Pass only the
488 variables used in the model to (*model->residual)().
490 * regression.q (regression_trns_pred_proc): Pass only the
491 variables used in the model to (*model->pred)().
493 2006-04-26 Jason Stover <jhs@math.gcsu.edu>
495 * regression.q: Added support for multiple transformations.
497 * regression.q (regression_trns_resid_proc): New function.
499 * regression.q (regression_trns_pred_proc): New function.
501 * regression.q (subcommand_save): Added support for saving
504 * regression.q (regression_trns_free): New function.
506 * regression.q (reg_get_name): New function.
508 * regression.q (reg_save_var): New function.
510 Tue Apr 25 13:18:56 2006 Ben Pfaff <blp@gnu.org>
512 * rank.q (parse_rank_function): Use SE instead of ME for parse
515 Tue Apr 25 13:16:28 2006 Ben Pfaff <blp@gnu.org>
517 * flip.c (flip_sink_write): Use snprintf() to simplify a bit of
520 2006-04-21 Jason Stover <jhs@math.gcsu.edu>
522 * regression.q (try_name): New function. (Partly copied from
523 try_name in descriptives.c.)
525 * regression.q (subcommand_save): Choose residual variable names
528 2006-04-20 Jason Stover <jhs@math.gcsu.edu>
530 * regression.q (cmd_regression): Moved call to subcommand_save()
531 outside multipass_procedure_with_splits().
533 * regression.q (regression_trns_proc): Fixed value counter n_vals
534 before calling *model->residual().
536 2006-04-19 Jason Stover <jhs@math.gcsu.edu>
538 * regression.q (regression_trns_proc): Fixed the look-up of the
541 2006-04-18 Jason Stover <jhs@math.gcsu.edu>
543 * regression.q (regression_trns_proc): Look up the residual
544 variable in the linear regression cache.
546 * regression.q (subcommand_save): Set the residual variable in the
547 linear regression cache.
549 2006-04-17 Jason Stover <jason@wonko.gcsu.edu>
551 * regression.q (regression_trns_proc): Accept case_idx as an int
552 to match the definition of trns_proc_func.
554 2006-04-17 Jason Stover <jhs@math.gcsu.edu>
556 * regression.q (regression_trns_proc): New function.
558 * regression.q (subcommand_save): Create variable residuals and
559 add a transformation to assign values to them. Also free the
560 linreg_cache if the SAVE command was not called. Removed the
561 casereading loop. Placed actual computation of residuals in
562 regression_trns_proc.
564 * regression.q (run_regression): Moved call to free
565 pspp_linreg_cache to subcommand_save.
567 Sat Apr 15 18:01:03 2006 Ben Pfaff <blp@gnu.org>
569 * examine.q (output_examine): Add casts to fix warnings.
571 2006-04-07 Jason Stover <jhs@math.gcsu.edu>
573 * regression.q (subcommand_save): New function.
575 2006-04-04 Jason Stover <jhs@math.gcsu.edu>
577 * regression.q: New function reg_has_categorical () to tell
578 whether a pspp_linreg_struct was made with any variables of type
581 * regression.q: (subcommand_export): Call
582 reg_print_categorical_encoding() only if the model uses any
583 categorical variables.
585 Mon Mar 27 16:00:42 2006 Ben Pfaff <blp@gnu.org>
587 * crosstabs.q: (output_pivot_table) Drop spurious space from
590 2006-03-15 Jason Stover <jhs@math.gcsu.edu>
592 * regression.q: Added custom syntax parser for VARIABLES subcommand
594 * regression.q: Moved most instructions for run_regression ()
595 inside the loop over dependent variables.
597 Thu Mar 2 08:40:33 WST 2006 John Darrington <john@darrington.wattle.id.au>
599 * Moved files from src directory