1 2008-06-21 Jason Stover <jhs@math.gcsu.edu>
3 * regression.q (reg_stats_coeff): Use new accessor function
4 pspp_coeff_get_sd. Fixed bug 23567. No longer call compute_moments
5 in run_regression. Pass entire design_matrix to pspp_linreg ().
7 2008-05-29 John Darrington <john@darrington.wattle.id.au>
9 * examine.q: Fixed bug where incorrect levels of dereferencing
10 were applied to pointers.
12 2008-04-09 John Darrington <john@darrington.wattle.id.au>
14 * regression.q: Fix display of degrees of freedom.
16 2008-04-08 Jason Stover <jhs@math.gcsu.edu>
18 * regression.q (identify_indep_vars): Don't panic unless
21 2008-03-16 Ben Pfaff <blp@gnu.org>
23 Bug #22037. Thanks to John Darrington for reporting this bug.
25 * crosstabs.q (calc_general): Only the short string prefix of long
26 string variables are tabulated, so we must not copy or zero out
29 2008-03-10 Jason Stover <jhs@math.gcsu.edu>
31 * regression.q (run_regression): Removed code for EXPORT
32 subcommand. Remove use of coefficient 0 as the intercept.
34 2008-02-14 John Darrington <john@darrington.wattle.id.au>
36 * examine.q: Fixed counts of missing variables. Thanks to
37 Jason Stover for reporting this problem.
39 2008-01-02 John Darrington <john@darrington.wattle.id.au>
41 * binomial.c chisquare.c examine.q frequencies.q oneway.q regression.q : updated
42 all users of var_get_value_name to use replacement function var_append_value_name.
44 2007-12-07 Ben Pfaff <blp@gnu.org>
48 * crosstabs.q (precalc): Initialize data structures even if the
49 first case cannot be read.
51 * frequencies.q (precalc): Ditto.
53 2007-11-03 Ben Pfaff <blp@gnu.org>
55 Allow output files to overwrite input files (bug #21280).
57 * aggregate.c (cmd_aggregate): Manage file handle reference
60 * correlations.q (internal_cmd_frequencies): Ditto.
61 (cor_custom_matrix): Ditto.
63 * regression.q (regression_custom_export): Ditto.
64 (cmd_regression): Ditto.
66 2007-10-12 Ben Pfaff <blp@gnu.org>
68 * flip.c (flip_file): No need to conditionally substitute for
69 "fseeko" and "off_t" manually anymore, as gnulib takes care of it
72 2007-09-21 Jason Stover <jhs@wonko.gcsu.edu>
74 * regression.q (run_regression): Partial fix of memory leak, bug
77 2007-09-19 Ben Pfaff <blp@gnu.org>
81 * aggregate.c (cmd_aggregate): Destroy casereader consistently,
82 even if casereader fails.
84 * examine.q (run_examine): Ditto.
86 * glm.q (run_glm): Ditto.
88 * oneway.q (run_oneway): Ditto.
90 * regression.q (run_regression): Ditto.
92 * t-test.q (calculate): Ditto.
94 * descriptives.c (calc_descriptives): Ditto. Also avoid
95 gratuitous casereader_clone.
97 2007-09-13 Jason Stover <jhs@math.gcsu.edu>
99 * regression.q (cmd_regression): Move declaration of models in to
100 definition of cmd_regression.
102 * regression.q (run_regression): Free mom to fix memory leak.
104 2007-09-12 Ben Pfaff <blp@gnu.org>
106 * crosstabs.q (postcalc): Free sorted_tab and the structures that
107 it points to, to plug a memory leak. Fixes bug #20910. Thanks to
108 John Darrington for reporting this bug and for review.
110 2007-09-04 Ben Pfaff <blp@gnu.org>
112 * crosstabs.q (cmd_crosstabs): Free xtab and the structures that
113 it points to, to plug a memory leak. Fixes bug #18315.
115 2007-08-15 Jason Stover <jhs@wonko.gcsu.edu>
117 * regression.q (identify_indep_vars): Print an error if dependent
118 and independent variables are the same. Fixes bug 19819.
120 2007-08-12 Ben Pfaff <blp@gnu.org>
122 * flip.c: Drop use of dict_get_compacted_dict_index_to_case_index
123 and just use the ordinary case indexes. There seemed to be no
124 reason for the former method.
126 2007-08-03 Ben Pfaff <blp@gnu.org>
128 * rank.q (rank_cmd): Instead of sorting by SPLIT FILE vars, group
129 by them. Fixes bug #17239.
130 Reviewed by John Darrington.
132 2007-08-01 Ben Pfaff <blp@gnu.org>
134 Clean up handling of median, by treating it almost like any other
135 percentile. Fixes bug #17424. Thanks to John Darrington for
137 * frequencies.q (internal_cmd_frequencies): Fix handling of bit
138 masks for `stats' variable. If median is selected, turn it off
139 and add a 50th percentile.
140 (add_percentile): Simplify code a little.
141 (calc_stats): Drop special casing of median.
142 (dump_statistics): Ditto, except that we label the 50th percentile
143 as "50 (Median)" to make it clear that it's also the median.
145 2007-07-31 Ben Pfaff <blp@gnu.org>
147 Remove integer mode from FREQUENCIES and incidentally fix bug
148 #17421. Reviewed by John Darrington.
149 * frequencies.q (int_pool): Rename data_pool.
150 (gen_pool): Rename syntax_pool.
151 (enum FRQM_*): Removed.
152 (struct freq_tab): Removed `mode', `vector', `min', `max',
153 `out_of_range', `sysmis' members.
154 (calc): Delete support for integer mode.
156 (postprocess_freq_tab): Ditto.
157 (cleanup_freq_tab): Ditto.
158 (frq_custom_variables): Ditto.
160 2007-07-28 John Darrington <john@darrington.wattle.id.au>
162 * t-test.q: Moved the order in which groups are displayed in the
163 independent samples case, where a cut point is given.
165 2007-07-27 Ben Pfaff <blp@gnu.org>
167 * regression.q (run_regression): Move casereader_destroy call so
168 that it always gets called, not just if there was some valid
169 data. Fixes bug #19581.
170 Reviewed by Jason Stover.
172 2007-07-24 Ben Pfaff <blp@gnu.org>
174 * flip.c (struct flip_pgm): Remove `case_size' member (now
176 (cmd_flip): Pass var_cnt as number of cases instead of case_cnt,
177 to fix bug #20494. Don't assign to `case_size' member, which was
178 unused after assignment.
179 (build_dictionary): When NEWNAMES not used, get the number of
180 variables right, to fix bug #20493.
182 2007-07-10 Jason Stover <jhs@math.gcsu.edu>
184 * glm.q: Initial version of the GLM procedure.
186 2007-06-06 Ben Pfaff <blp@gnu.org>
188 Adapt case sources, sinks, and clients of procedure code to the
191 * aggregate.c: Simplify greatly since everything is more uniform
194 * autorecode.c: Adapt to new procedure code.
196 * chisquare.c: Ditto.
197 * crosstabs.q: Ditto.
198 * descriptives.c: Ditto.
200 * npar-summary.c: Ditto.
201 * frequencies.q: Ditto.
204 * regression.q: Ditto.
205 * sort-cases.c: Ditto.
208 * sort-criteria.c: Rewrite to output a struct case_ordering.
210 * flip.c: Rewrite to be a casereader.
212 * rank.q: Simplify greatly since casereaders are much more
213 flexible than what we had before.
215 2007-05-15 Jason Stover <jhs@math.gcsu.edu>
217 * regression.q (run_regression): Tell the user when the data
218 contain no valid cases.
220 2007-05-08 Jason Stover <jhs@math.gcsu.edu>
222 * regression.q: Partial fix of bug which caused a crash if
223 dependent variable and independent variable were the same.
225 2007-04-16 John Darrington <john@darrington.wattle.id.au>
227 * t-test.q: Changed the output width of reported counts and
228 degrees of freedom, to avoid truncating these values. Thanks
229 to Seth Woolley for reporting this problem. A proper fix involves
230 re-thinking the output driver.
232 2007-04-12 Jason Stover <jhs@math.gcsu.edu>
234 * regression.q (run_regression): Added if (n_data >0) to fix bug
237 2007-03-29 Jason Stover <jhs@math.gcsu.edu>
239 * regression.q (prepare_data): New function.
241 * regression.q (compute_moments): New function.
243 2007-03-18 Ben Pfaff <blp@gnu.org>
245 * crosstabs.q (static var write): Rename write_style to avoid
246 conflict with POSIX function of same name.
248 2007-03-16 Jason Stover <jhs@math.gcsu.edu>
250 * regression.q (run_regression): Added support for moments.
252 Sat Feb 17 08:16:00 2007 Ben Pfaff <blp@gnu.org>
254 * flip.c (flip_sink_create): Improve error message when temporary
255 file cannot be created.
257 Tue Feb 6 19:58:03 2007 Ben Pfaff <blp@gnu.org>
259 * flip.c (flip_file): Give better error message on end-of-file.
261 2007-02-04 Jason Stover <jhs@math.gcsu.edu>
263 * regression.q: Fixed p-value computation in the test for
264 individual regression coefficients.
266 Mon Jan 15 11:03:20 2007 Ben Pfaff <blp@gnu.org>
268 Fix bugs found by valgrind when --enable-debug is used with the
269 new case code. These bugs are hidden when the data set is small
270 enough to find in memory; if a bigger data set that would overflow
271 to disk were used, then data corruption would occur.
273 * chisquare.c (create_freq_hash): Pass free_freq_mutable_hash to
274 hsh_create as free function. Make copy of data put into hash.
276 * oneway.q (free_value): New function.
277 (run_oneway): Use free_value as arg to hsh_create. Make copy of
280 * rank.q (rank_cases): Don't access data in case after we've given
283 Tue Jan 9 19:16:11 2007 Ben Pfaff <blp@gnu.org>
287 * aggregate.c (parse_aggregate_functions) Initialize
290 Fri Dec 22 14:04:09 2006 Ben Pfaff <blp@gnu.org>
292 Simplify missing value handling.
294 * aggregate.c (struct agr_var): Remove `bool include_missing', add
295 `enum mv_class exclude'. Remove `int missing', add `bool
296 saw_missing'. Update users.
298 * descriptives.c (struct dsc_trns): Removed `int
299 include_user_missing', add `enum mv_class exclude'. Update users.
300 (struct dsc_proc): Ditto.
302 * examine.q: (static var value_is_missing): Rename
303 `exclude_values', change type to `enum mv_class'. Update users.
307 Fri Dec 22 19:22:18 WST 2006 John Darrington <john@darrington.wattle.id.au>
309 * frequencies.q : Fixed bug #17420, where the table bounds were overun
310 when /FORMAT=nolabels was given.
312 Wed Dec 20 18:45:31 WST 2006 John Darrington <john@darrington.wattle.id.au>
314 * binomial.c binomial.h : New files. Thanks to Jason Stover
315 for assistance with these.
317 * chisquare.c chisquare.h freq.c freq.h npar-summary.c npar-summary.h
318 npar.h npar.q: New files. Implementing NPAR TESTS.
320 * frequencies.q : Moved structure definitions into freq.[ch]
322 Sat Dec 16 22:26:44 2006 Ben Pfaff <blp@gnu.org>
324 Make it possible to pull cases from the active file with a
325 function call, instead of requiring indirection through a callback
328 * aggregate.c (cmd_aggregate): Take advantage of new procedure
330 (agr_to_active_file): Removed.
331 (presorted_agr_to_sysfile): Removed.
333 * autorecode.c (cmd_autorecode): Take advantage of new procedure
335 (autorecode_proc_func): Removed.
337 * flip.c (struct flip_pgm): New members to allow conformance with
338 new case_source_class interface.
339 (cmd_flip): Adapt to new case_source_class interface.
340 (flip_source_read): Ditto.
341 (flip_source_destroy): Ditto.
343 Sat Dec 16 12:54:27 2006 Ben Pfaff <blp@gnu.org>
345 * rank.q (rank_custom_variables): Allow grouping variables to be
346 strings. Fixes bug #18533. Thanks to John Darrington for review.
348 Sat Dec 9 18:47:51 2006 Ben Pfaff <blp@gnu.org>
350 * regression.q (is_depvar): Compare variable pointers instead of
353 Thu Dec 7 15:26:25 WST 2006 John Darrington <john@darrington.wattle.id.au>
355 * examine.q: Allocated the categorical values for the dependent and
356 independent variables, on the heap. Hence they can be of any width.
358 Wed Dec 6 21:14:26 2006 Ben Pfaff <blp@gnu.org>
360 * regression.q (reg_inserted): Compare variable pointers instead
363 Mon Dec 4 22:33:46 2006 Ben Pfaff <blp@gnu.org>
365 * crosstabs.q (insert_summary): Use var_to_string for labeling.
366 (output_pivot_table) Ditto.
369 * frequencies.q (setup_z_trns): Ditto.
371 (dump_condensed) Ditto.
372 (dump_statistics) Ditto.
375 Sun Nov 5 08:31:42 WST 2006 John Darrington <john@darrington.wattle.id.au>
377 * t-test.q, oneway.q: Changed to use the new casefilter structure.
379 Sat Oct 14 16:52:28 2006 Ben Pfaff <blp@gnu.org>
381 * rank.q: (rank_sorted_casefile) Add some missing case_destroy()
382 calls to fix a memory leak.
384 Sun Oct 8 09:45:40 WST 2006 John Darrington <john@darrington.wattle.id.au>
386 * rank.q: Plugged a small memory leak which occurred under error
389 Sat Oct 7 11:06:01 WST 2006 John Darrington <john@darrington.wattle.id.au>
391 * rank.q: Implemented most of the RANK command.
393 2006-07-14 Jason Stover <jhs@math.gcsu.edu>
395 * regression.q (run_regression): New function to move knowledge of
396 pspp_linreg_cache out of math/coefficient.[ch].
398 Sat Jul 1 17:41:46 2006 Ben Pfaff <blp@gnu.org>
400 Fix bug #11612, "q2c documentation does not agree with code".
402 * examine.q: Audit use of q2c "+" prefixes that indicate that a
403 command may appear multiple times.
405 * frequencies.q: Ditto.
409 * regression.q: Ditto.
413 Fri Jun 23 14:18:22 2006 Ben Pfaff <blp@gnu.org>
415 Support long string variables on FREQUENCIES, as
416 an extension when in enhanced algorithms mode. For Greg Hunt
417 <greg@firmansyah.com>.
419 * frequencies.q: (struct freq) Change `v' member from union value
420 to union value *. Update all references.
421 (struct var_freqs) Add width, print members to represent effective
422 variable width and display format.
423 (calc) Copy entire long string value into the hash table.
424 (frq_custom_variables) Set new width, print members.
425 (hash_value_alpha) Get width from var_freqs.
426 (compare_value_alpha_a) Ditto.
427 (compare_freq_alpha_a) Ditto.
428 (compare_freq_alpha_d) Ditto.
429 (dump_full) Get display format from var_freqs.
430 (dump_condensed) Ditto.
432 Mon Jun 19 22:07:13 2006 Ben Pfaff <blp@gnu.org>
434 * frequencies.q: (dump_full) Only put the first MAX_SHORT_STRING
435 bytes of string variables into the output cells, seeing as we only
437 (dump_condensed) Ditto.
439 Mon Jun 19 21:52:05 2006 Ben Pfaff <blp@gnu.org>
441 Fixes a bug reported by Greg Hunt <greg@firmansyah.com>.
443 * frequencies.q: (hsh_hash_bytes) We only copy the first
444 MAX_SHORT_STRING bytes of string variables, so we must only
445 compare that many bytes, even if the string variable is longer.
446 (compare_value_alpha_a) Ditto.
447 (compare_freq_alpha_a) Ditto.
448 (compare_freq_alpha_d) Ditto.
450 2006-05-11 Jason Stover <jhs@math.gcsu.edu>
452 * regression.q: Adjusted code to account for cache->coeff being a
453 pspp_linreg_coeff **.
455 Sun May 7 18:31:25 2006 Ben Pfaff <blp@gnu.org>
459 * aggregate.c (cmd_aggregate): Free default_dict before replacing
462 Sun May 7 17:09:19 2006 Ben Pfaff <blp@gnu.org>
464 * flip.c (flip_file): Check return value of pool_fclose().
466 Sat May 6 16:00:13 2006 Ben Pfaff <blp@gnu.org>
468 Get rid of `char *c' member in union value, for cleanliness.
470 * aggregate.c: (union agr_argument) New union.
471 (struct agr_var) Change element type of arg[] from union value to
473 (parse_aggregate_functions) Change local variable types likewise.
475 * autorecode.c: (union arc_value) New union.
476 (struct arc_item) Change "from" from union value to union
478 (recode) Change local variable from union value to union
480 (autorecode_trns_proc) Ditto.
481 (compare_alpha_value) Ditto.
482 (hash_alpha_value) Ditto.
483 (compare_numeric_value) Ditto.
484 (hash_numeric_value) Ditto.
485 (autorecode_proc_func) Ditto.
487 Sat May 6 10:43:33 2006 Ben Pfaff <blp@gnu.org>
489 Continue reforming procedure execution. In this phase, get rid of
490 the output code for SPLIT FILE groups in procedure.c, which really
491 shouldn't be doing any output. Move it into the individual
492 procedures instead. This also adds some flexibility.
494 * crosstabs.q (precalc): Call output_split_file_values().
496 * descriptives.c (calc_descriptives): Ditto.
498 * examine.q (run_examine): Ditto.
500 * frequencies.q (precalc): Ditto.
502 * oneway.q (run_oneway): Ditto.
504 * regression.q (run_regression): Ditto.
506 * t-test.q (calculate): Ditto.
508 Wed May 3 23:05:31 2006 Ben Pfaff <blp@gnu.org>
510 Continue reforming procedure execution. In this phase, get rid of
511 many global variables, consolidating procedure execution in
512 procedure.c. Encapsulate transformations in new "struct
513 trns_chain". Also, change implementation of N OF CASES, FILTER,
514 and PROCESS IF from special cases to transformations.
516 * aggregate.c (cmd_aggregate) Use discard_variables().
518 2006-04-28 Jason Stover <jhs@math.gcsu.edu>
520 * regression.q (regression_trns_resid_proc): Pass only the
521 variables used in the model to (*model->residual)().
523 * regression.q (regression_trns_pred_proc): Pass only the
524 variables used in the model to (*model->pred)().
526 2006-04-26 Jason Stover <jhs@math.gcsu.edu>
528 * regression.q: Added support for multiple transformations.
530 * regression.q (regression_trns_resid_proc): New function.
532 * regression.q (regression_trns_pred_proc): New function.
534 * regression.q (subcommand_save): Added support for saving
537 * regression.q (regression_trns_free): New function.
539 * regression.q (reg_get_name): New function.
541 * regression.q (reg_save_var): New function.
543 Tue Apr 25 13:18:56 2006 Ben Pfaff <blp@gnu.org>
545 * rank.q (parse_rank_function): Use SE instead of ME for parse
548 Tue Apr 25 13:16:28 2006 Ben Pfaff <blp@gnu.org>
550 * flip.c (flip_sink_write): Use snprintf() to simplify a bit of
553 2006-04-21 Jason Stover <jhs@math.gcsu.edu>
555 * regression.q (try_name): New function. (Partly copied from
556 try_name in descriptives.c.)
558 * regression.q (subcommand_save): Choose residual variable names
561 2006-04-20 Jason Stover <jhs@math.gcsu.edu>
563 * regression.q (cmd_regression): Moved call to subcommand_save()
564 outside multipass_procedure_with_splits().
566 * regression.q (regression_trns_proc): Fixed value counter n_vals
567 before calling *model->residual().
569 2006-04-19 Jason Stover <jhs@math.gcsu.edu>
571 * regression.q (regression_trns_proc): Fixed the look-up of the
574 2006-04-18 Jason Stover <jhs@math.gcsu.edu>
576 * regression.q (regression_trns_proc): Look up the residual
577 variable in the linear regression cache.
579 * regression.q (subcommand_save): Set the residual variable in the
580 linear regression cache.
582 2006-04-17 Jason Stover <jason@wonko.gcsu.edu>
584 * regression.q (regression_trns_proc): Accept case_idx as an int
585 to match the definition of trns_proc_func.
587 2006-04-17 Jason Stover <jhs@math.gcsu.edu>
589 * regression.q (regression_trns_proc): New function.
591 * regression.q (subcommand_save): Create variable residuals and
592 add a transformation to assign values to them. Also free the
593 linreg_cache if the SAVE command was not called. Removed the
594 casereading loop. Placed actual computation of residuals in
595 regression_trns_proc.
597 * regression.q (run_regression): Moved call to free
598 pspp_linreg_cache to subcommand_save.
600 Sat Apr 15 18:01:03 2006 Ben Pfaff <blp@gnu.org>
602 * examine.q (output_examine): Add casts to fix warnings.
604 2006-04-07 Jason Stover <jhs@math.gcsu.edu>
606 * regression.q (subcommand_save): New function.
608 2006-04-04 Jason Stover <jhs@math.gcsu.edu>
610 * regression.q: New function reg_has_categorical () to tell
611 whether a pspp_linreg_struct was made with any variables of type
614 * regression.q: (subcommand_export): Call
615 reg_print_categorical_encoding() only if the model uses any
616 categorical variables.
618 Mon Mar 27 16:00:42 2006 Ben Pfaff <blp@gnu.org>
620 * crosstabs.q: (output_pivot_table) Drop spurious space from
623 2006-03-15 Jason Stover <jhs@math.gcsu.edu>
625 * regression.q: Added custom syntax parser for VARIABLES subcommand
627 * regression.q: Moved most instructions for run_regression ()
628 inside the loop over dependent variables.
630 Thu Mar 2 08:40:33 WST 2006 John Darrington <john@darrington.wattle.id.au>
632 * Moved files from src directory