1 2007-10-12 Ben Pfaff <blp@gnu.org>
3 * flip.c (flip_file): No need to conditionally substitute for
4 "fseeko" and "off_t" manually anymore, as gnulib takes care of it
7 2007-09-21 Jason Stover <jhs@wonko.gcsu.edu>
9 * regression.q (run_regression): Partial fix of memory leak, bug
12 2007-09-19 Ben Pfaff <blp@gnu.org>
16 * aggregate.c (cmd_aggregate): Destroy casereader consistently,
17 even if casereader fails.
19 * examine.q (run_examine): Ditto.
21 * glm.q (run_glm): Ditto.
23 * oneway.q (run_oneway): Ditto.
25 * regression.q (run_regression): Ditto.
27 * t-test.q (calculate): Ditto.
29 * descriptives.c (calc_descriptives): Ditto. Also avoid
30 gratuitous casereader_clone.
32 2007-09-13 Jason Stover <jhs@debs.hoobahooba.net>
34 * regression.q (cmd_regression): Move declaration of models in to
35 definition of cmd_regression.
37 * regression.q (run_regression): Free mom to fix memory leak.
39 2007-09-12 Ben Pfaff <blp@gnu.org>
41 * crosstabs.q (postcalc): Free sorted_tab and the structures that
42 it points to, to plug a memory leak. Fixes bug #20910. Thanks to
43 John Darrington for reporting this bug and for review.
45 2007-09-04 Ben Pfaff <blp@gnu.org>
47 * crosstabs.q (cmd_crosstabs): Free xtab and the structures that
48 it points to, to plug a memory leak. Fixes bug #18315.
50 2007-08-15 Jason Stover <jhs@wonko.gcsu.edu>
52 * regression.q (identify_indep_vars): Print an error if dependent
53 and independent variables are the same. Fixes bug 19819.
55 2007-08-12 Ben Pfaff <blp@gnu.org>
57 * flip.c: Drop use of dict_get_compacted_dict_index_to_case_index
58 and just use the ordinary case indexes. There seemed to be no
59 reason for the former method.
61 2007-08-03 Ben Pfaff <blp@gnu.org>
63 * rank.q (rank_cmd): Instead of sorting by SPLIT FILE vars, group
64 by them. Fixes bug #17239.
65 Reviewed by John Darrington.
67 2007-08-01 Ben Pfaff <blp@gnu.org>
69 Clean up handling of median, by treating it almost like any other
70 percentile. Fixes bug #17424. Thanks to John Darrington for
72 * frequencies.q (internal_cmd_frequencies): Fix handling of bit
73 masks for `stats' variable. If median is selected, turn it off
74 and add a 50th percentile.
75 (add_percentile): Simplify code a little.
76 (calc_stats): Drop special casing of median.
77 (dump_statistics): Ditto, except that we label the 50th percentile
78 as "50 (Median)" to make it clear that it's also the median.
80 2007-07-31 Ben Pfaff <blp@gnu.org>
82 Remove integer mode from FREQUENCIES and incidentally fix bug
83 #17421. Reviewed by John Darrington.
84 * frequencies.q (int_pool): Rename data_pool.
85 (gen_pool): Rename syntax_pool.
86 (enum FRQM_*): Removed.
87 (struct freq_tab): Removed `mode', `vector', `min', `max',
88 `out_of_range', `sysmis' members.
89 (calc): Delete support for integer mode.
91 (postprocess_freq_tab): Ditto.
92 (cleanup_freq_tab): Ditto.
93 (frq_custom_variables): Ditto.
95 2007-07-28 John Darrington <john@darrington.wattle.id.au>
97 * t-test.q: Moved the order in which groups are displayed in the
98 independent samples case, where a cut point is given.
100 2007-07-27 Ben Pfaff <blp@gnu.org>
102 * regression.q (run_regression): Move casereader_destroy call so
103 that it always gets called, not just if there was some valid
104 data. Fixes bug #19581.
105 Reviewed by Jason Stover.
107 2007-07-24 Ben Pfaff <blp@gnu.org>
109 * flip.c (struct flip_pgm): Remove `case_size' member (now
111 (cmd_flip): Pass var_cnt as number of cases instead of case_cnt,
112 to fix bug #20494. Don't assign to `case_size' member, which was
113 unused after assignment.
114 (build_dictionary): When NEWNAMES not used, get the number of
115 variables right, to fix bug #20493.
117 2007-07-10 Jason Stover <jhs@math.gcsu.edu>
119 * glm.q: Initial version of the GLM procedure.
121 2007-06-06 Ben Pfaff <blp@gnu.org>
123 Adapt case sources, sinks, and clients of procedure code to the
126 * aggregate.c: Simplify greatly since everything is more uniform
129 * autorecode.c: Adapt to new procedure code.
131 * chisquare.c: Ditto.
132 * crosstabs.q: Ditto.
133 * descriptives.c: Ditto.
135 * npar-summary.c: Ditto.
136 * frequencies.q: Ditto.
139 * regression.q: Ditto.
140 * sort-cases.c: Ditto.
143 * sort-criteria.c: Rewrite to output a struct case_ordering.
145 * flip.c: Rewrite to be a casereader.
147 * rank.q: Simplify greatly since casereaders are much more
148 flexible than what we had before.
150 2007-05-15 Jason Stover <jhs@math.gcsu.edu>
152 * regression.q (run_regression): Tell the user when the data
153 contain no valid cases.
155 2007-05-08 Jason Stover <jhs@math.gcsu.edu>
157 * regression.q: Partial fix of bug which caused a crash if
158 dependent variable and independent variable were the same.
160 2007-04-16 John Darrington <john@darrington.wattle.id.au>
162 * t-test.q: Changed the output width of reported counts and
163 degrees of freedom, to avoid truncating these values. Thanks
164 to Seth Woolley for reporting this problem. A proper fix involves
165 re-thinking the output driver.
167 2007-04-12 Jason Stover <jhs@math.gcsu.edu>
169 * regression.q (run_regression): Added if (n_data >0) to fix bug
172 2007-03-29 Jason Stover <jhs@math.gcsu.edu>
174 * regression.q (prepare_data): New function.
176 * regression.q (compute_moments): New function.
178 2007-03-18 Ben Pfaff <blp@gnu.org>
180 * crosstabs.q (static var write): Rename write_style to avoid
181 conflict with POSIX function of same name.
183 2007-03-16 Jason Stover <jhs@math.gcsu.edu>
185 * regression.q (run_regression): Added support for moments.
187 Sat Feb 17 08:16:00 2007 Ben Pfaff <blp@gnu.org>
189 * flip.c (flip_sink_create): Improve error message when temporary
190 file cannot be created.
192 Tue Feb 6 19:58:03 2007 Ben Pfaff <blp@gnu.org>
194 * flip.c (flip_file): Give better error message on end-of-file.
196 2007-02-04 Jason Stover <jhs@math.gcsu.edu>
198 * regression.q: Fixed p-value computation in the test for
199 individual regression coefficients.
201 Mon Jan 15 11:03:20 2007 Ben Pfaff <blp@gnu.org>
203 Fix bugs found by valgrind when --enable-debug is used with the
204 new case code. These bugs are hidden when the data set is small
205 enough to find in memory; if a bigger data set that would overflow
206 to disk were used, then data corruption would occur.
208 * chisquare.c (create_freq_hash): Pass free_freq_mutable_hash to
209 hsh_create as free function. Make copy of data put into hash.
211 * oneway.q (free_value): New function.
212 (run_oneway): Use free_value as arg to hsh_create. Make copy of
215 * rank.q (rank_cases): Don't access data in case after we've given
218 Tue Jan 9 19:16:11 2007 Ben Pfaff <blp@gnu.org>
222 * aggregate.c (parse_aggregate_functions) Initialize
225 Fri Dec 22 14:04:09 2006 Ben Pfaff <blp@gnu.org>
227 Simplify missing value handling.
229 * aggregate.c (struct agr_var): Remove `bool include_missing', add
230 `enum mv_class exclude'. Remove `int missing', add `bool
231 saw_missing'. Update users.
233 * descriptives.c (struct dsc_trns): Removed `int
234 include_user_missing', add `enum mv_class exclude'. Update users.
235 (struct dsc_proc): Ditto.
237 * examine.q: (static var value_is_missing): Rename
238 `exclude_values', change type to `enum mv_class'. Update users.
242 Fri Dec 22 19:22:18 WST 2006 John Darrington <john@darrington.wattle.id.au>
244 * frequencies.q : Fixed bug #17420, where the table bounds were overun
245 when /FORMAT=nolabels was given.
247 Wed Dec 20 18:45:31 WST 2006 John Darrington <john@darrington.wattle.id.au>
249 * binomial.c binomial.h : New files. Thanks to Jason Stover
250 for assistance with these.
252 * chisquare.c chisquare.h freq.c freq.h npar-summary.c npar-summary.h
253 npar.h npar.q: New files. Implementing NPAR TESTS.
255 * frequencies.q : Moved structure definitions into freq.[ch]
257 Sat Dec 16 22:26:44 2006 Ben Pfaff <blp@gnu.org>
259 Make it possible to pull cases from the active file with a
260 function call, instead of requiring indirection through a callback
263 * aggregate.c (cmd_aggregate): Take advantage of new procedure
265 (agr_to_active_file): Removed.
266 (presorted_agr_to_sysfile): Removed.
268 * autorecode.c (cmd_autorecode): Take advantage of new procedure
270 (autorecode_proc_func): Removed.
272 * flip.c (struct flip_pgm): New members to allow conformance with
273 new case_source_class interface.
274 (cmd_flip): Adapt to new case_source_class interface.
275 (flip_source_read): Ditto.
276 (flip_source_destroy): Ditto.
278 Sat Dec 16 12:54:27 2006 Ben Pfaff <blp@gnu.org>
280 * rank.q (rank_custom_variables): Allow grouping variables to be
281 strings. Fixes bug #18533. Thanks to John Darrington for review.
283 Sat Dec 9 18:47:51 2006 Ben Pfaff <blp@gnu.org>
285 * regression.q (is_depvar): Compare variable pointers instead of
288 Thu Dec 7 15:26:25 WST 2006 John Darrington <john@darrington.wattle.id.au>
290 * examine.q: Allocated the categorical values for the dependent and
291 independent variables, on the heap. Hence they can be of any width.
293 Wed Dec 6 21:14:26 2006 Ben Pfaff <blp@gnu.org>
295 * regression.q (reg_inserted): Compare variable pointers instead
298 Mon Dec 4 22:33:46 2006 Ben Pfaff <blp@gnu.org>
300 * crosstabs.q (insert_summary): Use var_to_string for labeling.
301 (output_pivot_table) Ditto.
304 * frequencies.q (setup_z_trns): Ditto.
306 (dump_condensed) Ditto.
307 (dump_statistics) Ditto.
310 Sun Nov 5 08:31:42 WST 2006 John Darrington <john@darrington.wattle.id.au>
312 * t-test.q, oneway.q: Changed to use the new casefilter structure.
314 Sat Oct 14 16:52:28 2006 Ben Pfaff <blp@gnu.org>
316 * rank.q: (rank_sorted_casefile) Add some missing case_destroy()
317 calls to fix a memory leak.
319 Sun Oct 8 09:45:40 WST 2006 John Darrington <john@darrington.wattle.id.au>
321 * rank.q: Plugged a small memory leak which occurred under error
324 Sat Oct 7 11:06:01 WST 2006 John Darrington <john@darrington.wattle.id.au>
326 * rank.q: Implemented most of the RANK command.
328 2006-07-14 Jason Stover <jhs@math.gcsu.edu>
330 * regression.q (run_regression): New function to move knowledge of
331 pspp_linreg_cache out of math/coefficient.[ch].
333 Sat Jul 1 17:41:46 2006 Ben Pfaff <blp@gnu.org>
335 Fix bug #11612, "q2c documentation does not agree with code".
337 * examine.q: Audit use of q2c "+" prefixes that indicate that a
338 command may appear multiple times.
340 * frequencies.q: Ditto.
344 * regression.q: Ditto.
348 Fri Jun 23 14:18:22 2006 Ben Pfaff <blp@gnu.org>
350 Support long string variables on FREQUENCIES, as
351 an extension when in enhanced algorithms mode. For Greg Hunt
352 <greg@firmansyah.com>.
354 * frequencies.q: (struct freq) Change `v' member from union value
355 to union value *. Update all references.
356 (struct var_freqs) Add width, print members to represent effective
357 variable width and display format.
358 (calc) Copy entire long string value into the hash table.
359 (frq_custom_variables) Set new width, print members.
360 (hash_value_alpha) Get width from var_freqs.
361 (compare_value_alpha_a) Ditto.
362 (compare_freq_alpha_a) Ditto.
363 (compare_freq_alpha_d) Ditto.
364 (dump_full) Get display format from var_freqs.
365 (dump_condensed) Ditto.
367 Mon Jun 19 22:07:13 2006 Ben Pfaff <blp@gnu.org>
369 * frequencies.q: (dump_full) Only put the first MAX_SHORT_STRING
370 bytes of string variables into the output cells, seeing as we only
372 (dump_condensed) Ditto.
374 Mon Jun 19 21:52:05 2006 Ben Pfaff <blp@gnu.org>
376 Fixes a bug reported by Greg Hunt <greg@firmansyah.com>.
378 * frequencies.q: (hsh_hash_bytes) We only copy the first
379 MAX_SHORT_STRING bytes of string variables, so we must only
380 compare that many bytes, even if the string variable is longer.
381 (compare_value_alpha_a) Ditto.
382 (compare_freq_alpha_a) Ditto.
383 (compare_freq_alpha_d) Ditto.
385 2006-05-11 Jason Stover <jhs@math.gcsu.edu>
387 * regression.q: Adjusted code to account for cache->coeff being a
388 pspp_linreg_coeff **.
390 Sun May 7 18:31:25 2006 Ben Pfaff <blp@gnu.org>
394 * aggregate.c (cmd_aggregate): Free default_dict before replacing
397 Sun May 7 17:09:19 2006 Ben Pfaff <blp@gnu.org>
399 * flip.c (flip_file): Check return value of pool_fclose().
401 Sat May 6 16:00:13 2006 Ben Pfaff <blp@gnu.org>
403 Get rid of `char *c' member in union value, for cleanliness.
405 * aggregate.c: (union agr_argument) New union.
406 (struct agr_var) Change element type of arg[] from union value to
408 (parse_aggregate_functions) Change local variable types likewise.
410 * autorecode.c: (union arc_value) New union.
411 (struct arc_item) Change "from" from union value to union
413 (recode) Change local variable from union value to union
415 (autorecode_trns_proc) Ditto.
416 (compare_alpha_value) Ditto.
417 (hash_alpha_value) Ditto.
418 (compare_numeric_value) Ditto.
419 (hash_numeric_value) Ditto.
420 (autorecode_proc_func) Ditto.
422 Sat May 6 10:43:33 2006 Ben Pfaff <blp@gnu.org>
424 Continue reforming procedure execution. In this phase, get rid of
425 the output code for SPLIT FILE groups in procedure.c, which really
426 shouldn't be doing any output. Move it into the individual
427 procedures instead. This also adds some flexibility.
429 * crosstabs.q (precalc): Call output_split_file_values().
431 * descriptives.c (calc_descriptives): Ditto.
433 * examine.q (run_examine): Ditto.
435 * frequencies.q (precalc): Ditto.
437 * oneway.q (run_oneway): Ditto.
439 * regression.q (run_regression): Ditto.
441 * t-test.q (calculate): Ditto.
443 Wed May 3 23:05:31 2006 Ben Pfaff <blp@gnu.org>
445 Continue reforming procedure execution. In this phase, get rid of
446 many global variables, consolidating procedure execution in
447 procedure.c. Encapsulate transformations in new "struct
448 trns_chain". Also, change implementation of N OF CASES, FILTER,
449 and PROCESS IF from special cases to transformations.
451 * aggregate.c (cmd_aggregate) Use discard_variables().
453 2006-04-28 Jason Stover <jhs@math.gcsu.edu>
455 * regression.q (regression_trns_resid_proc): Pass only the
456 variables used in the model to (*model->residual)().
458 * regression.q (regression_trns_pred_proc): Pass only the
459 variables used in the model to (*model->pred)().
461 2006-04-26 Jason Stover <jhs@math.gcsu.edu>
463 * regression.q: Added support for multiple transformations.
465 * regression.q (regression_trns_resid_proc): New function.
467 * regression.q (regression_trns_pred_proc): New function.
469 * regression.q (subcommand_save): Added support for saving
472 * regression.q (regression_trns_free): New function.
474 * regression.q (reg_get_name): New function.
476 * regression.q (reg_save_var): New function.
478 Tue Apr 25 13:18:56 2006 Ben Pfaff <blp@gnu.org>
480 * rank.q (parse_rank_function): Use SE instead of ME for parse
483 Tue Apr 25 13:16:28 2006 Ben Pfaff <blp@gnu.org>
485 * flip.c (flip_sink_write): Use snprintf() to simplify a bit of
488 2006-04-21 Jason Stover <jhs@math.gcsu.edu>
490 * regression.q (try_name): New function. (Partly copied from
491 try_name in descriptives.c.)
493 * regression.q (subcommand_save): Choose residual variable names
496 2006-04-20 Jason Stover <jhs@math.gcsu.edu>
498 * regression.q (cmd_regression): Moved call to subcommand_save()
499 outside multipass_procedure_with_splits().
501 * regression.q (regression_trns_proc): Fixed value counter n_vals
502 before calling *model->residual().
504 2006-04-19 Jason Stover <jhs@math.gcsu.edu>
506 * regression.q (regression_trns_proc): Fixed the look-up of the
509 2006-04-18 Jason Stover <jhs@math.gcsu.edu>
511 * regression.q (regression_trns_proc): Look up the residual
512 variable in the linear regression cache.
514 * regression.q (subcommand_save): Set the residual variable in the
515 linear regression cache.
517 2006-04-17 Jason Stover <jason@wonko.gcsu.edu>
519 * regression.q (regression_trns_proc): Accept case_idx as an int
520 to match the definition of trns_proc_func.
522 2006-04-17 Jason Stover <jhs@math.gcsu.edu>
524 * regression.q (regression_trns_proc): New function.
526 * regression.q (subcommand_save): Create variable residuals and
527 add a transformation to assign values to them. Also free the
528 linreg_cache if the SAVE command was not called. Removed the
529 casereading loop. Placed actual computation of residuals in
530 regression_trns_proc.
532 * regression.q (run_regression): Moved call to free
533 pspp_linreg_cache to subcommand_save.
535 Sat Apr 15 18:01:03 2006 Ben Pfaff <blp@gnu.org>
537 * examine.q (output_examine): Add casts to fix warnings.
539 2006-04-07 Jason Stover <jhs@math.gcsu.edu>
541 * regression.q (subcommand_save): New function.
543 2006-04-04 Jason Stover <jhs@math.gcsu.edu>
545 * regression.q: New function reg_has_categorical () to tell
546 whether a pspp_linreg_struct was made with any variables of type
549 * regression.q: (subcommand_export): Call
550 reg_print_categorical_encoding() only if the model uses any
551 categorical variables.
553 Mon Mar 27 16:00:42 2006 Ben Pfaff <blp@gnu.org>
555 * crosstabs.q: (output_pivot_table) Drop spurious space from
558 2006-03-15 Jason Stover <jhs@math.gcsu.edu>
560 * regression.q: Added custom syntax parser for VARIABLES subcommand
562 * regression.q: Moved most instructions for run_regression ()
563 inside the loop over dependent variables.
565 Thu Mar 2 08:40:33 WST 2006 John Darrington <john@darrington.wattle.id.au>
567 * Moved files from src directory