1 2007-08-15 Jason Stover <jhs@wonko.gcsu.edu>
3 * regression.q (identify_indep_vars): Print an error if dependent
4 and independent variables are the same. Fixes bug 19819.
6 2007-08-12 Ben Pfaff <blp@gnu.org>
8 * flip.c: Drop use of dict_get_compacted_dict_index_to_case_index
9 and just use the ordinary case indexes. There seemed to be no
10 reason for the former method.
12 2007-08-03 Ben Pfaff <blp@gnu.org>
14 * rank.q (rank_cmd): Instead of sorting by SPLIT FILE vars, group
15 by them. Fixes bug #17239.
16 Reviewed by John Darrington.
18 2007-08-01 Ben Pfaff <blp@gnu.org>
20 Clean up handling of median, by treating it almost like any other
21 percentile. Fixes bug #17424. Thanks to John Darrington for
23 * frequencies.q (internal_cmd_frequencies): Fix handling of bit
24 masks for `stats' variable. If median is selected, turn it off
25 and add a 50th percentile.
26 (add_percentile): Simplify code a little.
27 (calc_stats): Drop special casing of median.
28 (dump_statistics): Ditto, except that we label the 50th percentile
29 as "50 (Median)" to make it clear that it's also the median.
31 2007-07-31 Ben Pfaff <blp@gnu.org>
33 Remove integer mode from FREQUENCIES and incidentally fix bug
34 #17421. Reviewed by John Darrington.
35 * frequencies.q (int_pool): Rename data_pool.
36 (gen_pool): Rename syntax_pool.
37 (enum FRQM_*): Removed.
38 (struct freq_tab): Removed `mode', `vector', `min', `max',
39 `out_of_range', `sysmis' members.
40 (calc): Delete support for integer mode.
42 (postprocess_freq_tab): Ditto.
43 (cleanup_freq_tab): Ditto.
44 (frq_custom_variables): Ditto.
46 2007-07-28 John Darrington <john@darrington.wattle.id.au>
48 * t-test.q: Moved the order in which groups are displayed in the
49 independent samples case, where a cut point is given.
51 2007-07-27 Ben Pfaff <blp@gnu.org>
53 * regression.q (run_regression): Move casereader_destroy call so
54 that it always gets called, not just if there was some valid
55 data. Fixes bug #19581.
56 Reviewed by Jason Stover.
58 2007-07-24 Ben Pfaff <blp@gnu.org>
60 * flip.c (struct flip_pgm): Remove `case_size' member (now
62 (cmd_flip): Pass var_cnt as number of cases instead of case_cnt,
63 to fix bug #20494. Don't assign to `case_size' member, which was
64 unused after assignment.
65 (build_dictionary): When NEWNAMES not used, get the number of
66 variables right, to fix bug #20493.
68 2007-07-10 Jason Stover <jhs@math.gcsu.edu>
70 * glm.q: Initial version of the GLM procedure.
72 2007-06-06 Ben Pfaff <blp@gnu.org>
74 Adapt case sources, sinks, and clients of procedure code to the
77 * aggregate.c: Simplify greatly since everything is more uniform
80 * autorecode.c: Adapt to new procedure code.
84 * descriptives.c: Ditto.
86 * npar-summary.c: Ditto.
87 * frequencies.q: Ditto.
90 * regression.q: Ditto.
91 * sort-cases.c: Ditto.
94 * sort-criteria.c: Rewrite to output a struct case_ordering.
96 * flip.c: Rewrite to be a casereader.
98 * rank.q: Simplify greatly since casereaders are much more
99 flexible than what we had before.
101 2007-05-15 Jason Stover <jhs@math.gcsu.edu>
103 * regression.q (run_regression): Tell the user when the data
104 contain no valid cases.
106 2007-05-08 Jason Stover <jhs@math.gcsu.edu>
108 * regression.q: Partial fix of bug which caused a crash if
109 dependent variable and independent variable were the same.
111 2007-04-16 John Darrington <john@darrington.wattle.id.au>
113 * t-test.q: Changed the output width of reported counts and
114 degrees of freedom, to avoid truncating these values. Thanks
115 to Seth Woolley for reporting this problem. A proper fix involves
116 re-thinking the output driver.
118 2007-04-12 Jason Stover <jhs@math.gcsu.edu>
120 * regression.q (run_regression): Added if (n_data >0) to fix bug
123 2007-03-29 Jason Stover <jhs@math.gcsu.edu>
125 * regression.q (prepare_data): New function.
127 * regression.q (compute_moments): New function.
129 2007-03-18 Ben Pfaff <blp@gnu.org>
131 * crosstabs.q (static var write): Rename write_style to avoid
132 conflict with POSIX function of same name.
134 2007-03-16 Jason Stover <jhs@math.gcsu.edu>
136 * regression.q (run_regression): Added support for moments.
138 Sat Feb 17 08:16:00 2007 Ben Pfaff <blp@gnu.org>
140 * flip.c (flip_sink_create): Improve error message when temporary
141 file cannot be created.
143 Tue Feb 6 19:58:03 2007 Ben Pfaff <blp@gnu.org>
145 * flip.c (flip_file): Give better error message on end-of-file.
147 2007-02-04 Jason Stover <jhs@math.gcsu.edu>
149 * regression.q: Fixed p-value computation in the test for
150 individual regression coefficients.
152 Mon Jan 15 11:03:20 2007 Ben Pfaff <blp@gnu.org>
154 Fix bugs found by valgrind when --enable-debug is used with the
155 new case code. These bugs are hidden when the data set is small
156 enough to find in memory; if a bigger data set that would overflow
157 to disk were used, then data corruption would occur.
159 * chisquare.c (create_freq_hash): Pass free_freq_mutable_hash to
160 hsh_create as free function. Make copy of data put into hash.
162 * oneway.q (free_value): New function.
163 (run_oneway): Use free_value as arg to hsh_create. Make copy of
166 * rank.q (rank_cases): Don't access data in case after we've given
169 Tue Jan 9 19:16:11 2007 Ben Pfaff <blp@gnu.org>
173 * aggregate.c (parse_aggregate_functions) Initialize
176 Fri Dec 22 14:04:09 2006 Ben Pfaff <blp@gnu.org>
178 Simplify missing value handling.
180 * aggregate.c (struct agr_var): Remove `bool include_missing', add
181 `enum mv_class exclude'. Remove `int missing', add `bool
182 saw_missing'. Update users.
184 * descriptives.c (struct dsc_trns): Removed `int
185 include_user_missing', add `enum mv_class exclude'. Update users.
186 (struct dsc_proc): Ditto.
188 * examine.q: (static var value_is_missing): Rename
189 `exclude_values', change type to `enum mv_class'. Update users.
193 Fri Dec 22 19:22:18 WST 2006 John Darrington <john@darrington.wattle.id.au>
195 * frequencies.q : Fixed bug #17420, where the table bounds were overun
196 when /FORMAT=nolabels was given.
198 Wed Dec 20 18:45:31 WST 2006 John Darrington <john@darrington.wattle.id.au>
200 * binomial.c binomial.h : New files. Thanks to Jason Stover
201 for assistance with these.
203 * chisquare.c chisquare.h freq.c freq.h npar-summary.c npar-summary.h
204 npar.h npar.q: New files. Implementing NPAR TESTS.
206 * frequencies.q : Moved structure definitions into freq.[ch]
208 Sat Dec 16 22:26:44 2006 Ben Pfaff <blp@gnu.org>
210 Make it possible to pull cases from the active file with a
211 function call, instead of requiring indirection through a callback
214 * aggregate.c (cmd_aggregate): Take advantage of new procedure
216 (agr_to_active_file): Removed.
217 (presorted_agr_to_sysfile): Removed.
219 * autorecode.c (cmd_autorecode): Take advantage of new procedure
221 (autorecode_proc_func): Removed.
223 * flip.c (struct flip_pgm): New members to allow conformance with
224 new case_source_class interface.
225 (cmd_flip): Adapt to new case_source_class interface.
226 (flip_source_read): Ditto.
227 (flip_source_destroy): Ditto.
229 Sat Dec 16 12:54:27 2006 Ben Pfaff <blp@gnu.org>
231 * rank.q (rank_custom_variables): Allow grouping variables to be
232 strings. Fixes bug #18533. Thanks to John Darrington for review.
234 Sat Dec 9 18:47:51 2006 Ben Pfaff <blp@gnu.org>
236 * regression.q (is_depvar): Compare variable pointers instead of
239 Thu Dec 7 15:26:25 WST 2006 John Darrington <john@darrington.wattle.id.au>
241 * examine.q: Allocated the categorical values for the dependent and
242 independent variables, on the heap. Hence they can be of any width.
244 Wed Dec 6 21:14:26 2006 Ben Pfaff <blp@gnu.org>
246 * regression.q (reg_inserted): Compare variable pointers instead
249 Mon Dec 4 22:33:46 2006 Ben Pfaff <blp@gnu.org>
251 * crosstabs.q (insert_summary): Use var_to_string for labeling.
252 (output_pivot_table) Ditto.
255 * frequencies.q (setup_z_trns): Ditto.
257 (dump_condensed) Ditto.
258 (dump_statistics) Ditto.
261 Sun Nov 5 08:31:42 WST 2006 John Darrington <john@darrington.wattle.id.au>
263 * t-test.q, oneway.q: Changed to use the new casefilter structure.
265 Sat Oct 14 16:52:28 2006 Ben Pfaff <blp@gnu.org>
267 * rank.q: (rank_sorted_casefile) Add some missing case_destroy()
268 calls to fix a memory leak.
270 Sun Oct 8 09:45:40 WST 2006 John Darrington <john@darrington.wattle.id.au>
272 * rank.q: Plugged a small memory leak which occurred under error
275 Sat Oct 7 11:06:01 WST 2006 John Darrington <john@darrington.wattle.id.au>
277 * rank.q: Implemented most of the RANK command.
279 2006-07-14 Jason Stover <jhs@math.gcsu.edu>
281 * regression.q (run_regression): New function to move knowledge of
282 pspp_linreg_cache out of math/coefficient.[ch].
284 Sat Jul 1 17:41:46 2006 Ben Pfaff <blp@gnu.org>
286 Fix bug #11612, "q2c documentation does not agree with code".
288 * examine.q: Audit use of q2c "+" prefixes that indicate that a
289 command may appear multiple times.
291 * frequencies.q: Ditto.
295 * regression.q: Ditto.
299 Fri Jun 23 14:18:22 2006 Ben Pfaff <blp@gnu.org>
301 Support long string variables on FREQUENCIES, as
302 an extension when in enhanced algorithms mode. For Greg Hunt
303 <greg@firmansyah.com>.
305 * frequencies.q: (struct freq) Change `v' member from union value
306 to union value *. Update all references.
307 (struct var_freqs) Add width, print members to represent effective
308 variable width and display format.
309 (calc) Copy entire long string value into the hash table.
310 (frq_custom_variables) Set new width, print members.
311 (hash_value_alpha) Get width from var_freqs.
312 (compare_value_alpha_a) Ditto.
313 (compare_freq_alpha_a) Ditto.
314 (compare_freq_alpha_d) Ditto.
315 (dump_full) Get display format from var_freqs.
316 (dump_condensed) Ditto.
318 Mon Jun 19 22:07:13 2006 Ben Pfaff <blp@gnu.org>
320 * frequencies.q: (dump_full) Only put the first MAX_SHORT_STRING
321 bytes of string variables into the output cells, seeing as we only
323 (dump_condensed) Ditto.
325 Mon Jun 19 21:52:05 2006 Ben Pfaff <blp@gnu.org>
327 Fixes a bug reported by Greg Hunt <greg@firmansyah.com>.
329 * frequencies.q: (hsh_hash_bytes) We only copy the first
330 MAX_SHORT_STRING bytes of string variables, so we must only
331 compare that many bytes, even if the string variable is longer.
332 (compare_value_alpha_a) Ditto.
333 (compare_freq_alpha_a) Ditto.
334 (compare_freq_alpha_d) Ditto.
336 2006-05-11 Jason Stover <jhs@math.gcsu.edu>
338 * regression.q: Adjusted code to account for cache->coeff being a
339 pspp_linreg_coeff **.
341 Sun May 7 18:31:25 2006 Ben Pfaff <blp@gnu.org>
345 * aggregate.c (cmd_aggregate): Free default_dict before replacing
348 Sun May 7 17:09:19 2006 Ben Pfaff <blp@gnu.org>
350 * flip.c (flip_file): Check return value of pool_fclose().
352 Sat May 6 16:00:13 2006 Ben Pfaff <blp@gnu.org>
354 Get rid of `char *c' member in union value, for cleanliness.
356 * aggregate.c: (union agr_argument) New union.
357 (struct agr_var) Change element type of arg[] from union value to
359 (parse_aggregate_functions) Change local variable types likewise.
361 * autorecode.c: (union arc_value) New union.
362 (struct arc_item) Change "from" from union value to union
364 (recode) Change local variable from union value to union
366 (autorecode_trns_proc) Ditto.
367 (compare_alpha_value) Ditto.
368 (hash_alpha_value) Ditto.
369 (compare_numeric_value) Ditto.
370 (hash_numeric_value) Ditto.
371 (autorecode_proc_func) Ditto.
373 Sat May 6 10:43:33 2006 Ben Pfaff <blp@gnu.org>
375 Continue reforming procedure execution. In this phase, get rid of
376 the output code for SPLIT FILE groups in procedure.c, which really
377 shouldn't be doing any output. Move it into the individual
378 procedures instead. This also adds some flexibility.
380 * crosstabs.q (precalc): Call output_split_file_values().
382 * descriptives.c (calc_descriptives): Ditto.
384 * examine.q (run_examine): Ditto.
386 * frequencies.q (precalc): Ditto.
388 * oneway.q (run_oneway): Ditto.
390 * regression.q (run_regression): Ditto.
392 * t-test.q (calculate): Ditto.
394 Wed May 3 23:05:31 2006 Ben Pfaff <blp@gnu.org>
396 Continue reforming procedure execution. In this phase, get rid of
397 many global variables, consolidating procedure execution in
398 procedure.c. Encapsulate transformations in new "struct
399 trns_chain". Also, change implementation of N OF CASES, FILTER,
400 and PROCESS IF from special cases to transformations.
402 * aggregate.c (cmd_aggregate) Use discard_variables().
404 2006-04-28 Jason Stover <jhs@math.gcsu.edu>
406 * regression.q (regression_trns_resid_proc): Pass only the
407 variables used in the model to (*model->residual)().
409 * regression.q (regression_trns_pred_proc): Pass only the
410 variables used in the model to (*model->pred)().
412 2006-04-26 Jason Stover <jhs@math.gcsu.edu>
414 * regression.q: Added support for multiple transformations.
416 * regression.q (regression_trns_resid_proc): New function.
418 * regression.q (regression_trns_pred_proc): New function.
420 * regression.q (subcommand_save): Added support for saving
423 * regression.q (regression_trns_free): New function.
425 * regression.q (reg_get_name): New function.
427 * regression.q (reg_save_var): New function.
429 Tue Apr 25 13:18:56 2006 Ben Pfaff <blp@gnu.org>
431 * rank.q (parse_rank_function): Use SE instead of ME for parse
434 Tue Apr 25 13:16:28 2006 Ben Pfaff <blp@gnu.org>
436 * flip.c (flip_sink_write): Use snprintf() to simplify a bit of
439 2006-04-21 Jason Stover <jhs@math.gcsu.edu>
441 * regression.q (try_name): New function. (Partly copied from
442 try_name in descriptives.c.)
444 * regression.q (subcommand_save): Choose residual variable names
447 2006-04-20 Jason Stover <jhs@math.gcsu.edu>
449 * regression.q (cmd_regression): Moved call to subcommand_save()
450 outside multipass_procedure_with_splits().
452 * regression.q (regression_trns_proc): Fixed value counter n_vals
453 before calling *model->residual().
455 2006-04-19 Jason Stover <jhs@math.gcsu.edu>
457 * regression.q (regression_trns_proc): Fixed the look-up of the
460 2006-04-18 Jason Stover <jhs@math.gcsu.edu>
462 * regression.q (regression_trns_proc): Look up the residual
463 variable in the linear regression cache.
465 * regression.q (subcommand_save): Set the residual variable in the
466 linear regression cache.
468 2006-04-17 Jason Stover <jason@wonko.gcsu.edu>
470 * regression.q (regression_trns_proc): Accept case_idx as an int
471 to match the definition of trns_proc_func.
473 2006-04-17 Jason Stover <jhs@math.gcsu.edu>
475 * regression.q (regression_trns_proc): New function.
477 * regression.q (subcommand_save): Create variable residuals and
478 add a transformation to assign values to them. Also free the
479 linreg_cache if the SAVE command was not called. Removed the
480 casereading loop. Placed actual computation of residuals in
481 regression_trns_proc.
483 * regression.q (run_regression): Moved call to free
484 pspp_linreg_cache to subcommand_save.
486 Sat Apr 15 18:01:03 2006 Ben Pfaff <blp@gnu.org>
488 * examine.q (output_examine): Add casts to fix warnings.
490 2006-04-07 Jason Stover <jhs@math.gcsu.edu>
492 * regression.q (subcommand_save): New function.
494 2006-04-04 Jason Stover <jhs@math.gcsu.edu>
496 * regression.q: New function reg_has_categorical () to tell
497 whether a pspp_linreg_struct was made with any variables of type
500 * regression.q: (subcommand_export): Call
501 reg_print_categorical_encoding() only if the model uses any
502 categorical variables.
504 Mon Mar 27 16:00:42 2006 Ben Pfaff <blp@gnu.org>
506 * crosstabs.q: (output_pivot_table) Drop spurious space from
509 2006-03-15 Jason Stover <jhs@math.gcsu.edu>
511 * regression.q: Added custom syntax parser for VARIABLES subcommand
513 * regression.q: Moved most instructions for run_regression ()
514 inside the loop over dependent variables.
516 Thu Mar 2 08:40:33 WST 2006 John Darrington <john@darrington.wattle.id.au>
518 * Moved files from src directory