Debian Codegraph @

ChangeLog

raw
1 v1.24.1 Sat, 04 Jan 2025 09:02:21 +0100 mohawk2
2 - use `signed char` to accumulate logical values - #19
3 + same problem as https://github.com/PDLPorters/pdl/issues/502
4 + PR #20 - fixups
5 - set dimsize in Pars
6 - CI min 5.16
7
8 v1.24.0 Fri, 03 Jan 2025 08:59:42 +0100
9 + github issue #16: overhaul PP output dimension initialization (use RedoDimsCode)
10 + add and use CCS_PDL_IS_NULL macro
11 + double-check ccsutils type signatures (remove int+)
12 + moderinize bad-handling
13 + workaround for failing bandover,borover tests under PDL-2.096
14 + expand TABs to spaces in *.{pm,pd,t,PL,perl} files
15 + fix indentation, opt-in to $multi_c from local submodule builds
16 + add deprecation comment for ccs_indx() etc. to auto-generated Config.pm
17 + require perl >= v5.10 for //= operator
18 + fix typo for 'Func' export tag in Config.PL
19
20 v1.23.29 Fri, 27 Dec 2024 17:59:08 +0100
21 + support BAD non-missing values in ccs_matmult2d_zdd
22 - TODO: support BAD values (including missing) in ccs_matmult2d_sdd
23
24 v1.23.28 Fri, 20 Dec 2024 20:12:18 +0100
25 + fix 'Runtime error: Tried to convert(null)' when multiplying all-missing CCS::Nd
26 - reported as https://github.com/moocow-the-bovine/PDL-CCS/issues/14#issuecomment-2556862635
27
28 v1.23.27 Fri, 20 Dec 2024 11:10:00 +0100 mohawk2
29 + PR #15 assign "wanted" data of correct type in _ind and qsorti tests
30
31 v1.23.26 Fri, 20 Dec 2024 10:55:18 +0100
32 + allow ccs_accum_sumover() to accept empty nzValsIn
33 + add tests for https://github.com/moocow-the-bovine/PDL-CCS/issues/14
34 + set out_type=>indx for ccs_accum_nbad() - https://github.com/moocow-the-bovine/PDL-CCS/issues/6
35
36 v1.23.25 Thu, 14 Nov 2024 16:29:14 +0100
37 + fix embarrassing typo in ChangeLog
38
39 v1.23.24 Thu, 14 Nov 2024 16:06:52 +0100
40 + PR #13 if want $PDL::VERSION, load PDL
41 + PR #12 ccs_pointerlen removal
42 - keep pure-perl PDL::CCS::Functions::ccs_pointerlen for paranoia reasons
43
44 v1.23.23 Wed, 15 May 2024 21:51:55 +0200 mohawk2
45 + PR #8: no strict "refs" instead of string-eval
46 + PR #9: switch from $PP() to $P()
47
48 v1.23.22 Fri, 14 Apr 2023 14:11:53 +0200 mohawk2
49 + PR #7 from mowhawk2/tweaks: error-handling tweaks & bugfixes for PDL v2.082
50 - stop passing in outputs in ccs_binop_vector_mia
51 - DRY in CCS::Functions
52 - stop passing in outputs in recode
53 - done_testing means no need count-padding in CCS/t/06_matops.t
54 - DRY in CCS/t/06_matops.t
55 - use params not globals in CCS/t/06_matops.t
56 - _ufuncsub to give stack-trace if $vals1 is empty
57 - zap old files
58 - MANIFEST.SKIP expansion
59 - zap patches
60 - zap nullbarf file
61 - zap CVS remnants
62 - if empty nzvalsIn, stack-trace instead of SEGV
63 - empty nzvalsIn, stack-trace instead of SEGV
64
65 v1.23.21 Sat Apr 8 12:01:01 2023 +0200 moocow
66 + adjust test expectations for CCS/Utils/t/02_encode.t
67 + port mohawk2 test tweaks from https://github.com/moocow-the-bovine/PDL-HMM/pull/2/
68
69 v1.23.20 Tue, 19 Apr 2022 11:03:00 +0200 moocow
70 + fixed variable-clobbering warning in 02_encode.t
71 + type-mismatch fixes for ccs ufunc counters https://github.com/moocow-the-bovine/PDL-CCS/issues/6
72 - ccs ufunc counters (nbad, ngood, nnz) set out_type=>'indx' rather than out_type=>'int+'
73 - ccs ufunc counters always clear nzvalsOut bad-flag in CopyBadStatusCode
74 - ccs ufunc tests ensure $missing->type==$nzvalsIn->type
75
76 v1.23.19 Thu, 14 Apr 2022 17:05:42 +0200 mohawk2
77 + update default config for recent PDL
78 + fix for ccs_accum_hash() code generator in ccsufunc.pd
79 + cleaner tests (use strict+warnings, zap redundant use_ok-only tests)
80
81 v1.23.18 Fri, 18 Feb 2022 17:10:21 +0100 moocow
82 + fixes for PDL v2.073, contributed by mohawk2
83 + shared github actions, contributed by zmughal
84
85 v1.23.17 Tue, 18 Jan 2022 21:26:32 +0100 moocow
86 + merged changes for PDL 2.066+ from mohawk2
87 - opt in to PDL 2.058 multi-C, dep on PDL v2.019
88 - simplify for 2.014+ types, compat with PDL 2.066+
89 + updated copyright notices in PODs
90
91 v1.23.16 Thu, 29 Apr 2021 08:06:04 +0200 moocow
92 + fixed $PDL::VERSION checks in Config.PL to reflect reality
93 - integer-type downcasting behavior actually changed in PDL-2.037 with commit #f892aeb4ae on PDL/Basic/Ufunc/ufunc.pd
94 - should fix new cpantesters failures, e.g. http://www.cpantesters.org/cpan/report/1fc08e78-a7e3-11eb-aa01-337c1f24ea8f
95 + added missing "resources" level to META_MERGE section in Makefile.PL
96
97 v1.23.15 Tue, 27 Apr 2021 13:13:20 +0200 moocow
98 + added new PDL-2.039 integer-type listing code to Config.PL
99 + fixed downcasting in b*over methods be PDL-compatible (problem was 'max_int_type' option to ccs_accum_def())
100 + re-enabled b*over tests skipped in v1.23.14
101 + added Makefile.PL META_MERGE section pointing to new public github repo moocow-the-bovine/PDL-CCS
102 - github repo is just a fork of read-only ZDL git mirror of upstream SVN repository
103
104 v1.23.14 Mon, 26 Apr 2021 14:58:50 +0200 moocow
105 + skip some b(and|or)over type-check tests to avoid test failures for PDL >= v2.039
106 - CCS implementations are returning 'indx' type here, dense PDL versions are giving 'longlong' for input type=double
107
108 v1.23.13 Thu, 19 Nov 2020 06:54:54 +0100 moocow
109 + fix RT bug #133772, reported by Sebastiaan Couwenberg (spelling error in POD)
110
111 v1.23.12 Tue, 28 Aug 2018 09:20:08 +0200 moocow
112 + RT bug #126924, part III
113 - added exception to CCS::Nd::borover() test in CCS/t/03_ufuncs.t
114 - analogous to v1.23.11 fix for CCS/Ufunc/t/01_ufunc.t
115
116 v1.23.11 Mon, 27 Aug 2018 14:35:47 +0200 moocow
117 + workaround for RT bug #126294, reported by G. Herrmann
118 - skip ufunc "borover:missing=BAD" test in CCS/Ufunc/t/01_ufunc.t if PDL::borover() is broken
119 - upstream patch submitted to PDL maintainers as https://sourceforge.net/p/pdl/bugs/446/
120
121 v1.23.10 Fri, 24 Aug 2018 10:14:40 +0200 moocow
122 + fixed typo in failed test label-reporting labstr() in t/common.plt, added 'use strict'
123 + may help to diagnose RT bug #126294 (CCS/Ufunc/t/01_ufunc.t test 'borover:missing=BAD' fails on armv6l-linux ~ rpi)
124
125 v1.23.9 Fri, 22 Jun 2018 13:55:35 +0200 moocow
126 + ufunc.pd: avoid "|=" and "&=" operators (attempt to get build working on ARM64, reported by L. Baillet)
127 - see http://www.cpantesters.org/cpan/report/eaad8962-7102-11e8-905e-5ddc267117a8
128 - see https://buildd.debian.org/status/package.php?p=libpdl-ccs-perl
129 + more verbose diagnostics for failed pdlok() tests in t/common.plt
130
131 v1.23.8 Fri, 15 Jun 2018 13:45:06 +0200 moocow
132 + various fixes for debian packaging (RT bug #125587),
133 patches provided by L. Baillet and G. Herrmann
134
135 v1.23.7 Wed, 06 Jun 2018 09:18:55 +0200 moocow
136 + CCS/IO 'clean' target: remove test temporaries t/ccs3.* t/dense3.*
137
138 v1.23.6 Tue, 05 Jun 2018 16:39:46 +0200 moocow
139 + fixed "do 'Config.PL'" call in Makefile.PL
140
141 v1.23.5 Tue, 05 Jun 2018 15:05:00 +0200 moocow
142 + fixed typos reported by L. Baillet (RT bug #125493)
143
144 v1.23.4 Tue, 06 Jun 2017 10:17:44 +0200 moocow
145 + fixed bogus bareword pdl() call in CCS/Nd.pm (RT bug #121952)
146 + added 'use lib "."' to Makefile.PL (RT bug #121661)
147
148 v1.23.3 Mon, 06 Jun 2016 14:45:29 +0200 moocow
149 + fixed ccs_wfits() typecast-to-long hack for indx types
150 - feature request including patch posted to https://sourceforge.net/p/pdl/bugs/421/
151
152 v1.23.2 Mon, 06 Jun 2016 11:20:12 +0200 moocow
153 + win32/NaN fixes for tests (RT bug #115078)
154
155 v1.23.1 Tue, 12 Jan 2016 13:24:00 +0100 moocow
156 + fixed index overflow bug picking maximum output dimension in perl-side ccs_xindex2d()
157 + added optional pass-in $anorm() for ccs_vcos_zdd()
158 + added pointer-optimized sparse-crs matrix vs. sparse-coo vector cosine method ccs_vcos_pzd()
159
160 v1.23.0 Tue, 15 Dec 2015 13:25:35 +0100 moocow
161 + fixed "uninitialized value" warnings for PDL->can($type) in PDL::CCS::IO::*
162 + added support for ndims>2 to PDL::CCS::IO::MatrixMarket
163 + added sparse/dense vector-cosine ccs_vcos_zdd() in PDL::CCS::MatrixOps, with wrapper PDL::CCS::Nd::vcos_zdd()
164 - dense/dense variant in PDL::VectorValued::Utils::vv_vcos() for PDL::VectorValued v1.0.5
165 + fixed BAD handling in CCS::Compat::ccsget(), CCS::Compat::ccsget2d()
166 + fixed I/O type handling in integer ufuncs (borover)
167 + fixed wrongly succeeding bogus tests with unary ok()
168 - tests now use Test::More and re-factored common test subroutines
169 + fixed C-level abs() function in CCS::Utils::nnza(); now dispatches to one of {abs,labs,llabs,fabsf,fabs} using PDL::PP types(...) macro
170 + fixed CCS ufunc type-promotion logic to be compatible with PDL v2.015
171 + pared down CCS/t/06_matops.t to test only missing==0 : matrix ops don't work correctly with missing!=0
172 + pdlmaker.plm doesn't distribute generated PM files any more (PDL now does this for us)
173
174 v1.22.6 Wed, 25 Nov 2015 16:27:24 +0100 moocow
175 + added CCS::IO::FITS, CCS::IO::MatrixMarket, CCS::IO::LDAC, CCS::IO::PETSc
176 + moved common I/O utilities to CCS::IO::Common
177 + added CCS::IO tests
178
179 v1.22.5 Mon, 23 Nov 2015 12:34:25 +0100 moocow
180 + no real joy with Makefile.PL workaround (UNKNOWN results are still pretty wonky)
181 + updating PDL::VectorValued to use shared $VERSION via perl-reversion script from module Perl::Version
182 + PDL::CCS can now depend directly on PDL::VectorValued
183
184 v1.22.4 Tue, 17 Nov 2015 09:54:23 +0100 mocoow
185 + Makefile.PL workaround for PDL::VectorValued(::Version) strangeness on cpantesters
186 - see http://sourceforge.net/p/pdl/mailman/message/34623263/ ("headaches with indirect PDL-related dependencies on cpantesters", 2015-11-16 13:16:44)
187
188 v1.22.3 Thu, 05 Nov 2015 10:43:54 +0100 moocow
189 + workaround for PDL::clump(-N) bug in PDL-v2.014: compute non-negative clump() arguments in CCS/Compat.pm
190 - see RT bug #108472; PDL bug https://sourceforge.net/p/pdl/bugs/406/
191 + workaround for changed PDL::reshape() behavior in CCS/t/05_binops.t, CCS/t/06_matops.t : getting ugly realloc errors without it
192 - see RT bug #107829
193
194 v1.22.2 Tue, 18 Aug 2015 13:04:09 +0200 moocow
195 + added clearptr($pdimi) method
196
197 v1.22.1 Wed, 08 Apr 2015 16:09:43 +0200 moocow
198 + fixed ccs_xindex2d() utility and added CCS::Nd::xsubset2d() wrapper
199
200 v1.22.0 Wed, 08 Apr 2015 13:49:14 +0200
201 + added ccs_xindex2d() utility function: fast Cartesian product indexing of sparse 2d matrices
202
203 v1.21.0 Mon, 16 Mar 2015 13:22:34 +0100 moocow
204 + added PDL::IO::FastRaw wrappers (incl mapfraw) for PDL::CCS::Nd objects
205 + PDL::CCS::Nd->fromWhich() now accepts ARRAY-refs for 'pdims' and 'vdims' options
206
207 v1.20.2 Wed, 05 Nov 2014 13:24:55 +0100 moocow
208 + more empty-piddle fixes for PDL-v2.4.11 (kaskade / debian wheezy)
209
210 v1.20.1 Wed, 05 Nov 2014 10:41:19 +0100 moocow
211 + improved handling of empty pdls in PDL::CCS::Nd
212
213 v1.19.1 Thu, 26 Sep 2013 08:57:11 +0200 moocow
214 + use ExtUtils::MakeMaker::prompt() for configuration questions; fixes RT #88972
215
216 v1.19.0 Wed, 25 Sep 2013 12:13:27 +0200 moocow
217 + added (optional) support for 64-bit indices via PDL_Indx (requires PDL >= v2.007)
218
219 v1.18.0 Wed, 07 Nov 2012 13:57:26 +0100
220 + added CCS::Functions::ccs_qsort(), CCS::Nd::qsort(), CCS::Nd::qsorti()
221 + new qsort code requires PDL::VectorValued >= v0.06 (for enumvec())
222 + added :lvalue attribute to selected CCS::Nd and CCS::Functions subs
223
224 v1.16 Mon, 02 Jan 2012 13:38:48 +0100 moocow
225 + cpan-friendly distribution with pdlmaker.plm
226
227 v1.15 2011-12-20 moocow
228 * [r5936] band-aided barf()ing PDL::CCS::Functions::ccs_decode()
229 due to mismatched dimensions in empty index and value piddles
230
231 v1.14 2011-03-31 moocow
232 * [r5596] CCS/Makefile.PL, CCS/Nd.pm, CCS/Version.pm,
233 CCS/testme.perl, ChangeLog, Makefile.PL, testme.perl: + v1.14:
234 updated for PDL::VectorValued 0.04 (qsortveci -> vv_qsortveci)
235
236 v1.13 2010-02-26 moocow
237 * [r4085] CCS/MatrixOps/ccsmatops.pd, CCS/Nd.pm, CCS/Version.pm,
238 testme.perl: + updated MatrixOps::ccs_matmult2d_sdd : 2d matrix
239 mult with arbitrary finite missing values
240 * [r4084] MANIFEST: + updated MANIFEST (added CCS/MatrixOps/
241 subdir)
242
243 v1.12 2009-11-04 moocow
244 * [r3653] CCS/Nd.pm: + v1.12: fixed empty-dimension bug in
245 CCS::Nd::dice_axis() [not indexND as in last log message]
246 * [r3652] CCS/Nd.pm, CCS/Utils/ccsutils.pd, CCS/Version.pm: +
247 v1.12: fixed empty-dimension bug in CCS::Nd::indexND()
248
249 v1.11 2009-10-31 moocow
250 * [r3621] CCS/MatrixOps/ccsmatops.pd, CCS/Nd.pm, CCS/t/06_matops.t,
251 testme.perl: + added matmult2d_zdd() variant: should really work
252 * [r3618] CCS.pm, CCS/Attic, CCS/Makefile.PL, CCS/MatrixOps,
253 CCS/MatrixOps/Makefile.PL, CCS/MatrixOps/ccsmatops.pd,
254 CCS/MatrixOps/t, CCS/MatrixOps/t/00_basic.t,
255 CCS/MatrixOps/t/common.plt, CCS/Nd.pm, CCS/Version.pm,
256 CCS/t/06_matops.t, testme.perl: + v1.11: added
257 CCS::Nd::matmult2d_sdd for correct matrix multiplication with
258 dense 2nd operand and output
259
260 v1.10 2009-10-22 moocow
261 * [r3567] CCS/Nd.pm, CCS/Version.pm: + v1.10: added isbad(),
262 isgood()
263
264 v1.09 2009-10-19 moocow
265 * [r3540] CCS/Ufunc/ccsufunc.pd:
266 * [r3539] CCS/Ufunc/ccsufunc.pd, CCS/Version.pm:
267 * [r3538] CCS/Ufunc/ccsufunc.pd:
268
269 v1.08 2009-07-16 moocow
270 * [r3395] CCS/Nd.pm, CCS/Version.pm, ChangeLog, testme.perl: +
271 added CCS::Nd methods interpolate(), interpol()
272
273 v1.07 2008-07-26 moocow
274 * [r2534] CCS/Nd.pm: + improved BAD handling in
275 _ccsnd_binary_op_mia()
276 * [r2533] CCS/Nd.pm, CCS/Version.pm: + improved BAD handling in
277 _ccsnd_binary_op_mia()
278
279 v1.06 2008-06-26 moocow
280 * [r2490] CCS/Version.pm: + v1.06: added CCS::Nd::_missing()
281 * [r2489] CCS/Nd.pm: + added '_missing()' method
282
283 v1.05 Fri, 02 May 2008 13:00:22 +0200
284 + added CCS::Nd::ismissing(), CCS::Nd::ispresent() mask methods
285 + added CCS::Nd::maximum_ind(), CCS::Nd::minimum_ind()
286
287 v1.04 Mon, 28 Apr 2008 23:48:57 +0200
288 + added PDL::CCS::Ufunc::ccs_accum_average()
289 + added PDL::CCS::Nd wrappers: average_nz, avg_nz, average, avg
290 + documented PDL::CCS::Nd method _nzvals(), added alias _whichVals()
291 + added CCS::Nd::badflag()
292
293 v1.03 Wed, 20 Feb 2008 10:30:39 +0100
294 + fixed some bugs in PDL::CCS::Nd::to_physically_indexed()
295 - bad use of vdims for pdims (output pdl should be physically ordered)
296 - missing value wasn't getting appended to output pdl
297
298 v1.02 Thu, 14 Feb 2008 12:50:11 +0100
299 + fixed some bugs in PDL::CCS::Nd::string(), ::appendWhich() for objects with empty index pdls
300 + fixed virtual-dimension indexing bug in PDL::CCS::Nd::indexNDi
301 causing indexing to fail for e.g. transposed sparse matrices
302
303 v1.01 Tue, 24 Apr 2007 01:24:05 +0200 (moocow)
304 + added PDL::CCS::Nd perl class for sparse Nd piddle-like structures
305 + separated out submodules Utils, Ufunc, Ops
306 + added PDL::CCS::Compat for backwards-compatibility
307
308 2007-03-27 moocow
309 * [r1933] Attic/CCS.pd, CCS.pd: + minor documentation fixes
310
311 2007-02-27 moocow
312 * [r1847] Attic/CCS.pd, CCS.pd, t/02_encode.t, t/03_ops.t,
313 t/Attic/02_encode.t, t/Attic/03_ops.t, t/common.plt, testme.perl:
314 + added partial decoding, encoding from indices, ufuncs
315
316 2007-02-26 moocow
317 * [r1845] Attic/CCS.pd, CCS.pd, t/03_ops.t, t/Attic/03_ops.t,
318 testme.perl: + added whichND, which, transpose, and basic vector
319 ops
320 * [r1842] Attic/CCS.pd, CCS.pd: + added bad-processing stuff to
321 PDL::CCS
322
323 2007-02-26 moocow
324 * [r1845] Attic/CCS.pd, CCS.pd, t/03_ops.t, t/Attic/03_ops.t,
325 testme.perl: + added whichND, which, transpose, and basic vector
326 ops
327 * [r1842] Attic/CCS.pd, CCS.pd: + added bad-processing stuff to
328 PDL::CCS
329
330 2005-08-02 moocow
331 * [r1215] Attic/CCS.pd, CCS.pd, ChangeLog: re-import (gaspode)
332
333 v0.01 Sat, 11 Jun 2005 10:32:05 +0200 (moocow)
334 + initial version, 2d pdls only