1 2007-08-12 Ben Pfaff <blp@gnu.org>
3 * flip.c: Drop use of dict_get_compacted_dict_index_to_case_index
4 and just use the ordinary case indexes. There seemed to be no
5 reason for the former method.
7 2007-08-03 Ben Pfaff <blp@gnu.org>
9 * rank.q (rank_cmd): Instead of sorting by SPLIT FILE vars, group
10 by them. Fixes bug #17239.
11 Reviewed by John Darrington.
13 2007-08-01 Ben Pfaff <blp@gnu.org>
15 Clean up handling of median, by treating it almost like any other
16 percentile. Fixes bug #17424. Thanks to John Darrington for
18 * frequencies.q (internal_cmd_frequencies): Fix handling of bit
19 masks for `stats' variable. If median is selected, turn it off
20 and add a 50th percentile.
21 (add_percentile): Simplify code a little.
22 (calc_stats): Drop special casing of median.
23 (dump_statistics): Ditto, except that we label the 50th percentile
24 as "50 (Median)" to make it clear that it's also the median.
26 2007-07-31 Ben Pfaff <blp@gnu.org>
28 Remove integer mode from FREQUENCIES and incidentally fix bug
29 #17421. Reviewed by John Darrington.
30 * frequencies.q (int_pool): Rename data_pool.
31 (gen_pool): Rename syntax_pool.
32 (enum FRQM_*): Removed.
33 (struct freq_tab): Removed `mode', `vector', `min', `max',
34 `out_of_range', `sysmis' members.
35 (calc): Delete support for integer mode.
37 (postprocess_freq_tab): Ditto.
38 (cleanup_freq_tab): Ditto.
39 (frq_custom_variables): Ditto.
41 2007-07-28 John Darrington <john@darrington.wattle.id.au>
43 * t-test.q: Moved the order in which groups are displayed in the
44 independent samples case, where a cut point is given.
46 2007-07-27 Ben Pfaff <blp@gnu.org>
48 * regression.q (run_regression): Move casereader_destroy call so
49 that it always gets called, not just if there was some valid
50 data. Fixes bug #19581.
51 Reviewed by Jason Stover.
53 2007-07-24 Ben Pfaff <blp@gnu.org>
55 * flip.c (struct flip_pgm): Remove `case_size' member (now
57 (cmd_flip): Pass var_cnt as number of cases instead of case_cnt,
58 to fix bug #20494. Don't assign to `case_size' member, which was
59 unused after assignment.
60 (build_dictionary): When NEWNAMES not used, get the number of
61 variables right, to fix bug #20493.
63 2007-07-10 Jason Stover <jhs@math.gcsu.edu>
65 * glm.q: Initial version of the GLM procedure.
67 2007-06-06 Ben Pfaff <blp@gnu.org>
69 Adapt case sources, sinks, and clients of procedure code to the
72 * aggregate.c: Simplify greatly since everything is more uniform
75 * autorecode.c: Adapt to new procedure code.
79 * descriptives.c: Ditto.
81 * npar-summary.c: Ditto.
82 * frequencies.q: Ditto.
85 * regression.q: Ditto.
86 * sort-cases.c: Ditto.
89 * sort-criteria.c: Rewrite to output a struct case_ordering.
91 * flip.c: Rewrite to be a casereader.
93 * rank.q: Simplify greatly since casereaders are much more
94 flexible than what we had before.
96 2007-05-15 Jason Stover <jhs@math.gcsu.edu>
98 * regression.q (run_regression): Tell the user when the data
99 contain no valid cases.
101 2007-05-08 Jason Stover <jhs@math.gcsu.edu>
103 * regression.q: Partial fix of bug which caused a crash if
104 dependent variable and independent variable were the same.
106 2007-04-16 John Darrington <john@darrington.wattle.id.au>
108 * t-test.q: Changed the output width of reported counts and
109 degrees of freedom, to avoid truncating these values. Thanks
110 to Seth Woolley for reporting this problem. A proper fix involves
111 re-thinking the output driver.
113 2007-04-12 Jason Stover <jhs@math.gcsu.edu>
115 * regression.q (run_regression): Added if (n_data >0) to fix bug
118 2007-03-29 Jason Stover <jhs@math.gcsu.edu>
120 * regression.q (prepare_data): New function.
122 * regression.q (compute_moments): New function.
124 2007-03-18 Ben Pfaff <blp@gnu.org>
126 * crosstabs.q (static var write): Rename write_style to avoid
127 conflict with POSIX function of same name.
129 2007-03-16 Jason Stover <jhs@math.gcsu.edu>
131 * regression.q (run_regression): Added support for moments.
133 Sat Feb 17 08:16:00 2007 Ben Pfaff <blp@gnu.org>
135 * flip.c (flip_sink_create): Improve error message when temporary
136 file cannot be created.
138 Tue Feb 6 19:58:03 2007 Ben Pfaff <blp@gnu.org>
140 * flip.c (flip_file): Give better error message on end-of-file.
142 2007-02-04 Jason Stover <jhs@math.gcsu.edu>
144 * regression.q: Fixed p-value computation in the test for
145 individual regression coefficients.
147 Mon Jan 15 11:03:20 2007 Ben Pfaff <blp@gnu.org>
149 Fix bugs found by valgrind when --enable-debug is used with the
150 new case code. These bugs are hidden when the data set is small
151 enough to find in memory; if a bigger data set that would overflow
152 to disk were used, then data corruption would occur.
154 * chisquare.c (create_freq_hash): Pass free_freq_mutable_hash to
155 hsh_create as free function. Make copy of data put into hash.
157 * oneway.q (free_value): New function.
158 (run_oneway): Use free_value as arg to hsh_create. Make copy of
161 * rank.q (rank_cases): Don't access data in case after we've given
164 Tue Jan 9 19:16:11 2007 Ben Pfaff <blp@gnu.org>
168 * aggregate.c (parse_aggregate_functions) Initialize
171 Fri Dec 22 14:04:09 2006 Ben Pfaff <blp@gnu.org>
173 Simplify missing value handling.
175 * aggregate.c (struct agr_var): Remove `bool include_missing', add
176 `enum mv_class exclude'. Remove `int missing', add `bool
177 saw_missing'. Update users.
179 * descriptives.c (struct dsc_trns): Removed `int
180 include_user_missing', add `enum mv_class exclude'. Update users.
181 (struct dsc_proc): Ditto.
183 * examine.q: (static var value_is_missing): Rename
184 `exclude_values', change type to `enum mv_class'. Update users.
188 Fri Dec 22 19:22:18 WST 2006 John Darrington <john@darrington.wattle.id.au>
190 * frequencies.q : Fixed bug #17420, where the table bounds were overun
191 when /FORMAT=nolabels was given.
193 Wed Dec 20 18:45:31 WST 2006 John Darrington <john@darrington.wattle.id.au>
195 * binomial.c binomial.h : New files. Thanks to Jason Stover
196 for assistance with these.
198 * chisquare.c chisquare.h freq.c freq.h npar-summary.c npar-summary.h
199 npar.h npar.q: New files. Implementing NPAR TESTS.
201 * frequencies.q : Moved structure definitions into freq.[ch]
203 Sat Dec 16 22:26:44 2006 Ben Pfaff <blp@gnu.org>
205 Make it possible to pull cases from the active file with a
206 function call, instead of requiring indirection through a callback
209 * aggregate.c (cmd_aggregate): Take advantage of new procedure
211 (agr_to_active_file): Removed.
212 (presorted_agr_to_sysfile): Removed.
214 * autorecode.c (cmd_autorecode): Take advantage of new procedure
216 (autorecode_proc_func): Removed.
218 * flip.c (struct flip_pgm): New members to allow conformance with
219 new case_source_class interface.
220 (cmd_flip): Adapt to new case_source_class interface.
221 (flip_source_read): Ditto.
222 (flip_source_destroy): Ditto.
224 Sat Dec 16 12:54:27 2006 Ben Pfaff <blp@gnu.org>
226 * rank.q (rank_custom_variables): Allow grouping variables to be
227 strings. Fixes bug #18533. Thanks to John Darrington for review.
229 Sat Dec 9 18:47:51 2006 Ben Pfaff <blp@gnu.org>
231 * regression.q (is_depvar): Compare variable pointers instead of
234 Thu Dec 7 15:26:25 WST 2006 John Darrington <john@darrington.wattle.id.au>
236 * examine.q: Allocated the categorical values for the dependent and
237 independent variables, on the heap. Hence they can be of any width.
239 Wed Dec 6 21:14:26 2006 Ben Pfaff <blp@gnu.org>
241 * regression.q (reg_inserted): Compare variable pointers instead
244 Mon Dec 4 22:33:46 2006 Ben Pfaff <blp@gnu.org>
246 * crosstabs.q (insert_summary): Use var_to_string for labeling.
247 (output_pivot_table) Ditto.
250 * frequencies.q (setup_z_trns): Ditto.
252 (dump_condensed) Ditto.
253 (dump_statistics) Ditto.
256 Sun Nov 5 08:31:42 WST 2006 John Darrington <john@darrington.wattle.id.au>
258 * t-test.q, oneway.q: Changed to use the new casefilter structure.
260 Sat Oct 14 16:52:28 2006 Ben Pfaff <blp@gnu.org>
262 * rank.q: (rank_sorted_casefile) Add some missing case_destroy()
263 calls to fix a memory leak.
265 Sun Oct 8 09:45:40 WST 2006 John Darrington <john@darrington.wattle.id.au>
267 * rank.q: Plugged a small memory leak which occurred under error
270 Sat Oct 7 11:06:01 WST 2006 John Darrington <john@darrington.wattle.id.au>
272 * rank.q: Implemented most of the RANK command.
274 2006-07-14 Jason Stover <jhs@math.gcsu.edu>
276 * regression.q (run_regression): New function to move knowledge of
277 pspp_linreg_cache out of math/coefficient.[ch].
279 Sat Jul 1 17:41:46 2006 Ben Pfaff <blp@gnu.org>
281 Fix bug #11612, "q2c documentation does not agree with code".
283 * examine.q: Audit use of q2c "+" prefixes that indicate that a
284 command may appear multiple times.
286 * frequencies.q: Ditto.
290 * regression.q: Ditto.
294 Fri Jun 23 14:18:22 2006 Ben Pfaff <blp@gnu.org>
296 Support long string variables on FREQUENCIES, as
297 an extension when in enhanced algorithms mode. For Greg Hunt
298 <greg@firmansyah.com>.
300 * frequencies.q: (struct freq) Change `v' member from union value
301 to union value *. Update all references.
302 (struct var_freqs) Add width, print members to represent effective
303 variable width and display format.
304 (calc) Copy entire long string value into the hash table.
305 (frq_custom_variables) Set new width, print members.
306 (hash_value_alpha) Get width from var_freqs.
307 (compare_value_alpha_a) Ditto.
308 (compare_freq_alpha_a) Ditto.
309 (compare_freq_alpha_d) Ditto.
310 (dump_full) Get display format from var_freqs.
311 (dump_condensed) Ditto.
313 Mon Jun 19 22:07:13 2006 Ben Pfaff <blp@gnu.org>
315 * frequencies.q: (dump_full) Only put the first MAX_SHORT_STRING
316 bytes of string variables into the output cells, seeing as we only
318 (dump_condensed) Ditto.
320 Mon Jun 19 21:52:05 2006 Ben Pfaff <blp@gnu.org>
322 Fixes a bug reported by Greg Hunt <greg@firmansyah.com>.
324 * frequencies.q: (hsh_hash_bytes) We only copy the first
325 MAX_SHORT_STRING bytes of string variables, so we must only
326 compare that many bytes, even if the string variable is longer.
327 (compare_value_alpha_a) Ditto.
328 (compare_freq_alpha_a) Ditto.
329 (compare_freq_alpha_d) Ditto.
331 2006-05-11 Jason Stover <jhs@math.gcsu.edu>
333 * regression.q: Adjusted code to account for cache->coeff being a
334 pspp_linreg_coeff **.
336 Sun May 7 18:31:25 2006 Ben Pfaff <blp@gnu.org>
340 * aggregate.c (cmd_aggregate): Free default_dict before replacing
343 Sun May 7 17:09:19 2006 Ben Pfaff <blp@gnu.org>
345 * flip.c (flip_file): Check return value of pool_fclose().
347 Sat May 6 16:00:13 2006 Ben Pfaff <blp@gnu.org>
349 Get rid of `char *c' member in union value, for cleanliness.
351 * aggregate.c: (union agr_argument) New union.
352 (struct agr_var) Change element type of arg[] from union value to
354 (parse_aggregate_functions) Change local variable types likewise.
356 * autorecode.c: (union arc_value) New union.
357 (struct arc_item) Change "from" from union value to union
359 (recode) Change local variable from union value to union
361 (autorecode_trns_proc) Ditto.
362 (compare_alpha_value) Ditto.
363 (hash_alpha_value) Ditto.
364 (compare_numeric_value) Ditto.
365 (hash_numeric_value) Ditto.
366 (autorecode_proc_func) Ditto.
368 Sat May 6 10:43:33 2006 Ben Pfaff <blp@gnu.org>
370 Continue reforming procedure execution. In this phase, get rid of
371 the output code for SPLIT FILE groups in procedure.c, which really
372 shouldn't be doing any output. Move it into the individual
373 procedures instead. This also adds some flexibility.
375 * crosstabs.q (precalc): Call output_split_file_values().
377 * descriptives.c (calc_descriptives): Ditto.
379 * examine.q (run_examine): Ditto.
381 * frequencies.q (precalc): Ditto.
383 * oneway.q (run_oneway): Ditto.
385 * regression.q (run_regression): Ditto.
387 * t-test.q (calculate): Ditto.
389 Wed May 3 23:05:31 2006 Ben Pfaff <blp@gnu.org>
391 Continue reforming procedure execution. In this phase, get rid of
392 many global variables, consolidating procedure execution in
393 procedure.c. Encapsulate transformations in new "struct
394 trns_chain". Also, change implementation of N OF CASES, FILTER,
395 and PROCESS IF from special cases to transformations.
397 * aggregate.c (cmd_aggregate) Use discard_variables().
399 2006-04-28 Jason Stover <jhs@math.gcsu.edu>
401 * regression.q (regression_trns_resid_proc): Pass only the
402 variables used in the model to (*model->residual)().
404 * regression.q (regression_trns_pred_proc): Pass only the
405 variables used in the model to (*model->pred)().
407 2006-04-26 Jason Stover <jhs@math.gcsu.edu>
409 * regression.q: Added support for multiple transformations.
411 * regression.q (regression_trns_resid_proc): New function.
413 * regression.q (regression_trns_pred_proc): New function.
415 * regression.q (subcommand_save): Added support for saving
418 * regression.q (regression_trns_free): New function.
420 * regression.q (reg_get_name): New function.
422 * regression.q (reg_save_var): New function.
424 Tue Apr 25 13:18:56 2006 Ben Pfaff <blp@gnu.org>
426 * rank.q (parse_rank_function): Use SE instead of ME for parse
429 Tue Apr 25 13:16:28 2006 Ben Pfaff <blp@gnu.org>
431 * flip.c (flip_sink_write): Use snprintf() to simplify a bit of
434 2006-04-21 Jason Stover <jhs@math.gcsu.edu>
436 * regression.q (try_name): New function. (Partly copied from
437 try_name in descriptives.c.)
439 * regression.q (subcommand_save): Choose residual variable names
442 2006-04-20 Jason Stover <jhs@math.gcsu.edu>
444 * regression.q (cmd_regression): Moved call to subcommand_save()
445 outside multipass_procedure_with_splits().
447 * regression.q (regression_trns_proc): Fixed value counter n_vals
448 before calling *model->residual().
450 2006-04-19 Jason Stover <jhs@math.gcsu.edu>
452 * regression.q (regression_trns_proc): Fixed the look-up of the
455 2006-04-18 Jason Stover <jhs@math.gcsu.edu>
457 * regression.q (regression_trns_proc): Look up the residual
458 variable in the linear regression cache.
460 * regression.q (subcommand_save): Set the residual variable in the
461 linear regression cache.
463 2006-04-17 Jason Stover <jason@wonko.gcsu.edu>
465 * regression.q (regression_trns_proc): Accept case_idx as an int
466 to match the definition of trns_proc_func.
468 2006-04-17 Jason Stover <jhs@math.gcsu.edu>
470 * regression.q (regression_trns_proc): New function.
472 * regression.q (subcommand_save): Create variable residuals and
473 add a transformation to assign values to them. Also free the
474 linreg_cache if the SAVE command was not called. Removed the
475 casereading loop. Placed actual computation of residuals in
476 regression_trns_proc.
478 * regression.q (run_regression): Moved call to free
479 pspp_linreg_cache to subcommand_save.
481 Sat Apr 15 18:01:03 2006 Ben Pfaff <blp@gnu.org>
483 * examine.q (output_examine): Add casts to fix warnings.
485 2006-04-07 Jason Stover <jhs@math.gcsu.edu>
487 * regression.q (subcommand_save): New function.
489 2006-04-04 Jason Stover <jhs@math.gcsu.edu>
491 * regression.q: New function reg_has_categorical () to tell
492 whether a pspp_linreg_struct was made with any variables of type
495 * regression.q: (subcommand_export): Call
496 reg_print_categorical_encoding() only if the model uses any
497 categorical variables.
499 Mon Mar 27 16:00:42 2006 Ben Pfaff <blp@gnu.org>
501 * crosstabs.q: (output_pivot_table) Drop spurious space from
504 2006-03-15 Jason Stover <jhs@math.gcsu.edu>
506 * regression.q: Added custom syntax parser for VARIABLES subcommand
508 * regression.q: Moved most instructions for run_regression ()
509 inside the loop over dependent variables.
511 Thu Mar 2 08:40:33 WST 2006 John Darrington <john@darrington.wattle.id.au>
513 * Moved files from src directory