1 2007-02-04 Jason Stover <jhs@math.gcsu.edu>
3 * regression.q: Fixed p-value computation in the test for
4 individual regression coefficients.
6 Mon Jan 15 11:03:20 2007 Ben Pfaff <blp@gnu.org>
8 Fix bugs found by valgrind when --enable-debug is used with the
9 new case code. These bugs are hidden when the data set is small
10 enough to find in memory; if a bigger data set that would overflow
11 to disk were used, then data corruption would occur.
13 * chisquare.c (create_freq_hash): Pass free_freq_mutable_hash to
14 hsh_create as free function. Make copy of data put into hash.
16 * oneway.q (free_value): New function.
17 (run_oneway): Use free_value as arg to hsh_create. Make copy of
20 * rank.q (rank_cases): Don't access data in case after we've given
23 Tue Jan 9 19:16:11 2007 Ben Pfaff <blp@gnu.org>
27 * aggregate.c (parse_aggregate_functions) Initialize
30 Fri Dec 22 14:04:09 2006 Ben Pfaff <blp@gnu.org>
32 Simplify missing value handling.
34 * aggregate.c (struct agr_var): Remove `bool include_missing', add
35 `enum mv_class exclude'. Remove `int missing', add `bool
36 saw_missing'. Update users.
38 * descriptives.c (struct dsc_trns): Removed `int
39 include_user_missing', add `enum mv_class exclude'. Update users.
40 (struct dsc_proc): Ditto.
42 * examine.q: (static var value_is_missing): Rename
43 `exclude_values', change type to `enum mv_class'. Update users.
47 Fri Dec 22 19:22:18 WST 2006 John Darrington <john@darrington.wattle.id.au>
49 * frequencies.q : Fixed bug #17420, where the table bounds were overun
50 when /FORMAT=nolabels was given.
52 Wed Dec 20 18:45:31 WST 2006 John Darrington <john@darrington.wattle.id.au>
54 * binomial.c binomial.h : New files. Thanks to Jason Stover
55 for assistance with these.
57 * chisquare.c chisquare.h freq.c freq.h npar-summary.c npar-summary.h
58 npar.h npar.q: New files. Implementing NPAR TESTS.
60 * frequencies.q : Moved structure definitions into freq.[ch]
62 Sat Dec 16 22:26:44 2006 Ben Pfaff <blp@gnu.org>
64 Make it possible to pull cases from the active file with a
65 function call, instead of requiring indirection through a callback
68 * aggregate.c (cmd_aggregate): Take advantage of new procedure
70 (agr_to_active_file): Removed.
71 (presorted_agr_to_sysfile): Removed.
73 * autorecode.c (cmd_autorecode): Take advantage of new procedure
75 (autorecode_proc_func): Removed.
77 * flip.c (struct flip_pgm): New members to allow conformance with
78 new case_source_class interface.
79 (cmd_flip): Adapt to new case_source_class interface.
80 (flip_source_read): Ditto.
81 (flip_source_destroy): Ditto.
83 Sat Dec 16 12:54:27 2006 Ben Pfaff <blp@gnu.org>
85 * rank.q (rank_custom_variables): Allow grouping variables to be
86 strings. Fixes bug #18533. Thanks to John Darrington for review.
88 Sat Dec 9 18:47:51 2006 Ben Pfaff <blp@gnu.org>
90 * regression.q (is_depvar): Compare variable pointers instead of
93 Thu Dec 7 15:26:25 WST 2006 John Darrington <john@darrington.wattle.id.au>
95 * examine.q: Allocated the categorical values for the dependent and
96 independent variables, on the heap. Hence they can be of any width.
98 Wed Dec 6 21:14:26 2006 Ben Pfaff <blp@gnu.org>
100 * regression.q (reg_inserted): Compare variable pointers instead
103 Mon Dec 4 22:33:46 2006 Ben Pfaff <blp@gnu.org>
105 * crosstabs.q (insert_summary): Use var_to_string for labeling.
106 (output_pivot_table) Ditto.
109 * frequencies.q (setup_z_trns): Ditto.
111 (dump_condensed) Ditto.
112 (dump_statistics) Ditto.
115 Sun Nov 5 08:31:42 WST 2006 John Darrington <john@darrington.wattle.id.au>
117 * t-test.q, oneway.q: Changed to use the new casefilter structure.
119 Sat Oct 14 16:52:28 2006 Ben Pfaff <blp@gnu.org>
121 * rank.q: (rank_sorted_casefile) Add some missing case_destroy()
122 calls to fix a memory leak.
124 Sun Oct 8 09:45:40 WST 2006 John Darrington <john@darrington.wattle.id.au>
126 * rank.q: Plugged a small memory leak which occurred under error
129 Sat Oct 7 11:06:01 WST 2006 John Darrington <john@darrington.wattle.id.au>
131 * rank.q: Implemented most of the RANK command.
133 2006-07-14 Jason Stover <jhs@math.gcsu.edu>
135 * regression.q (run_regression): New function to move knowledge of
136 pspp_linreg_cache out of math/coefficient.[ch].
138 Sat Jul 1 17:41:46 2006 Ben Pfaff <blp@gnu.org>
140 Fix bug #11612, "q2c documentation does not agree with code".
142 * examine.q: Audit use of q2c "+" prefixes that indicate that a
143 command may appear multiple times.
145 * frequencies.q: Ditto.
149 * regression.q: Ditto.
153 Fri Jun 23 14:18:22 2006 Ben Pfaff <blp@gnu.org>
155 Support long string variables on FREQUENCIES, as
156 an extension when in enhanced algorithms mode. For Greg Hunt
157 <greg@firmansyah.com>.
159 * frequencies.q: (struct freq) Change `v' member from union value
160 to union value *. Update all references.
161 (struct var_freqs) Add width, print members to represent effective
162 variable width and display format.
163 (calc) Copy entire long string value into the hash table.
164 (frq_custom_variables) Set new width, print members.
165 (hash_value_alpha) Get width from var_freqs.
166 (compare_value_alpha_a) Ditto.
167 (compare_freq_alpha_a) Ditto.
168 (compare_freq_alpha_d) Ditto.
169 (dump_full) Get display format from var_freqs.
170 (dump_condensed) Ditto.
172 Mon Jun 19 22:07:13 2006 Ben Pfaff <blp@gnu.org>
174 * frequencies.q: (dump_full) Only put the first MAX_SHORT_STRING
175 bytes of string variables into the output cells, seeing as we only
177 (dump_condensed) Ditto.
179 Mon Jun 19 21:52:05 2006 Ben Pfaff <blp@gnu.org>
181 Fixes a bug reported by Greg Hunt <greg@firmansyah.com>.
183 * frequencies.q: (hsh_hash_bytes) We only copy the first
184 MAX_SHORT_STRING bytes of string variables, so we must only
185 compare that many bytes, even if the string variable is longer.
186 (compare_value_alpha_a) Ditto.
187 (compare_freq_alpha_a) Ditto.
188 (compare_freq_alpha_d) Ditto.
190 2006-05-11 Jason Stover <jhs@math.gcsu.edu>
192 * regression.q: Adjusted code to account for cache->coeff being a
193 pspp_linreg_coeff **.
195 Sun May 7 18:31:25 2006 Ben Pfaff <blp@gnu.org>
199 * aggregate.c (cmd_aggregate): Free default_dict before replacing
202 Sun May 7 17:09:19 2006 Ben Pfaff <blp@gnu.org>
204 * flip.c (flip_file): Check return value of pool_fclose().
206 Sat May 6 16:00:13 2006 Ben Pfaff <blp@gnu.org>
208 Get rid of `char *c' member in union value, for cleanliness.
210 * aggregate.c: (union agr_argument) New union.
211 (struct agr_var) Change element type of arg[] from union value to
213 (parse_aggregate_functions) Change local variable types likewise.
215 * autorecode.c: (union arc_value) New union.
216 (struct arc_item) Change "from" from union value to union
218 (recode) Change local variable from union value to union
220 (autorecode_trns_proc) Ditto.
221 (compare_alpha_value) Ditto.
222 (hash_alpha_value) Ditto.
223 (compare_numeric_value) Ditto.
224 (hash_numeric_value) Ditto.
225 (autorecode_proc_func) Ditto.
227 Sat May 6 10:43:33 2006 Ben Pfaff <blp@gnu.org>
229 Continue reforming procedure execution. In this phase, get rid of
230 the output code for SPLIT FILE groups in procedure.c, which really
231 shouldn't be doing any output. Move it into the individual
232 procedures instead. This also adds some flexibility.
234 * crosstabs.q (precalc): Call output_split_file_values().
236 * descriptives.c (calc_descriptives): Ditto.
238 * examine.q (run_examine): Ditto.
240 * frequencies.q (precalc): Ditto.
242 * oneway.q (run_oneway): Ditto.
244 * regression.q (run_regression): Ditto.
246 * t-test.q (calculate): Ditto.
248 Wed May 3 23:05:31 2006 Ben Pfaff <blp@gnu.org>
250 Continue reforming procedure execution. In this phase, get rid of
251 many global variables, consolidating procedure execution in
252 procedure.c. Encapsulate transformations in new "struct
253 trns_chain". Also, change implementation of N OF CASES, FILTER,
254 and PROCESS IF from special cases to transformations.
256 * aggregate.c (cmd_aggregate) Use discard_variables().
258 2006-04-28 Jason Stover <jhs@math.gcsu.edu>
260 * regression.q (regression_trns_resid_proc): Pass only the
261 variables used in the model to (*model->residual)().
263 * regression.q (regression_trns_pred_proc): Pass only the
264 variables used in the model to (*model->pred)().
266 2006-04-26 Jason Stover <jhs@math.gcsu.edu>
268 * regression.q: Added support for multiple transformations.
270 * regression.q (regression_trns_resid_proc): New function.
272 * regression.q (regression_trns_pred_proc): New function.
274 * regression.q (subcommand_save): Added support for saving
277 * regression.q (regression_trns_free): New function.
279 * regression.q (reg_get_name): New function.
281 * regression.q (reg_save_var): New function.
283 Tue Apr 25 13:18:56 2006 Ben Pfaff <blp@gnu.org>
285 * rank.q (parse_rank_function): Use SE instead of ME for parse
288 Tue Apr 25 13:16:28 2006 Ben Pfaff <blp@gnu.org>
290 * flip.c (flip_sink_write): Use snprintf() to simplify a bit of
293 2006-04-21 Jason Stover <jhs@math.gcsu.edu>
295 * regression.q (try_name): New function. (Partly copied from
296 try_name in descriptives.c.)
298 * regression.q (subcommand_save): Choose residual variable names
301 2006-04-20 Jason Stover <jhs@math.gcsu.edu>
303 * regression.q (cmd_regression): Moved call to subcommand_save()
304 outside multipass_procedure_with_splits().
306 * regression.q (regression_trns_proc): Fixed value counter n_vals
307 before calling *model->residual().
309 2006-04-19 Jason Stover <jhs@math.gcsu.edu>
311 * regression.q (regression_trns_proc): Fixed the look-up of the
314 2006-04-18 Jason Stover <jhs@math.gcsu.edu>
316 * regression.q (regression_trns_proc): Look up the residual
317 variable in the linear regression cache.
319 * regression.q (subcommand_save): Set the residual variable in the
320 linear regression cache.
322 2006-04-17 Jason Stover <jason@wonko.gcsu.edu>
324 * regression.q (regression_trns_proc): Accept case_idx as an int
325 to match the definition of trns_proc_func.
327 2006-04-17 Jason Stover <jhs@math.gcsu.edu>
329 * regression.q (regression_trns_proc): New function.
331 * regression.q (subcommand_save): Create variable residuals and
332 add a transformation to assign values to them. Also free the
333 linreg_cache if the SAVE command was not called. Removed the
334 casereading loop. Placed actual computation of residuals in
335 regression_trns_proc.
337 * regression.q (run_regression): Moved call to free
338 pspp_linreg_cache to subcommand_save.
340 Sat Apr 15 18:01:03 2006 Ben Pfaff <blp@gnu.org>
342 * examine.q (output_examine): Add casts to fix warnings.
344 2006-04-07 Jason Stover <jhs@math.gcsu.edu>
346 * regression.q (subcommand_save): New function.
348 2006-04-04 Jason Stover <jhs@math.gcsu.edu>
350 * regression.q: New function reg_has_categorical () to tell
351 whether a pspp_linreg_struct was made with any variables of type
354 * regression.q: (subcommand_export): Call
355 reg_print_categorical_encoding() only if the model uses any
356 categorical variables.
358 Mon Mar 27 16:00:42 2006 Ben Pfaff <blp@gnu.org>
360 * crosstabs.q: (output_pivot_table) Drop spurious space from
363 2006-03-15 Jason Stover <jhs@math.gcsu.edu>
365 * regression.q: Added custom syntax parser for VARIABLES subcommand
367 * regression.q: Moved most instructions for run_regression ()
368 inside the loop over dependent variables.
370 Thu Mar 2 08:40:33 WST 2006 John Darrington <john@darrington.wattle.id.au>
372 * Moved files from src directory