1 2007-09-04 Ben Pfaff <blp@gnu.org>
3 * crosstabs.q (cmd_crosstabs): Free xtab and the structures that
4 it points to, to plug a memory leak. Fixes bug #18315.
6 2007-08-15 Jason Stover <jhs@wonko.gcsu.edu>
8 * regression.q (identify_indep_vars): Print an error if dependent
9 and independent variables are the same. Fixes bug 19819.
11 2007-08-12 Ben Pfaff <blp@gnu.org>
13 * flip.c: Drop use of dict_get_compacted_dict_index_to_case_index
14 and just use the ordinary case indexes. There seemed to be no
15 reason for the former method.
17 2007-08-03 Ben Pfaff <blp@gnu.org>
19 * rank.q (rank_cmd): Instead of sorting by SPLIT FILE vars, group
20 by them. Fixes bug #17239.
21 Reviewed by John Darrington.
23 2007-08-01 Ben Pfaff <blp@gnu.org>
25 Clean up handling of median, by treating it almost like any other
26 percentile. Fixes bug #17424. Thanks to John Darrington for
28 * frequencies.q (internal_cmd_frequencies): Fix handling of bit
29 masks for `stats' variable. If median is selected, turn it off
30 and add a 50th percentile.
31 (add_percentile): Simplify code a little.
32 (calc_stats): Drop special casing of median.
33 (dump_statistics): Ditto, except that we label the 50th percentile
34 as "50 (Median)" to make it clear that it's also the median.
36 2007-07-31 Ben Pfaff <blp@gnu.org>
38 Remove integer mode from FREQUENCIES and incidentally fix bug
39 #17421. Reviewed by John Darrington.
40 * frequencies.q (int_pool): Rename data_pool.
41 (gen_pool): Rename syntax_pool.
42 (enum FRQM_*): Removed.
43 (struct freq_tab): Removed `mode', `vector', `min', `max',
44 `out_of_range', `sysmis' members.
45 (calc): Delete support for integer mode.
47 (postprocess_freq_tab): Ditto.
48 (cleanup_freq_tab): Ditto.
49 (frq_custom_variables): Ditto.
51 2007-07-28 John Darrington <john@darrington.wattle.id.au>
53 * t-test.q: Moved the order in which groups are displayed in the
54 independent samples case, where a cut point is given.
56 2007-07-27 Ben Pfaff <blp@gnu.org>
58 * regression.q (run_regression): Move casereader_destroy call so
59 that it always gets called, not just if there was some valid
60 data. Fixes bug #19581.
61 Reviewed by Jason Stover.
63 2007-07-24 Ben Pfaff <blp@gnu.org>
65 * flip.c (struct flip_pgm): Remove `case_size' member (now
67 (cmd_flip): Pass var_cnt as number of cases instead of case_cnt,
68 to fix bug #20494. Don't assign to `case_size' member, which was
69 unused after assignment.
70 (build_dictionary): When NEWNAMES not used, get the number of
71 variables right, to fix bug #20493.
73 2007-07-10 Jason Stover <jhs@math.gcsu.edu>
75 * glm.q: Initial version of the GLM procedure.
77 2007-06-06 Ben Pfaff <blp@gnu.org>
79 Adapt case sources, sinks, and clients of procedure code to the
82 * aggregate.c: Simplify greatly since everything is more uniform
85 * autorecode.c: Adapt to new procedure code.
89 * descriptives.c: Ditto.
91 * npar-summary.c: Ditto.
92 * frequencies.q: Ditto.
95 * regression.q: Ditto.
96 * sort-cases.c: Ditto.
99 * sort-criteria.c: Rewrite to output a struct case_ordering.
101 * flip.c: Rewrite to be a casereader.
103 * rank.q: Simplify greatly since casereaders are much more
104 flexible than what we had before.
106 2007-05-15 Jason Stover <jhs@math.gcsu.edu>
108 * regression.q (run_regression): Tell the user when the data
109 contain no valid cases.
111 2007-05-08 Jason Stover <jhs@math.gcsu.edu>
113 * regression.q: Partial fix of bug which caused a crash if
114 dependent variable and independent variable were the same.
116 2007-04-16 John Darrington <john@darrington.wattle.id.au>
118 * t-test.q: Changed the output width of reported counts and
119 degrees of freedom, to avoid truncating these values. Thanks
120 to Seth Woolley for reporting this problem. A proper fix involves
121 re-thinking the output driver.
123 2007-04-12 Jason Stover <jhs@math.gcsu.edu>
125 * regression.q (run_regression): Added if (n_data >0) to fix bug
128 2007-03-29 Jason Stover <jhs@math.gcsu.edu>
130 * regression.q (prepare_data): New function.
132 * regression.q (compute_moments): New function.
134 2007-03-18 Ben Pfaff <blp@gnu.org>
136 * crosstabs.q (static var write): Rename write_style to avoid
137 conflict with POSIX function of same name.
139 2007-03-16 Jason Stover <jhs@math.gcsu.edu>
141 * regression.q (run_regression): Added support for moments.
143 Sat Feb 17 08:16:00 2007 Ben Pfaff <blp@gnu.org>
145 * flip.c (flip_sink_create): Improve error message when temporary
146 file cannot be created.
148 Tue Feb 6 19:58:03 2007 Ben Pfaff <blp@gnu.org>
150 * flip.c (flip_file): Give better error message on end-of-file.
152 2007-02-04 Jason Stover <jhs@math.gcsu.edu>
154 * regression.q: Fixed p-value computation in the test for
155 individual regression coefficients.
157 Mon Jan 15 11:03:20 2007 Ben Pfaff <blp@gnu.org>
159 Fix bugs found by valgrind when --enable-debug is used with the
160 new case code. These bugs are hidden when the data set is small
161 enough to find in memory; if a bigger data set that would overflow
162 to disk were used, then data corruption would occur.
164 * chisquare.c (create_freq_hash): Pass free_freq_mutable_hash to
165 hsh_create as free function. Make copy of data put into hash.
167 * oneway.q (free_value): New function.
168 (run_oneway): Use free_value as arg to hsh_create. Make copy of
171 * rank.q (rank_cases): Don't access data in case after we've given
174 Tue Jan 9 19:16:11 2007 Ben Pfaff <blp@gnu.org>
178 * aggregate.c (parse_aggregate_functions) Initialize
181 Fri Dec 22 14:04:09 2006 Ben Pfaff <blp@gnu.org>
183 Simplify missing value handling.
185 * aggregate.c (struct agr_var): Remove `bool include_missing', add
186 `enum mv_class exclude'. Remove `int missing', add `bool
187 saw_missing'. Update users.
189 * descriptives.c (struct dsc_trns): Removed `int
190 include_user_missing', add `enum mv_class exclude'. Update users.
191 (struct dsc_proc): Ditto.
193 * examine.q: (static var value_is_missing): Rename
194 `exclude_values', change type to `enum mv_class'. Update users.
198 Fri Dec 22 19:22:18 WST 2006 John Darrington <john@darrington.wattle.id.au>
200 * frequencies.q : Fixed bug #17420, where the table bounds were overun
201 when /FORMAT=nolabels was given.
203 Wed Dec 20 18:45:31 WST 2006 John Darrington <john@darrington.wattle.id.au>
205 * binomial.c binomial.h : New files. Thanks to Jason Stover
206 for assistance with these.
208 * chisquare.c chisquare.h freq.c freq.h npar-summary.c npar-summary.h
209 npar.h npar.q: New files. Implementing NPAR TESTS.
211 * frequencies.q : Moved structure definitions into freq.[ch]
213 Sat Dec 16 22:26:44 2006 Ben Pfaff <blp@gnu.org>
215 Make it possible to pull cases from the active file with a
216 function call, instead of requiring indirection through a callback
219 * aggregate.c (cmd_aggregate): Take advantage of new procedure
221 (agr_to_active_file): Removed.
222 (presorted_agr_to_sysfile): Removed.
224 * autorecode.c (cmd_autorecode): Take advantage of new procedure
226 (autorecode_proc_func): Removed.
228 * flip.c (struct flip_pgm): New members to allow conformance with
229 new case_source_class interface.
230 (cmd_flip): Adapt to new case_source_class interface.
231 (flip_source_read): Ditto.
232 (flip_source_destroy): Ditto.
234 Sat Dec 16 12:54:27 2006 Ben Pfaff <blp@gnu.org>
236 * rank.q (rank_custom_variables): Allow grouping variables to be
237 strings. Fixes bug #18533. Thanks to John Darrington for review.
239 Sat Dec 9 18:47:51 2006 Ben Pfaff <blp@gnu.org>
241 * regression.q (is_depvar): Compare variable pointers instead of
244 Thu Dec 7 15:26:25 WST 2006 John Darrington <john@darrington.wattle.id.au>
246 * examine.q: Allocated the categorical values for the dependent and
247 independent variables, on the heap. Hence they can be of any width.
249 Wed Dec 6 21:14:26 2006 Ben Pfaff <blp@gnu.org>
251 * regression.q (reg_inserted): Compare variable pointers instead
254 Mon Dec 4 22:33:46 2006 Ben Pfaff <blp@gnu.org>
256 * crosstabs.q (insert_summary): Use var_to_string for labeling.
257 (output_pivot_table) Ditto.
260 * frequencies.q (setup_z_trns): Ditto.
262 (dump_condensed) Ditto.
263 (dump_statistics) Ditto.
266 Sun Nov 5 08:31:42 WST 2006 John Darrington <john@darrington.wattle.id.au>
268 * t-test.q, oneway.q: Changed to use the new casefilter structure.
270 Sat Oct 14 16:52:28 2006 Ben Pfaff <blp@gnu.org>
272 * rank.q: (rank_sorted_casefile) Add some missing case_destroy()
273 calls to fix a memory leak.
275 Sun Oct 8 09:45:40 WST 2006 John Darrington <john@darrington.wattle.id.au>
277 * rank.q: Plugged a small memory leak which occurred under error
280 Sat Oct 7 11:06:01 WST 2006 John Darrington <john@darrington.wattle.id.au>
282 * rank.q: Implemented most of the RANK command.
284 2006-07-14 Jason Stover <jhs@math.gcsu.edu>
286 * regression.q (run_regression): New function to move knowledge of
287 pspp_linreg_cache out of math/coefficient.[ch].
289 Sat Jul 1 17:41:46 2006 Ben Pfaff <blp@gnu.org>
291 Fix bug #11612, "q2c documentation does not agree with code".
293 * examine.q: Audit use of q2c "+" prefixes that indicate that a
294 command may appear multiple times.
296 * frequencies.q: Ditto.
300 * regression.q: Ditto.
304 Fri Jun 23 14:18:22 2006 Ben Pfaff <blp@gnu.org>
306 Support long string variables on FREQUENCIES, as
307 an extension when in enhanced algorithms mode. For Greg Hunt
308 <greg@firmansyah.com>.
310 * frequencies.q: (struct freq) Change `v' member from union value
311 to union value *. Update all references.
312 (struct var_freqs) Add width, print members to represent effective
313 variable width and display format.
314 (calc) Copy entire long string value into the hash table.
315 (frq_custom_variables) Set new width, print members.
316 (hash_value_alpha) Get width from var_freqs.
317 (compare_value_alpha_a) Ditto.
318 (compare_freq_alpha_a) Ditto.
319 (compare_freq_alpha_d) Ditto.
320 (dump_full) Get display format from var_freqs.
321 (dump_condensed) Ditto.
323 Mon Jun 19 22:07:13 2006 Ben Pfaff <blp@gnu.org>
325 * frequencies.q: (dump_full) Only put the first MAX_SHORT_STRING
326 bytes of string variables into the output cells, seeing as we only
328 (dump_condensed) Ditto.
330 Mon Jun 19 21:52:05 2006 Ben Pfaff <blp@gnu.org>
332 Fixes a bug reported by Greg Hunt <greg@firmansyah.com>.
334 * frequencies.q: (hsh_hash_bytes) We only copy the first
335 MAX_SHORT_STRING bytes of string variables, so we must only
336 compare that many bytes, even if the string variable is longer.
337 (compare_value_alpha_a) Ditto.
338 (compare_freq_alpha_a) Ditto.
339 (compare_freq_alpha_d) Ditto.
341 2006-05-11 Jason Stover <jhs@math.gcsu.edu>
343 * regression.q: Adjusted code to account for cache->coeff being a
344 pspp_linreg_coeff **.
346 Sun May 7 18:31:25 2006 Ben Pfaff <blp@gnu.org>
350 * aggregate.c (cmd_aggregate): Free default_dict before replacing
353 Sun May 7 17:09:19 2006 Ben Pfaff <blp@gnu.org>
355 * flip.c (flip_file): Check return value of pool_fclose().
357 Sat May 6 16:00:13 2006 Ben Pfaff <blp@gnu.org>
359 Get rid of `char *c' member in union value, for cleanliness.
361 * aggregate.c: (union agr_argument) New union.
362 (struct agr_var) Change element type of arg[] from union value to
364 (parse_aggregate_functions) Change local variable types likewise.
366 * autorecode.c: (union arc_value) New union.
367 (struct arc_item) Change "from" from union value to union
369 (recode) Change local variable from union value to union
371 (autorecode_trns_proc) Ditto.
372 (compare_alpha_value) Ditto.
373 (hash_alpha_value) Ditto.
374 (compare_numeric_value) Ditto.
375 (hash_numeric_value) Ditto.
376 (autorecode_proc_func) Ditto.
378 Sat May 6 10:43:33 2006 Ben Pfaff <blp@gnu.org>
380 Continue reforming procedure execution. In this phase, get rid of
381 the output code for SPLIT FILE groups in procedure.c, which really
382 shouldn't be doing any output. Move it into the individual
383 procedures instead. This also adds some flexibility.
385 * crosstabs.q (precalc): Call output_split_file_values().
387 * descriptives.c (calc_descriptives): Ditto.
389 * examine.q (run_examine): Ditto.
391 * frequencies.q (precalc): Ditto.
393 * oneway.q (run_oneway): Ditto.
395 * regression.q (run_regression): Ditto.
397 * t-test.q (calculate): Ditto.
399 Wed May 3 23:05:31 2006 Ben Pfaff <blp@gnu.org>
401 Continue reforming procedure execution. In this phase, get rid of
402 many global variables, consolidating procedure execution in
403 procedure.c. Encapsulate transformations in new "struct
404 trns_chain". Also, change implementation of N OF CASES, FILTER,
405 and PROCESS IF from special cases to transformations.
407 * aggregate.c (cmd_aggregate) Use discard_variables().
409 2006-04-28 Jason Stover <jhs@math.gcsu.edu>
411 * regression.q (regression_trns_resid_proc): Pass only the
412 variables used in the model to (*model->residual)().
414 * regression.q (regression_trns_pred_proc): Pass only the
415 variables used in the model to (*model->pred)().
417 2006-04-26 Jason Stover <jhs@math.gcsu.edu>
419 * regression.q: Added support for multiple transformations.
421 * regression.q (regression_trns_resid_proc): New function.
423 * regression.q (regression_trns_pred_proc): New function.
425 * regression.q (subcommand_save): Added support for saving
428 * regression.q (regression_trns_free): New function.
430 * regression.q (reg_get_name): New function.
432 * regression.q (reg_save_var): New function.
434 Tue Apr 25 13:18:56 2006 Ben Pfaff <blp@gnu.org>
436 * rank.q (parse_rank_function): Use SE instead of ME for parse
439 Tue Apr 25 13:16:28 2006 Ben Pfaff <blp@gnu.org>
441 * flip.c (flip_sink_write): Use snprintf() to simplify a bit of
444 2006-04-21 Jason Stover <jhs@math.gcsu.edu>
446 * regression.q (try_name): New function. (Partly copied from
447 try_name in descriptives.c.)
449 * regression.q (subcommand_save): Choose residual variable names
452 2006-04-20 Jason Stover <jhs@math.gcsu.edu>
454 * regression.q (cmd_regression): Moved call to subcommand_save()
455 outside multipass_procedure_with_splits().
457 * regression.q (regression_trns_proc): Fixed value counter n_vals
458 before calling *model->residual().
460 2006-04-19 Jason Stover <jhs@math.gcsu.edu>
462 * regression.q (regression_trns_proc): Fixed the look-up of the
465 2006-04-18 Jason Stover <jhs@math.gcsu.edu>
467 * regression.q (regression_trns_proc): Look up the residual
468 variable in the linear regression cache.
470 * regression.q (subcommand_save): Set the residual variable in the
471 linear regression cache.
473 2006-04-17 Jason Stover <jason@wonko.gcsu.edu>
475 * regression.q (regression_trns_proc): Accept case_idx as an int
476 to match the definition of trns_proc_func.
478 2006-04-17 Jason Stover <jhs@math.gcsu.edu>
480 * regression.q (regression_trns_proc): New function.
482 * regression.q (subcommand_save): Create variable residuals and
483 add a transformation to assign values to them. Also free the
484 linreg_cache if the SAVE command was not called. Removed the
485 casereading loop. Placed actual computation of residuals in
486 regression_trns_proc.
488 * regression.q (run_regression): Moved call to free
489 pspp_linreg_cache to subcommand_save.
491 Sat Apr 15 18:01:03 2006 Ben Pfaff <blp@gnu.org>
493 * examine.q (output_examine): Add casts to fix warnings.
495 2006-04-07 Jason Stover <jhs@math.gcsu.edu>
497 * regression.q (subcommand_save): New function.
499 2006-04-04 Jason Stover <jhs@math.gcsu.edu>
501 * regression.q: New function reg_has_categorical () to tell
502 whether a pspp_linreg_struct was made with any variables of type
505 * regression.q: (subcommand_export): Call
506 reg_print_categorical_encoding() only if the model uses any
507 categorical variables.
509 Mon Mar 27 16:00:42 2006 Ben Pfaff <blp@gnu.org>
511 * crosstabs.q: (output_pivot_table) Drop spurious space from
514 2006-03-15 Jason Stover <jhs@math.gcsu.edu>
516 * regression.q: Added custom syntax parser for VARIABLES subcommand
518 * regression.q: Moved most instructions for run_regression ()
519 inside the loop over dependent variables.
521 Thu Mar 2 08:40:33 WST 2006 John Darrington <john@darrington.wattle.id.au>
523 * Moved files from src directory