1 2007-08-03 Ben Pfaff <blp@gnu.org>
3 * rank.q (rank_cmd): Instead of sorting by SPLIT FILE vars, group
4 by them. Fixes bug #17239.
5 Reviewed by John Darrington.
7 2007-08-01 Ben Pfaff <blp@gnu.org>
9 Clean up handling of median, by treating it almost like any other
10 percentile. Fixes bug #17424. Thanks to John Darrington for
12 * frequencies.q (internal_cmd_frequencies): Fix handling of bit
13 masks for `stats' variable. If median is selected, turn it off
14 and add a 50th percentile.
15 (add_percentile): Simplify code a little.
16 (calc_stats): Drop special casing of median.
17 (dump_statistics): Ditto, except that we label the 50th percentile
18 as "50 (Median)" to make it clear that it's also the median.
20 2007-07-31 Ben Pfaff <blp@gnu.org>
22 Remove integer mode from FREQUENCIES and incidentally fix bug
23 #17421. Reviewed by John Darrington.
24 * frequencies.q (int_pool): Rename data_pool.
25 (gen_pool): Rename syntax_pool.
26 (enum FRQM_*): Removed.
27 (struct freq_tab): Removed `mode', `vector', `min', `max',
28 `out_of_range', `sysmis' members.
29 (calc): Delete support for integer mode.
31 (postprocess_freq_tab): Ditto.
32 (cleanup_freq_tab): Ditto.
33 (frq_custom_variables): Ditto.
35 2007-07-28 John Darrington <john@darrington.wattle.id.au>
37 * t-test.q: Moved the order in which groups are displayed in the
38 independent samples case, where a cut point is given.
40 2007-07-27 Ben Pfaff <blp@gnu.org>
42 * regression.q (run_regression): Move casereader_destroy call so
43 that it always gets called, not just if there was some valid
44 data. Fixes bug #19581.
45 Reviewed by Jason Stover.
47 2007-07-24 Ben Pfaff <blp@gnu.org>
49 * flip.c (struct flip_pgm): Remove `case_size' member (now
51 (cmd_flip): Pass var_cnt as number of cases instead of case_cnt,
52 to fix bug #20494. Don't assign to `case_size' member, which was
53 unused after assignment.
54 (build_dictionary): When NEWNAMES not used, get the number of
55 variables right, to fix bug #20493.
57 2007-07-10 Jason Stover <jhs@math.gcsu.edu>
59 * glm.q: Initial version of the GLM procedure.
61 2007-06-06 Ben Pfaff <blp@gnu.org>
63 Adapt case sources, sinks, and clients of procedure code to the
66 * aggregate.c: Simplify greatly since everything is more uniform
69 * autorecode.c: Adapt to new procedure code.
73 * descriptives.c: Ditto.
75 * npar-summary.c: Ditto.
76 * frequencies.q: Ditto.
79 * regression.q: Ditto.
80 * sort-cases.c: Ditto.
83 * sort-criteria.c: Rewrite to output a struct case_ordering.
85 * flip.c: Rewrite to be a casereader.
87 * rank.q: Simplify greatly since casereaders are much more
88 flexible than what we had before.
90 2007-05-15 Jason Stover <jhs@math.gcsu.edu>
92 * regression.q (run_regression): Tell the user when the data
93 contain no valid cases.
95 2007-05-08 Jason Stover <jhs@math.gcsu.edu>
97 * regression.q: Partial fix of bug which caused a crash if
98 dependent variable and independent variable were the same.
100 2007-04-16 John Darrington <john@darrington.wattle.id.au>
102 * t-test.q: Changed the output width of reported counts and
103 degrees of freedom, to avoid truncating these values. Thanks
104 to Seth Woolley for reporting this problem. A proper fix involves
105 re-thinking the output driver.
107 2007-04-12 Jason Stover <jhs@math.gcsu.edu>
109 * regression.q (run_regression): Added if (n_data >0) to fix bug
112 2007-03-29 Jason Stover <jhs@math.gcsu.edu>
114 * regression.q (prepare_data): New function.
116 * regression.q (compute_moments): New function.
118 2007-03-18 Ben Pfaff <blp@gnu.org>
120 * crosstabs.q (static var write): Rename write_style to avoid
121 conflict with POSIX function of same name.
123 2007-03-16 Jason Stover <jhs@math.gcsu.edu>
125 * regression.q (run_regression): Added support for moments.
127 Sat Feb 17 08:16:00 2007 Ben Pfaff <blp@gnu.org>
129 * flip.c (flip_sink_create): Improve error message when temporary
130 file cannot be created.
132 Tue Feb 6 19:58:03 2007 Ben Pfaff <blp@gnu.org>
134 * flip.c (flip_file): Give better error message on end-of-file.
136 2007-02-04 Jason Stover <jhs@math.gcsu.edu>
138 * regression.q: Fixed p-value computation in the test for
139 individual regression coefficients.
141 Mon Jan 15 11:03:20 2007 Ben Pfaff <blp@gnu.org>
143 Fix bugs found by valgrind when --enable-debug is used with the
144 new case code. These bugs are hidden when the data set is small
145 enough to find in memory; if a bigger data set that would overflow
146 to disk were used, then data corruption would occur.
148 * chisquare.c (create_freq_hash): Pass free_freq_mutable_hash to
149 hsh_create as free function. Make copy of data put into hash.
151 * oneway.q (free_value): New function.
152 (run_oneway): Use free_value as arg to hsh_create. Make copy of
155 * rank.q (rank_cases): Don't access data in case after we've given
158 Tue Jan 9 19:16:11 2007 Ben Pfaff <blp@gnu.org>
162 * aggregate.c (parse_aggregate_functions) Initialize
165 Fri Dec 22 14:04:09 2006 Ben Pfaff <blp@gnu.org>
167 Simplify missing value handling.
169 * aggregate.c (struct agr_var): Remove `bool include_missing', add
170 `enum mv_class exclude'. Remove `int missing', add `bool
171 saw_missing'. Update users.
173 * descriptives.c (struct dsc_trns): Removed `int
174 include_user_missing', add `enum mv_class exclude'. Update users.
175 (struct dsc_proc): Ditto.
177 * examine.q: (static var value_is_missing): Rename
178 `exclude_values', change type to `enum mv_class'. Update users.
182 Fri Dec 22 19:22:18 WST 2006 John Darrington <john@darrington.wattle.id.au>
184 * frequencies.q : Fixed bug #17420, where the table bounds were overun
185 when /FORMAT=nolabels was given.
187 Wed Dec 20 18:45:31 WST 2006 John Darrington <john@darrington.wattle.id.au>
189 * binomial.c binomial.h : New files. Thanks to Jason Stover
190 for assistance with these.
192 * chisquare.c chisquare.h freq.c freq.h npar-summary.c npar-summary.h
193 npar.h npar.q: New files. Implementing NPAR TESTS.
195 * frequencies.q : Moved structure definitions into freq.[ch]
197 Sat Dec 16 22:26:44 2006 Ben Pfaff <blp@gnu.org>
199 Make it possible to pull cases from the active file with a
200 function call, instead of requiring indirection through a callback
203 * aggregate.c (cmd_aggregate): Take advantage of new procedure
205 (agr_to_active_file): Removed.
206 (presorted_agr_to_sysfile): Removed.
208 * autorecode.c (cmd_autorecode): Take advantage of new procedure
210 (autorecode_proc_func): Removed.
212 * flip.c (struct flip_pgm): New members to allow conformance with
213 new case_source_class interface.
214 (cmd_flip): Adapt to new case_source_class interface.
215 (flip_source_read): Ditto.
216 (flip_source_destroy): Ditto.
218 Sat Dec 16 12:54:27 2006 Ben Pfaff <blp@gnu.org>
220 * rank.q (rank_custom_variables): Allow grouping variables to be
221 strings. Fixes bug #18533. Thanks to John Darrington for review.
223 Sat Dec 9 18:47:51 2006 Ben Pfaff <blp@gnu.org>
225 * regression.q (is_depvar): Compare variable pointers instead of
228 Thu Dec 7 15:26:25 WST 2006 John Darrington <john@darrington.wattle.id.au>
230 * examine.q: Allocated the categorical values for the dependent and
231 independent variables, on the heap. Hence they can be of any width.
233 Wed Dec 6 21:14:26 2006 Ben Pfaff <blp@gnu.org>
235 * regression.q (reg_inserted): Compare variable pointers instead
238 Mon Dec 4 22:33:46 2006 Ben Pfaff <blp@gnu.org>
240 * crosstabs.q (insert_summary): Use var_to_string for labeling.
241 (output_pivot_table) Ditto.
244 * frequencies.q (setup_z_trns): Ditto.
246 (dump_condensed) Ditto.
247 (dump_statistics) Ditto.
250 Sun Nov 5 08:31:42 WST 2006 John Darrington <john@darrington.wattle.id.au>
252 * t-test.q, oneway.q: Changed to use the new casefilter structure.
254 Sat Oct 14 16:52:28 2006 Ben Pfaff <blp@gnu.org>
256 * rank.q: (rank_sorted_casefile) Add some missing case_destroy()
257 calls to fix a memory leak.
259 Sun Oct 8 09:45:40 WST 2006 John Darrington <john@darrington.wattle.id.au>
261 * rank.q: Plugged a small memory leak which occurred under error
264 Sat Oct 7 11:06:01 WST 2006 John Darrington <john@darrington.wattle.id.au>
266 * rank.q: Implemented most of the RANK command.
268 2006-07-14 Jason Stover <jhs@math.gcsu.edu>
270 * regression.q (run_regression): New function to move knowledge of
271 pspp_linreg_cache out of math/coefficient.[ch].
273 Sat Jul 1 17:41:46 2006 Ben Pfaff <blp@gnu.org>
275 Fix bug #11612, "q2c documentation does not agree with code".
277 * examine.q: Audit use of q2c "+" prefixes that indicate that a
278 command may appear multiple times.
280 * frequencies.q: Ditto.
284 * regression.q: Ditto.
288 Fri Jun 23 14:18:22 2006 Ben Pfaff <blp@gnu.org>
290 Support long string variables on FREQUENCIES, as
291 an extension when in enhanced algorithms mode. For Greg Hunt
292 <greg@firmansyah.com>.
294 * frequencies.q: (struct freq) Change `v' member from union value
295 to union value *. Update all references.
296 (struct var_freqs) Add width, print members to represent effective
297 variable width and display format.
298 (calc) Copy entire long string value into the hash table.
299 (frq_custom_variables) Set new width, print members.
300 (hash_value_alpha) Get width from var_freqs.
301 (compare_value_alpha_a) Ditto.
302 (compare_freq_alpha_a) Ditto.
303 (compare_freq_alpha_d) Ditto.
304 (dump_full) Get display format from var_freqs.
305 (dump_condensed) Ditto.
307 Mon Jun 19 22:07:13 2006 Ben Pfaff <blp@gnu.org>
309 * frequencies.q: (dump_full) Only put the first MAX_SHORT_STRING
310 bytes of string variables into the output cells, seeing as we only
312 (dump_condensed) Ditto.
314 Mon Jun 19 21:52:05 2006 Ben Pfaff <blp@gnu.org>
316 Fixes a bug reported by Greg Hunt <greg@firmansyah.com>.
318 * frequencies.q: (hsh_hash_bytes) We only copy the first
319 MAX_SHORT_STRING bytes of string variables, so we must only
320 compare that many bytes, even if the string variable is longer.
321 (compare_value_alpha_a) Ditto.
322 (compare_freq_alpha_a) Ditto.
323 (compare_freq_alpha_d) Ditto.
325 2006-05-11 Jason Stover <jhs@math.gcsu.edu>
327 * regression.q: Adjusted code to account for cache->coeff being a
328 pspp_linreg_coeff **.
330 Sun May 7 18:31:25 2006 Ben Pfaff <blp@gnu.org>
334 * aggregate.c (cmd_aggregate): Free default_dict before replacing
337 Sun May 7 17:09:19 2006 Ben Pfaff <blp@gnu.org>
339 * flip.c (flip_file): Check return value of pool_fclose().
341 Sat May 6 16:00:13 2006 Ben Pfaff <blp@gnu.org>
343 Get rid of `char *c' member in union value, for cleanliness.
345 * aggregate.c: (union agr_argument) New union.
346 (struct agr_var) Change element type of arg[] from union value to
348 (parse_aggregate_functions) Change local variable types likewise.
350 * autorecode.c: (union arc_value) New union.
351 (struct arc_item) Change "from" from union value to union
353 (recode) Change local variable from union value to union
355 (autorecode_trns_proc) Ditto.
356 (compare_alpha_value) Ditto.
357 (hash_alpha_value) Ditto.
358 (compare_numeric_value) Ditto.
359 (hash_numeric_value) Ditto.
360 (autorecode_proc_func) Ditto.
362 Sat May 6 10:43:33 2006 Ben Pfaff <blp@gnu.org>
364 Continue reforming procedure execution. In this phase, get rid of
365 the output code for SPLIT FILE groups in procedure.c, which really
366 shouldn't be doing any output. Move it into the individual
367 procedures instead. This also adds some flexibility.
369 * crosstabs.q (precalc): Call output_split_file_values().
371 * descriptives.c (calc_descriptives): Ditto.
373 * examine.q (run_examine): Ditto.
375 * frequencies.q (precalc): Ditto.
377 * oneway.q (run_oneway): Ditto.
379 * regression.q (run_regression): Ditto.
381 * t-test.q (calculate): Ditto.
383 Wed May 3 23:05:31 2006 Ben Pfaff <blp@gnu.org>
385 Continue reforming procedure execution. In this phase, get rid of
386 many global variables, consolidating procedure execution in
387 procedure.c. Encapsulate transformations in new "struct
388 trns_chain". Also, change implementation of N OF CASES, FILTER,
389 and PROCESS IF from special cases to transformations.
391 * aggregate.c (cmd_aggregate) Use discard_variables().
393 2006-04-28 Jason Stover <jhs@math.gcsu.edu>
395 * regression.q (regression_trns_resid_proc): Pass only the
396 variables used in the model to (*model->residual)().
398 * regression.q (regression_trns_pred_proc): Pass only the
399 variables used in the model to (*model->pred)().
401 2006-04-26 Jason Stover <jhs@math.gcsu.edu>
403 * regression.q: Added support for multiple transformations.
405 * regression.q (regression_trns_resid_proc): New function.
407 * regression.q (regression_trns_pred_proc): New function.
409 * regression.q (subcommand_save): Added support for saving
412 * regression.q (regression_trns_free): New function.
414 * regression.q (reg_get_name): New function.
416 * regression.q (reg_save_var): New function.
418 Tue Apr 25 13:18:56 2006 Ben Pfaff <blp@gnu.org>
420 * rank.q (parse_rank_function): Use SE instead of ME for parse
423 Tue Apr 25 13:16:28 2006 Ben Pfaff <blp@gnu.org>
425 * flip.c (flip_sink_write): Use snprintf() to simplify a bit of
428 2006-04-21 Jason Stover <jhs@math.gcsu.edu>
430 * regression.q (try_name): New function. (Partly copied from
431 try_name in descriptives.c.)
433 * regression.q (subcommand_save): Choose residual variable names
436 2006-04-20 Jason Stover <jhs@math.gcsu.edu>
438 * regression.q (cmd_regression): Moved call to subcommand_save()
439 outside multipass_procedure_with_splits().
441 * regression.q (regression_trns_proc): Fixed value counter n_vals
442 before calling *model->residual().
444 2006-04-19 Jason Stover <jhs@math.gcsu.edu>
446 * regression.q (regression_trns_proc): Fixed the look-up of the
449 2006-04-18 Jason Stover <jhs@math.gcsu.edu>
451 * regression.q (regression_trns_proc): Look up the residual
452 variable in the linear regression cache.
454 * regression.q (subcommand_save): Set the residual variable in the
455 linear regression cache.
457 2006-04-17 Jason Stover <jason@wonko.gcsu.edu>
459 * regression.q (regression_trns_proc): Accept case_idx as an int
460 to match the definition of trns_proc_func.
462 2006-04-17 Jason Stover <jhs@math.gcsu.edu>
464 * regression.q (regression_trns_proc): New function.
466 * regression.q (subcommand_save): Create variable residuals and
467 add a transformation to assign values to them. Also free the
468 linreg_cache if the SAVE command was not called. Removed the
469 casereading loop. Placed actual computation of residuals in
470 regression_trns_proc.
472 * regression.q (run_regression): Moved call to free
473 pspp_linreg_cache to subcommand_save.
475 Sat Apr 15 18:01:03 2006 Ben Pfaff <blp@gnu.org>
477 * examine.q (output_examine): Add casts to fix warnings.
479 2006-04-07 Jason Stover <jhs@math.gcsu.edu>
481 * regression.q (subcommand_save): New function.
483 2006-04-04 Jason Stover <jhs@math.gcsu.edu>
485 * regression.q: New function reg_has_categorical () to tell
486 whether a pspp_linreg_struct was made with any variables of type
489 * regression.q: (subcommand_export): Call
490 reg_print_categorical_encoding() only if the model uses any
491 categorical variables.
493 Mon Mar 27 16:00:42 2006 Ben Pfaff <blp@gnu.org>
495 * crosstabs.q: (output_pivot_table) Drop spurious space from
498 2006-03-15 Jason Stover <jhs@math.gcsu.edu>
500 * regression.q: Added custom syntax parser for VARIABLES subcommand
502 * regression.q: Moved most instructions for run_regression ()
503 inside the loop over dependent variables.
505 Thu Mar 2 08:40:33 WST 2006 John Darrington <john@darrington.wattle.id.au>
507 * Moved files from src directory