1 2007-07-24 Ben Pfaff <blp@gnu.org>
3 * flip.c (struct flip_pgm): Remove `case_size' member (now
5 (cmd_flip): Pass var_cnt as number of cases instead of case_cnt,
6 to fix bug #20494. Don't assign to `case_size' member, which was
7 unused after assignment.
8 (build_dictionary): When NEWNAMES not used, get the number of
9 variables right, to fix bug #20493.
11 2007-07-10 Jason Stover <jhs@math.gcsu.edu>
13 * glm.q: Initial version of the GLM procedure.
15 2007-06-06 Ben Pfaff <blp@gnu.org>
17 Adapt case sources, sinks, and clients of procedure code to the
20 * aggregate.c: Simplify greatly since everything is more uniform
23 * autorecode.c: Adapt to new procedure code.
27 * descriptives.c: Ditto.
29 * npar-summary.c: Ditto.
30 * frequencies.q: Ditto.
33 * regression.q: Ditto.
34 * sort-cases.c: Ditto.
37 * sort-criteria.c: Rewrite to output a struct case_ordering.
39 * flip.c: Rewrite to be a casereader.
41 * rank.q: Simplify greatly since casereaders are much more
42 flexible than what we had before.
44 2007-05-15 Jason Stover <jhs@math.gcsu.edu>
46 * regression.q (run_regression): Tell the user when the data
47 contain no valid cases.
49 2007-05-08 Jason Stover <jhs@math.gcsu.edu>
51 * regression.q: Partial fix of bug which caused a crash if
52 dependent variable and independent variable were the same.
54 2007-04-16 John Darrington <john@darrington.wattle.id.au>
56 * t-test.q: Changed the output width of reported counts and
57 degrees of freedom, to avoid truncating these values. Thanks
58 to Seth Woolley for reporting this problem. A proper fix involves
59 re-thinking the output driver.
61 2007-04-12 Jason Stover <jhs@math.gcsu.edu>
63 * regression.q (run_regression): Added if (n_data >0) to fix bug
66 2007-03-29 Jason Stover <jhs@math.gcsu.edu>
68 * regression.q (prepare_data): New function.
70 * regression.q (compute_moments): New function.
72 2007-03-18 Ben Pfaff <blp@gnu.org>
74 * crosstabs.q (static var write): Rename write_style to avoid
75 conflict with POSIX function of same name.
77 2007-03-16 Jason Stover <jhs@math.gcsu.edu>
79 * regression.q (run_regression): Added support for moments.
81 Sat Feb 17 08:16:00 2007 Ben Pfaff <blp@gnu.org>
83 * flip.c (flip_sink_create): Improve error message when temporary
84 file cannot be created.
86 Tue Feb 6 19:58:03 2007 Ben Pfaff <blp@gnu.org>
88 * flip.c (flip_file): Give better error message on end-of-file.
90 2007-02-04 Jason Stover <jhs@math.gcsu.edu>
92 * regression.q: Fixed p-value computation in the test for
93 individual regression coefficients.
95 Mon Jan 15 11:03:20 2007 Ben Pfaff <blp@gnu.org>
97 Fix bugs found by valgrind when --enable-debug is used with the
98 new case code. These bugs are hidden when the data set is small
99 enough to find in memory; if a bigger data set that would overflow
100 to disk were used, then data corruption would occur.
102 * chisquare.c (create_freq_hash): Pass free_freq_mutable_hash to
103 hsh_create as free function. Make copy of data put into hash.
105 * oneway.q (free_value): New function.
106 (run_oneway): Use free_value as arg to hsh_create. Make copy of
109 * rank.q (rank_cases): Don't access data in case after we've given
112 Tue Jan 9 19:16:11 2007 Ben Pfaff <blp@gnu.org>
116 * aggregate.c (parse_aggregate_functions) Initialize
119 Fri Dec 22 14:04:09 2006 Ben Pfaff <blp@gnu.org>
121 Simplify missing value handling.
123 * aggregate.c (struct agr_var): Remove `bool include_missing', add
124 `enum mv_class exclude'. Remove `int missing', add `bool
125 saw_missing'. Update users.
127 * descriptives.c (struct dsc_trns): Removed `int
128 include_user_missing', add `enum mv_class exclude'. Update users.
129 (struct dsc_proc): Ditto.
131 * examine.q: (static var value_is_missing): Rename
132 `exclude_values', change type to `enum mv_class'. Update users.
136 Fri Dec 22 19:22:18 WST 2006 John Darrington <john@darrington.wattle.id.au>
138 * frequencies.q : Fixed bug #17420, where the table bounds were overun
139 when /FORMAT=nolabels was given.
141 Wed Dec 20 18:45:31 WST 2006 John Darrington <john@darrington.wattle.id.au>
143 * binomial.c binomial.h : New files. Thanks to Jason Stover
144 for assistance with these.
146 * chisquare.c chisquare.h freq.c freq.h npar-summary.c npar-summary.h
147 npar.h npar.q: New files. Implementing NPAR TESTS.
149 * frequencies.q : Moved structure definitions into freq.[ch]
151 Sat Dec 16 22:26:44 2006 Ben Pfaff <blp@gnu.org>
153 Make it possible to pull cases from the active file with a
154 function call, instead of requiring indirection through a callback
157 * aggregate.c (cmd_aggregate): Take advantage of new procedure
159 (agr_to_active_file): Removed.
160 (presorted_agr_to_sysfile): Removed.
162 * autorecode.c (cmd_autorecode): Take advantage of new procedure
164 (autorecode_proc_func): Removed.
166 * flip.c (struct flip_pgm): New members to allow conformance with
167 new case_source_class interface.
168 (cmd_flip): Adapt to new case_source_class interface.
169 (flip_source_read): Ditto.
170 (flip_source_destroy): Ditto.
172 Sat Dec 16 12:54:27 2006 Ben Pfaff <blp@gnu.org>
174 * rank.q (rank_custom_variables): Allow grouping variables to be
175 strings. Fixes bug #18533. Thanks to John Darrington for review.
177 Sat Dec 9 18:47:51 2006 Ben Pfaff <blp@gnu.org>
179 * regression.q (is_depvar): Compare variable pointers instead of
182 Thu Dec 7 15:26:25 WST 2006 John Darrington <john@darrington.wattle.id.au>
184 * examine.q: Allocated the categorical values for the dependent and
185 independent variables, on the heap. Hence they can be of any width.
187 Wed Dec 6 21:14:26 2006 Ben Pfaff <blp@gnu.org>
189 * regression.q (reg_inserted): Compare variable pointers instead
192 Mon Dec 4 22:33:46 2006 Ben Pfaff <blp@gnu.org>
194 * crosstabs.q (insert_summary): Use var_to_string for labeling.
195 (output_pivot_table) Ditto.
198 * frequencies.q (setup_z_trns): Ditto.
200 (dump_condensed) Ditto.
201 (dump_statistics) Ditto.
204 Sun Nov 5 08:31:42 WST 2006 John Darrington <john@darrington.wattle.id.au>
206 * t-test.q, oneway.q: Changed to use the new casefilter structure.
208 Sat Oct 14 16:52:28 2006 Ben Pfaff <blp@gnu.org>
210 * rank.q: (rank_sorted_casefile) Add some missing case_destroy()
211 calls to fix a memory leak.
213 Sun Oct 8 09:45:40 WST 2006 John Darrington <john@darrington.wattle.id.au>
215 * rank.q: Plugged a small memory leak which occurred under error
218 Sat Oct 7 11:06:01 WST 2006 John Darrington <john@darrington.wattle.id.au>
220 * rank.q: Implemented most of the RANK command.
222 2006-07-14 Jason Stover <jhs@math.gcsu.edu>
224 * regression.q (run_regression): New function to move knowledge of
225 pspp_linreg_cache out of math/coefficient.[ch].
227 Sat Jul 1 17:41:46 2006 Ben Pfaff <blp@gnu.org>
229 Fix bug #11612, "q2c documentation does not agree with code".
231 * examine.q: Audit use of q2c "+" prefixes that indicate that a
232 command may appear multiple times.
234 * frequencies.q: Ditto.
238 * regression.q: Ditto.
242 Fri Jun 23 14:18:22 2006 Ben Pfaff <blp@gnu.org>
244 Support long string variables on FREQUENCIES, as
245 an extension when in enhanced algorithms mode. For Greg Hunt
246 <greg@firmansyah.com>.
248 * frequencies.q: (struct freq) Change `v' member from union value
249 to union value *. Update all references.
250 (struct var_freqs) Add width, print members to represent effective
251 variable width and display format.
252 (calc) Copy entire long string value into the hash table.
253 (frq_custom_variables) Set new width, print members.
254 (hash_value_alpha) Get width from var_freqs.
255 (compare_value_alpha_a) Ditto.
256 (compare_freq_alpha_a) Ditto.
257 (compare_freq_alpha_d) Ditto.
258 (dump_full) Get display format from var_freqs.
259 (dump_condensed) Ditto.
261 Mon Jun 19 22:07:13 2006 Ben Pfaff <blp@gnu.org>
263 * frequencies.q: (dump_full) Only put the first MAX_SHORT_STRING
264 bytes of string variables into the output cells, seeing as we only
266 (dump_condensed) Ditto.
268 Mon Jun 19 21:52:05 2006 Ben Pfaff <blp@gnu.org>
270 Fixes a bug reported by Greg Hunt <greg@firmansyah.com>.
272 * frequencies.q: (hsh_hash_bytes) We only copy the first
273 MAX_SHORT_STRING bytes of string variables, so we must only
274 compare that many bytes, even if the string variable is longer.
275 (compare_value_alpha_a) Ditto.
276 (compare_freq_alpha_a) Ditto.
277 (compare_freq_alpha_d) Ditto.
279 2006-05-11 Jason Stover <jhs@math.gcsu.edu>
281 * regression.q: Adjusted code to account for cache->coeff being a
282 pspp_linreg_coeff **.
284 Sun May 7 18:31:25 2006 Ben Pfaff <blp@gnu.org>
288 * aggregate.c (cmd_aggregate): Free default_dict before replacing
291 Sun May 7 17:09:19 2006 Ben Pfaff <blp@gnu.org>
293 * flip.c (flip_file): Check return value of pool_fclose().
295 Sat May 6 16:00:13 2006 Ben Pfaff <blp@gnu.org>
297 Get rid of `char *c' member in union value, for cleanliness.
299 * aggregate.c: (union agr_argument) New union.
300 (struct agr_var) Change element type of arg[] from union value to
302 (parse_aggregate_functions) Change local variable types likewise.
304 * autorecode.c: (union arc_value) New union.
305 (struct arc_item) Change "from" from union value to union
307 (recode) Change local variable from union value to union
309 (autorecode_trns_proc) Ditto.
310 (compare_alpha_value) Ditto.
311 (hash_alpha_value) Ditto.
312 (compare_numeric_value) Ditto.
313 (hash_numeric_value) Ditto.
314 (autorecode_proc_func) Ditto.
316 Sat May 6 10:43:33 2006 Ben Pfaff <blp@gnu.org>
318 Continue reforming procedure execution. In this phase, get rid of
319 the output code for SPLIT FILE groups in procedure.c, which really
320 shouldn't be doing any output. Move it into the individual
321 procedures instead. This also adds some flexibility.
323 * crosstabs.q (precalc): Call output_split_file_values().
325 * descriptives.c (calc_descriptives): Ditto.
327 * examine.q (run_examine): Ditto.
329 * frequencies.q (precalc): Ditto.
331 * oneway.q (run_oneway): Ditto.
333 * regression.q (run_regression): Ditto.
335 * t-test.q (calculate): Ditto.
337 Wed May 3 23:05:31 2006 Ben Pfaff <blp@gnu.org>
339 Continue reforming procedure execution. In this phase, get rid of
340 many global variables, consolidating procedure execution in
341 procedure.c. Encapsulate transformations in new "struct
342 trns_chain". Also, change implementation of N OF CASES, FILTER,
343 and PROCESS IF from special cases to transformations.
345 * aggregate.c (cmd_aggregate) Use discard_variables().
347 2006-04-28 Jason Stover <jhs@math.gcsu.edu>
349 * regression.q (regression_trns_resid_proc): Pass only the
350 variables used in the model to (*model->residual)().
352 * regression.q (regression_trns_pred_proc): Pass only the
353 variables used in the model to (*model->pred)().
355 2006-04-26 Jason Stover <jhs@math.gcsu.edu>
357 * regression.q: Added support for multiple transformations.
359 * regression.q (regression_trns_resid_proc): New function.
361 * regression.q (regression_trns_pred_proc): New function.
363 * regression.q (subcommand_save): Added support for saving
366 * regression.q (regression_trns_free): New function.
368 * regression.q (reg_get_name): New function.
370 * regression.q (reg_save_var): New function.
372 Tue Apr 25 13:18:56 2006 Ben Pfaff <blp@gnu.org>
374 * rank.q (parse_rank_function): Use SE instead of ME for parse
377 Tue Apr 25 13:16:28 2006 Ben Pfaff <blp@gnu.org>
379 * flip.c (flip_sink_write): Use snprintf() to simplify a bit of
382 2006-04-21 Jason Stover <jhs@math.gcsu.edu>
384 * regression.q (try_name): New function. (Partly copied from
385 try_name in descriptives.c.)
387 * regression.q (subcommand_save): Choose residual variable names
390 2006-04-20 Jason Stover <jhs@math.gcsu.edu>
392 * regression.q (cmd_regression): Moved call to subcommand_save()
393 outside multipass_procedure_with_splits().
395 * regression.q (regression_trns_proc): Fixed value counter n_vals
396 before calling *model->residual().
398 2006-04-19 Jason Stover <jhs@math.gcsu.edu>
400 * regression.q (regression_trns_proc): Fixed the look-up of the
403 2006-04-18 Jason Stover <jhs@math.gcsu.edu>
405 * regression.q (regression_trns_proc): Look up the residual
406 variable in the linear regression cache.
408 * regression.q (subcommand_save): Set the residual variable in the
409 linear regression cache.
411 2006-04-17 Jason Stover <jason@wonko.gcsu.edu>
413 * regression.q (regression_trns_proc): Accept case_idx as an int
414 to match the definition of trns_proc_func.
416 2006-04-17 Jason Stover <jhs@math.gcsu.edu>
418 * regression.q (regression_trns_proc): New function.
420 * regression.q (subcommand_save): Create variable residuals and
421 add a transformation to assign values to them. Also free the
422 linreg_cache if the SAVE command was not called. Removed the
423 casereading loop. Placed actual computation of residuals in
424 regression_trns_proc.
426 * regression.q (run_regression): Moved call to free
427 pspp_linreg_cache to subcommand_save.
429 Sat Apr 15 18:01:03 2006 Ben Pfaff <blp@gnu.org>
431 * examine.q (output_examine): Add casts to fix warnings.
433 2006-04-07 Jason Stover <jhs@math.gcsu.edu>
435 * regression.q (subcommand_save): New function.
437 2006-04-04 Jason Stover <jhs@math.gcsu.edu>
439 * regression.q: New function reg_has_categorical () to tell
440 whether a pspp_linreg_struct was made with any variables of type
443 * regression.q: (subcommand_export): Call
444 reg_print_categorical_encoding() only if the model uses any
445 categorical variables.
447 Mon Mar 27 16:00:42 2006 Ben Pfaff <blp@gnu.org>
449 * crosstabs.q: (output_pivot_table) Drop spurious space from
452 2006-03-15 Jason Stover <jhs@math.gcsu.edu>
454 * regression.q: Added custom syntax parser for VARIABLES subcommand
456 * regression.q: Moved most instructions for run_regression ()
457 inside the loop over dependent variables.
459 Thu Mar 2 08:40:33 WST 2006 John Darrington <john@darrington.wattle.id.au>
461 * Moved files from src directory