1 2007-05-15 Jason Stover <jhs@math.gcsu.edu>
3 * regression.q (run_regression): Tell the user when the data
4 contain no valid cases.
6 2007-05-08 Jason Stover <jhs@math.gcsu.edu>
8 * regression.q: Partial fix of bug which caused a crash if
9 dependent variable and independent variable were the same.
11 2007-04-16 John Darrington <john@darrington.wattle.id.au>
13 * t-test.q: Changed the output width of reported counts and
14 degrees of freedom, to avoid truncating these values. Thanks
15 to Seth Woolley for reporting this problem. A proper fix involves
16 re-thinking the output driver.
18 2007-04-12 Jason Stover <jhs@math.gcsu.edu>
20 * regression.q (run_regression): Added if (n_data >0) to fix bug
23 2007-03-29 Jason Stover <jhs@math.gcsu.edu>
25 * regression.q (prepare_data): New function.
27 * regression.q (compute_moments): New function.
29 2007-03-18 Ben Pfaff <blp@gnu.org>
31 * crosstabs.q (static var write): Rename write_style to avoid
32 conflict with POSIX function of same name.
34 2007-03-16 Jason Stover <jhs@math.gcsu.edu>
36 * regression.q (run_regression): Added support for moments.
38 Sat Feb 17 08:16:00 2007 Ben Pfaff <blp@gnu.org>
40 * flip.c (flip_sink_create): Improve error message when temporary
41 file cannot be created.
43 Tue Feb 6 19:58:03 2007 Ben Pfaff <blp@gnu.org>
45 * flip.c (flip_file): Give better error message on end-of-file.
47 2007-02-04 Jason Stover <jhs@math.gcsu.edu>
49 * regression.q: Fixed p-value computation in the test for
50 individual regression coefficients.
52 Mon Jan 15 11:03:20 2007 Ben Pfaff <blp@gnu.org>
54 Fix bugs found by valgrind when --enable-debug is used with the
55 new case code. These bugs are hidden when the data set is small
56 enough to find in memory; if a bigger data set that would overflow
57 to disk were used, then data corruption would occur.
59 * chisquare.c (create_freq_hash): Pass free_freq_mutable_hash to
60 hsh_create as free function. Make copy of data put into hash.
62 * oneway.q (free_value): New function.
63 (run_oneway): Use free_value as arg to hsh_create. Make copy of
66 * rank.q (rank_cases): Don't access data in case after we've given
69 Tue Jan 9 19:16:11 2007 Ben Pfaff <blp@gnu.org>
73 * aggregate.c (parse_aggregate_functions) Initialize
76 Fri Dec 22 14:04:09 2006 Ben Pfaff <blp@gnu.org>
78 Simplify missing value handling.
80 * aggregate.c (struct agr_var): Remove `bool include_missing', add
81 `enum mv_class exclude'. Remove `int missing', add `bool
82 saw_missing'. Update users.
84 * descriptives.c (struct dsc_trns): Removed `int
85 include_user_missing', add `enum mv_class exclude'. Update users.
86 (struct dsc_proc): Ditto.
88 * examine.q: (static var value_is_missing): Rename
89 `exclude_values', change type to `enum mv_class'. Update users.
93 Fri Dec 22 19:22:18 WST 2006 John Darrington <john@darrington.wattle.id.au>
95 * frequencies.q : Fixed bug #17420, where the table bounds were overun
96 when /FORMAT=nolabels was given.
98 Wed Dec 20 18:45:31 WST 2006 John Darrington <john@darrington.wattle.id.au>
100 * binomial.c binomial.h : New files. Thanks to Jason Stover
101 for assistance with these.
103 * chisquare.c chisquare.h freq.c freq.h npar-summary.c npar-summary.h
104 npar.h npar.q: New files. Implementing NPAR TESTS.
106 * frequencies.q : Moved structure definitions into freq.[ch]
108 Sat Dec 16 22:26:44 2006 Ben Pfaff <blp@gnu.org>
110 Make it possible to pull cases from the active file with a
111 function call, instead of requiring indirection through a callback
114 * aggregate.c (cmd_aggregate): Take advantage of new procedure
116 (agr_to_active_file): Removed.
117 (presorted_agr_to_sysfile): Removed.
119 * autorecode.c (cmd_autorecode): Take advantage of new procedure
121 (autorecode_proc_func): Removed.
123 * flip.c (struct flip_pgm): New members to allow conformance with
124 new case_source_class interface.
125 (cmd_flip): Adapt to new case_source_class interface.
126 (flip_source_read): Ditto.
127 (flip_source_destroy): Ditto.
129 Sat Dec 16 12:54:27 2006 Ben Pfaff <blp@gnu.org>
131 * rank.q (rank_custom_variables): Allow grouping variables to be
132 strings. Fixes bug #18533. Thanks to John Darrington for review.
134 Sat Dec 9 18:47:51 2006 Ben Pfaff <blp@gnu.org>
136 * regression.q (is_depvar): Compare variable pointers instead of
139 Thu Dec 7 15:26:25 WST 2006 John Darrington <john@darrington.wattle.id.au>
141 * examine.q: Allocated the categorical values for the dependent and
142 independent variables, on the heap. Hence they can be of any width.
144 Wed Dec 6 21:14:26 2006 Ben Pfaff <blp@gnu.org>
146 * regression.q (reg_inserted): Compare variable pointers instead
149 Mon Dec 4 22:33:46 2006 Ben Pfaff <blp@gnu.org>
151 * crosstabs.q (insert_summary): Use var_to_string for labeling.
152 (output_pivot_table) Ditto.
155 * frequencies.q (setup_z_trns): Ditto.
157 (dump_condensed) Ditto.
158 (dump_statistics) Ditto.
161 Sun Nov 5 08:31:42 WST 2006 John Darrington <john@darrington.wattle.id.au>
163 * t-test.q, oneway.q: Changed to use the new casefilter structure.
165 Sat Oct 14 16:52:28 2006 Ben Pfaff <blp@gnu.org>
167 * rank.q: (rank_sorted_casefile) Add some missing case_destroy()
168 calls to fix a memory leak.
170 Sun Oct 8 09:45:40 WST 2006 John Darrington <john@darrington.wattle.id.au>
172 * rank.q: Plugged a small memory leak which occurred under error
175 Sat Oct 7 11:06:01 WST 2006 John Darrington <john@darrington.wattle.id.au>
177 * rank.q: Implemented most of the RANK command.
179 2006-07-14 Jason Stover <jhs@math.gcsu.edu>
181 * regression.q (run_regression): New function to move knowledge of
182 pspp_linreg_cache out of math/coefficient.[ch].
184 Sat Jul 1 17:41:46 2006 Ben Pfaff <blp@gnu.org>
186 Fix bug #11612, "q2c documentation does not agree with code".
188 * examine.q: Audit use of q2c "+" prefixes that indicate that a
189 command may appear multiple times.
191 * frequencies.q: Ditto.
195 * regression.q: Ditto.
199 Fri Jun 23 14:18:22 2006 Ben Pfaff <blp@gnu.org>
201 Support long string variables on FREQUENCIES, as
202 an extension when in enhanced algorithms mode. For Greg Hunt
203 <greg@firmansyah.com>.
205 * frequencies.q: (struct freq) Change `v' member from union value
206 to union value *. Update all references.
207 (struct var_freqs) Add width, print members to represent effective
208 variable width and display format.
209 (calc) Copy entire long string value into the hash table.
210 (frq_custom_variables) Set new width, print members.
211 (hash_value_alpha) Get width from var_freqs.
212 (compare_value_alpha_a) Ditto.
213 (compare_freq_alpha_a) Ditto.
214 (compare_freq_alpha_d) Ditto.
215 (dump_full) Get display format from var_freqs.
216 (dump_condensed) Ditto.
218 Mon Jun 19 22:07:13 2006 Ben Pfaff <blp@gnu.org>
220 * frequencies.q: (dump_full) Only put the first MAX_SHORT_STRING
221 bytes of string variables into the output cells, seeing as we only
223 (dump_condensed) Ditto.
225 Mon Jun 19 21:52:05 2006 Ben Pfaff <blp@gnu.org>
227 Fixes a bug reported by Greg Hunt <greg@firmansyah.com>.
229 * frequencies.q: (hsh_hash_bytes) We only copy the first
230 MAX_SHORT_STRING bytes of string variables, so we must only
231 compare that many bytes, even if the string variable is longer.
232 (compare_value_alpha_a) Ditto.
233 (compare_freq_alpha_a) Ditto.
234 (compare_freq_alpha_d) Ditto.
236 2006-05-11 Jason Stover <jhs@math.gcsu.edu>
238 * regression.q: Adjusted code to account for cache->coeff being a
239 pspp_linreg_coeff **.
241 Sun May 7 18:31:25 2006 Ben Pfaff <blp@gnu.org>
245 * aggregate.c (cmd_aggregate): Free default_dict before replacing
248 Sun May 7 17:09:19 2006 Ben Pfaff <blp@gnu.org>
250 * flip.c (flip_file): Check return value of pool_fclose().
252 Sat May 6 16:00:13 2006 Ben Pfaff <blp@gnu.org>
254 Get rid of `char *c' member in union value, for cleanliness.
256 * aggregate.c: (union agr_argument) New union.
257 (struct agr_var) Change element type of arg[] from union value to
259 (parse_aggregate_functions) Change local variable types likewise.
261 * autorecode.c: (union arc_value) New union.
262 (struct arc_item) Change "from" from union value to union
264 (recode) Change local variable from union value to union
266 (autorecode_trns_proc) Ditto.
267 (compare_alpha_value) Ditto.
268 (hash_alpha_value) Ditto.
269 (compare_numeric_value) Ditto.
270 (hash_numeric_value) Ditto.
271 (autorecode_proc_func) Ditto.
273 Sat May 6 10:43:33 2006 Ben Pfaff <blp@gnu.org>
275 Continue reforming procedure execution. In this phase, get rid of
276 the output code for SPLIT FILE groups in procedure.c, which really
277 shouldn't be doing any output. Move it into the individual
278 procedures instead. This also adds some flexibility.
280 * crosstabs.q (precalc): Call output_split_file_values().
282 * descriptives.c (calc_descriptives): Ditto.
284 * examine.q (run_examine): Ditto.
286 * frequencies.q (precalc): Ditto.
288 * oneway.q (run_oneway): Ditto.
290 * regression.q (run_regression): Ditto.
292 * t-test.q (calculate): Ditto.
294 Wed May 3 23:05:31 2006 Ben Pfaff <blp@gnu.org>
296 Continue reforming procedure execution. In this phase, get rid of
297 many global variables, consolidating procedure execution in
298 procedure.c. Encapsulate transformations in new "struct
299 trns_chain". Also, change implementation of N OF CASES, FILTER,
300 and PROCESS IF from special cases to transformations.
302 * aggregate.c (cmd_aggregate) Use discard_variables().
304 2006-04-28 Jason Stover <jhs@math.gcsu.edu>
306 * regression.q (regression_trns_resid_proc): Pass only the
307 variables used in the model to (*model->residual)().
309 * regression.q (regression_trns_pred_proc): Pass only the
310 variables used in the model to (*model->pred)().
312 2006-04-26 Jason Stover <jhs@math.gcsu.edu>
314 * regression.q: Added support for multiple transformations.
316 * regression.q (regression_trns_resid_proc): New function.
318 * regression.q (regression_trns_pred_proc): New function.
320 * regression.q (subcommand_save): Added support for saving
323 * regression.q (regression_trns_free): New function.
325 * regression.q (reg_get_name): New function.
327 * regression.q (reg_save_var): New function.
329 Tue Apr 25 13:18:56 2006 Ben Pfaff <blp@gnu.org>
331 * rank.q (parse_rank_function): Use SE instead of ME for parse
334 Tue Apr 25 13:16:28 2006 Ben Pfaff <blp@gnu.org>
336 * flip.c (flip_sink_write): Use snprintf() to simplify a bit of
339 2006-04-21 Jason Stover <jhs@math.gcsu.edu>
341 * regression.q (try_name): New function. (Partly copied from
342 try_name in descriptives.c.)
344 * regression.q (subcommand_save): Choose residual variable names
347 2006-04-20 Jason Stover <jhs@math.gcsu.edu>
349 * regression.q (cmd_regression): Moved call to subcommand_save()
350 outside multipass_procedure_with_splits().
352 * regression.q (regression_trns_proc): Fixed value counter n_vals
353 before calling *model->residual().
355 2006-04-19 Jason Stover <jhs@math.gcsu.edu>
357 * regression.q (regression_trns_proc): Fixed the look-up of the
360 2006-04-18 Jason Stover <jhs@math.gcsu.edu>
362 * regression.q (regression_trns_proc): Look up the residual
363 variable in the linear regression cache.
365 * regression.q (subcommand_save): Set the residual variable in the
366 linear regression cache.
368 2006-04-17 Jason Stover <jason@wonko.gcsu.edu>
370 * regression.q (regression_trns_proc): Accept case_idx as an int
371 to match the definition of trns_proc_func.
373 2006-04-17 Jason Stover <jhs@math.gcsu.edu>
375 * regression.q (regression_trns_proc): New function.
377 * regression.q (subcommand_save): Create variable residuals and
378 add a transformation to assign values to them. Also free the
379 linreg_cache if the SAVE command was not called. Removed the
380 casereading loop. Placed actual computation of residuals in
381 regression_trns_proc.
383 * regression.q (run_regression): Moved call to free
384 pspp_linreg_cache to subcommand_save.
386 Sat Apr 15 18:01:03 2006 Ben Pfaff <blp@gnu.org>
388 * examine.q (output_examine): Add casts to fix warnings.
390 2006-04-07 Jason Stover <jhs@math.gcsu.edu>
392 * regression.q (subcommand_save): New function.
394 2006-04-04 Jason Stover <jhs@math.gcsu.edu>
396 * regression.q: New function reg_has_categorical () to tell
397 whether a pspp_linreg_struct was made with any variables of type
400 * regression.q: (subcommand_export): Call
401 reg_print_categorical_encoding() only if the model uses any
402 categorical variables.
404 Mon Mar 27 16:00:42 2006 Ben Pfaff <blp@gnu.org>
406 * crosstabs.q: (output_pivot_table) Drop spurious space from
409 2006-03-15 Jason Stover <jhs@math.gcsu.edu>
411 * regression.q: Added custom syntax parser for VARIABLES subcommand
413 * regression.q: Moved most instructions for run_regression ()
414 inside the loop over dependent variables.
416 Thu Mar 2 08:40:33 WST 2006 John Darrington <john@darrington.wattle.id.au>
418 * Moved files from src directory