-
Notifications
You must be signed in to change notification settings - Fork 2
/
NEWS
1705 lines (1279 loc) · 75.8 KB
/
NEWS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
GLPK 4.52 (release date: Jul 18, 2013)
The clique cut generator was essentially reimplemented, and now
it is able to process very large and/or dense conflict graphs.
A simple rounding heuristic was added to the MIP optimizer.
Some bugs were fixed in the proximity search heuristic routine.
Thanks to Giorgio Sartor <[email protected]>.
New command-line option '--proxy [nnn]' was added to glpsol to
enable using the proximity search heuristic.
A bug (incorrect processing of LI column indicator) was fixed
in the mps format reading routine. Thanks to Charles Brixko for
bug report.
GLPK 4.51 (release date: Jun 19, 2013)
Singleton and dense phases were implemented on computing
LU-factorization with Gaussian elimination. The singleton phase
is a feature that allows processing row and column singletons
on initial elimination steps more efficiently. The dense phase
is a feature used on final elimination steps when the active
submatrix becomes relatively dense. It significantly reduces
the time needed, especially if the active submatrix fits in CPU
cache, and improves numerical accuracy due to full pivoting.
The API routine glp_adv_basis that constructs advanced initial
LP basis was replaced by an improved version, which (unlike the
old version) takes into account numerical values of constraint
coefficients.
The proximity search heuristic for MIP was included in the GLPK
integer optimizer glp_intopt. On API level the heuristic can be
enabled by setting the parameter ps_heur in glp_iocp to GLP_ON.
This feature is also available in the solver glpsol through
command-line option '--proxy'. Many thanks to Giorgio Sartor
<[email protected]> for contribution.
A bug was fixed that caused numerical instability in the FPUMP
heuristic.
GLPK 4.50 (release date: May 24, 2013)
A new version of LU-factorization routines were added.
Currently this version provides the same functionality as the
old one, however, the new version allows further improving.
Old routines for FHV-factorization used to update the basis
factorization were replaced by a new version conforming to the
new version of LU-factorization.
Some clarifications about using the name index routines were
added. Thanks to Xypron <[email protected]> for suggestion.
Some typos were corrected in the MathProg language reference.
Thanks to Jeffrey Kantor <[email protected]> for report.
A serious bug (out-of-range indexing error) was *tentatively*
fixed in the routine glp_relax4. Unfortunatly, this bug is
inherited from the original Fortran version of the RELAX-IV
code (for details please see ChangeLog), and since the code is
very intricate, the bug is still under investigation. Thanks to
Sylvain Fournier for bug report.
GLPK 4.49 (release date: Apr 16, 2013)
The new API routine glp_mincost_relax4, which is a driver to
relaxation method of Bertsekas and Tseng (RELAX-IV), was added
to the package. RELAX-IV is a code for solving minimum cost
flow problems. On large instances it is 100-1000 times faster
than the standard primal simplex method. Prof. Bertsekas, the
author of the original RELAX-IV Fortran code, kindly permitted
to include a C translation of his code in GLPK under GPLv3.
A bug (wrong dual feasibility test) was fixed in API routine
glp_warm_up. Thanks to David T. Price <[email protected]>
for bug report.
Obsolete API routine lpx_check_kkt was replaced by new routine
glp_check_kkt.
IMPORTANT: All old API routines whose names begin with 'lpx_'
were removed from API level and NO MORE AVAILABLE.
GLPK 4.48 (release date: Jan 28, 2013)
This is a maintainer release.
Some minor changes in API (glpk.h) were made. For details
please see ChangeLog.
Some bugs/typos were fixed. Thanks to
Raniere Gaia Costa da Silva,
Heinrich Schuchardt <[email protected]>, and
Robbie Morrison <[email protected]> for reports.
GLPK 4.47 (release date: Sep 09, 2011)
The new API routine glp_intfeas1 was added to the package.
This routine is a tentative implementation of the integer (0-1)
feasibility solver based on the CNF-SAT solver (which currently
is MiniSat). It may be used in the same way as glp_intopt to
find either any integer feasible solution or a solution, for
which the objective function is not worse than the specified
value. Detailed description of this routine can be found in the
document "CNF Satisfiability Problem", which is included in the
distribution (see doc/cnfsat.pdf).
The following two options were added to glpsol:
--minisat translate 0-1 feasibility problem to CNF-SAT
problem and solve it with glp_intfeas1/MiniSat
(if the problem instance is already in CNF-SAT
format, no translation is performed)
--objbnd bound add inequality obj <= bound (minimization) or
obj >= bound (maximization) to 0-1 feasibility
problem (this option assumes --minisat)
The paint-by-numbers puzzle model (pbn.mod) included in the
distribution is a nice example of the 0-1 feasibility problem,
which can be efficiently solved with glp_intfeas1/MiniSat. This
model along with a brief instruction (pbn.pdf) and benchmark
examples from <webpbn.com> encoded in GNU MathProg (*.dat) can
be found in subdirectory examples/pbn/.
The glpsol lp/mip solver was modified to bypass postprocessing
of MathProg models if the solution reported is neither optimal
nor feasible.
A minor bug in examples/Makefile.am was fixed to correctly
build glpk in a separate directory. Thanks to Marco Atzeri
<[email protected]> for bug report and patch.
GLPK 4.46 (release date: Aug 09, 2011)
The following new API routines were added:
glp_read_cnfsat read CNF-SAT problem data in DIMACS format
glp_check_cnfsat check for CNF-SAT problem instance
glp_write_cnfsat write CNF-SAT problem data in DIMACS format
glp_minisat1 solve CNF-SAT problem instance with MiniSat
The routine glp_minisat1 is a driver to MiniSat, a CNF-SAT
solver developed by Niklas Een and Niklas Sorensson, Chalmers
University of Technology, Sweden. This routine is similar to
the routine glp_intopt, however, it is intended to solve a 0-1
programming problem instance, which is the MIP translation of
a CNF-SAT problem instance.
Detailed description of these new API routines can be found in
the document "CNF Satisfiability Problem", which is included in
the distribution (see files doc/cnfsat.tex and doc/cnfsat.pdf).
The following new glpsol command-line options were added:
--cnf filename read CNF-SAT problem instance in DIMACS
format from filename and translate it to MIP
--wcnf filename write CNF-SAT problem instance in DIMACS
format to filename
--minisat solve CNF-SAT problem instance with MiniSat
solver
The zlib compression library (version 1.2.5) was ANSIfied,
modified according to GLPK requirements and included in the
distribution as an external software module. Thus, now this
feature is platform independent.
Some bugs were fixed in the SQL table driver. Thanks to Xypron
GLPK 4.45 (release date: Dec 05, 2010)
This is a bug-fix release.
Several bugs/typos were fixed. Thanks to
Xypron <[email protected]>,
Robbie Morrison <[email protected]>, and
Ali Baharev <[email protected]> for reports.
Some glpk documents was re-formatted and merged into a single
document. Now the glpk documentation consists of the following
three main documents (all included in the distribution):
GLPK: Reference Manual
GLPK: Graph and Network Routines
Modeling Language GNU MathProg: Language Reference
GLPK 4.44 (release date: Jun 03, 2010)
The following suffixes for variables and constraints were
implemented in the MathProg language:
.lb (lower bound),
.ub (upper bound),
.status (status in the solution),
.val (primal value), and
.dual (dual value).
Thanks to Xypron <[email protected]> for draft implementation
and testing.
Now the MathProg language allows comment records (marked by
'#' in the very first position) in CSV data files read with the
table statements. Note that the comment records may appear only
in the beginning of a CSV data file.
The API routine glp_cpp to solve the Critical Path Problem was
added and documented.
GLPK 4.43 (release date: Feb 20, 2010)
This is a maintainer release.
`configure.ac' was changed to allow building the package under
Mac OS and Darwin with ODBC support.
Thanks to Xypron <[email protected]> for suggestions and Noli
Sicad <[email protected]> for testing.
The SQL table driver was improved to process NULL data. Thanks
to Xypron <[email protected]>.
Some bugs were fixed in the LP/MIP preprocessor.
GLPK 4.42 (release date: Jan 13, 2010)
The following new API routines were added:
glp_check_dup check for duplicate elements in sparse
matrix
glp_sort_matrix sort elements of the constraint matrix
glp_read_prob read problem data in GLPK format
glp_write_prob write problem data in GLPK format
glp_analyze_bound analyze active bound of non-basic variable
glp_analyze_coef analyze objective coefficient at basic
variable
glp_print_ranges print sensitivity analysis report (this
routine replaces lpx_print_sens_bnds and
makes it deprecated)
For description of these new routines and the GLPK LP/MIP
format see a new edition of the reference manual included in
the distribution. (Chapter "Graph and network API routines" was
carried out from the main reference manual and included in the
distribution as a separate document.)
The following new command-line options were added to the stand-
alone solver glpsol:
--glp filename read problem data in GLPK format
--wglp filename write problem data in GLPK format
--ranges filename print sensitivity analysis report (this
option replaces --bounds)
Now all GLPK routines performing file I/O support special
filenames "/dev/stdin", "/dev/stdout", and "/dev/stderr", which
can be specified in the same way as regular filenames. This
feature is plaform-independent.
GLPK 4.41 (release date: Dec 21, 2009)
The following new API routies were added:
glp_transform_row transform explicitly specified row
glp_transform_col transform explicitly specified column
glp_prim_rtest perform primal ratio test
glp_dual_rtest perform dual ratio test
For description of these new routines see a new edition of the
reference manual included in the distribution.
The following API routines are deprecated: lpx_transform_row,
lpx_transform_col, lpx_prim_ratio_test, lpx_dual_ratio_test.
Some improvements were made in the MIP solver (glp_intopt).
The SQL table driver used to read/write data in MathProg models
was changed to allow multiple arguments separated by semicolon
in SQL statements. Thanks to Xypron <[email protected]>.
Two new options were added to the glpsol stand-alone solver:
--seed value (to initialize the pseudo-random number generator
used in MathProg models with specified value), and
--ini filename (to use a basis previously saved with -w option
as an initial basis on solving similar LP's).
Two new MathProg example models were included. Thanks to
Nigel Galloway <[email protected]> and Noli Sicad
<[email protected]> for contribution.
Scripts to build GLPK with Microsoft Visual Studio 2010 for
both 32-bit and 64-bit Windows were included. Thanks to Xypron
<[email protected]> for contribution and testing.
GLPK 4.40 (release date: Nov 03, 2009)
The following new API routines were added:
glp_del_vertices remove vertices from graph
glp_del_arc remove arc from graph
glp_wclique_exact find maximum weight clique with the exact
algorithm developed by Prof. P. Ostergard
glp_read_ccdata read graph in DIMACS clique/coloring
format
glp_write_ccdata write graph in DIMACS clique/coloring
format
For description of these new routines see a new edition of the
reference manual included in the distribution.
The hybrid pseudocost branching heuristic was included in the
MIP solver. It is available on API level (iocp.br_tech should
be set to GLP_BR_PCH) and in the stand-alone solver glpsol
(via the command-line option --pcost). This heuristic may be
useful on solving hard MIP instances.
The branching heuristic by Driebeck and Tomlin (used in the
MIP solver by default) was changed to switch to branching on
most fractional variable if an lower bound of degradation of
the objective is close to zero for all branching candidates.
A bug was fixed in the LP preprocessor (routine npp_empty_col).
Thanks to Stefan Vigerske <[email protected]> for the
bug report.
A bug was fixed and some improvements were made in the FPUMP
heuristic module. Thanks to Xypron <[email protected]>.
A bug was fixed in the API routine glp_warm_up (dual
feasibility test was incorrect in maximization case). Thanks to
Uday Venkatadri <[email protected]> for the bug report.
GLPK 4.39 (release date: Jul 26, 2009)
The following new API routines were added:
glp_warm_up "warm up" LP basis
glp_set_vertex_name assign (change) vertex name
glp_create_v_index create vertex name index
glp_find_vertex find vertex by its name
glp_delete_v_index delete vertex name index
glp_read_asnprob read assignment problem data in DIMACS
format
glp_write_asnprob write assignment problem data in DIMACS
format
glp_check_asnprob check correctness of assignment problem
data
glp_asnprob_lp convert assignment problem to LP
glp_asnprob_okalg solve assignment problem with the
out-of-kilter algorithm
glp_asnprob_hall find bipartite matching of maxumum
cardinality with Hall's algorithm
Also were added some API routines to read plain data files.
The API routines glp_read_lp and glp_write_lp to read/write
files in CPLEX LP format were re-implemented. Now glp_write_lp
correctly writes double-bounded (ranged) rows by introducing
slack variables rather than by duplicating the rows.
For description of these new routines see a new edition of the
reference manual included in the distribution.
The 'xfree(NULL)' bug was fixed in the AMD routines. Thanks to
Niels Klitgord <[email protected]> for bug report.
The message "Crashing..." was changed to "Constructing initial
basis..." due to suggestion by Thomas Kahle <[email protected]>.
Some typos were corrected in glpsol output messages. Thanks to
Xypron <[email protected]> for patch.
GLPK 4.38 (release date: May 02, 2009)
API routines glp_read_mps and glp_write_mps were improved.
Some improvements were made in the dual simplex routines.
Two external software modules AMD and COLAMD were included in
the distribution (for more details please see src/amd/README
and src/colamd/README). Now they are used in the interior-point
solver to reorder the matrix prior to Cholesky factorization.
API routine glp_ipt_status may return two new statuses due to
changes in the routine glp_interior. For details please see the
reference manual included in the distribution.
A minor bug was fixed in the graph/network routines. Thanks to
Nelson H. F. Beebe <[email protected]> for bug report.
GLPK 4.37 (release date: Mar 29, 2009)
The 0-1 Feasibility Pump heuristic was included in the GLPK
integer optimizer glp_intopt. On API level the heuristic can be
enabled by setting the parameter fp_heur in glp_iocp to GLP_ON.
This feature is also available in the solver glpsol through
command-line option '--fpump'. For more details please see the
reference manual included in the distribution.
The following new API routines were added:
glp_print_sol write basic solution in printable format
glp_print_ipt write interior-point solution in printable
format
glp_print_mip write MIP solution in printable format
glp_read_graph read (di)graph from plain text file
glp_write_graph write (di)graph to plain text file
glp_weak_comp find all weakly connected components
glp_strong_comp find all strongly connected components
The following API routines are deprecated: lpx_print_sol,
lpx_print_ips, lpx_print_mip, lpx_print_prob (the latter is
equivalent to glp_write_lp).
A bug was fixed in the interior-point solver (glp_interior) to
correctly compute dual solution components when the problem is
scaled.
The files configure.ac and Makefile.am were changed:
(a) to allow using autoreconf/autoheader;
(b) to allow building the package in a directory other than its
source directory.
Thanks to Marco Atzeri <[email protected]> for bug report.
An example model in the GNU MathProg language was added.
Thanks to Larry D'Agostino <Larry.D'[email protected]> for
contribution.
GLPK 4.36 (release date: Feb 06, 2009)
The following new API routines were added to the package:
glp_mincost_okalg find minimum-cost flow with out-of-kilter
algorithm
glp_maxflow_ffalg find maximal flow with Ford-Fulkerson
algorithm
For detailed description of these new routines and related data
structures see chapter "Graph and Network API Routines" in a new
edition of the reference manual included in the distribution.
The following two new command-line options were added to the
solver glpsol:
--mincost read min-cost flow data in DIMACS format
--maxflow read maximum flow data in DIMACS format
Duplicate symbols in the header glpk.h were removed to allow
using swig.
Thanks to Kelly Westbrooks <[email protected]> and
Nigel Galloway <[email protected]> for suggestion.
A minor defect was fixed in the routine glp_write_lp.
Thanks to Sebastien Briais <[email protected]> for bug report.
A minor bug was fixed in the SQL module.
Thanks to Xypron <[email protected]> for patch.
Some new example models in the GNU MathProg modeling language
were added. Thanks to Sebastian Nowozin <[email protected]> and
Nigel Galloway <[email protected]> for contribution.
GLPK 4.35 (release date: Jan 09, 2009)
The following new API routines were added to the package:
glp_create_graph create graph
glp_set_graph_name assign (change) graph name
glp_add_vertices add new vertices to graph
glp_add_arc add new arc to graph
glp_erase_graph erase graph content
glp_delete_graph delete graph
glp_read_mincost read minimum cost flow problem data in
DIMACS format
glp_write_mincost write minimum cost flow problem data in
DIMACS format
glp_mincost_lp convert minimum cost flow problem to LP
glp_netgen Klingman's network problem generator
glp_gridgen grid-like network problem generator
glp_read_maxflow read maximum flow problem data in DIMACS
format
glp_write_maxflow write maximum flow problem data in DIMACS
format
glp_maxflow_lp convert maximum flow problem to LP
glp_rmfgen Goldfarb's maximum flow problem generator
For detailed description of these new routines and related data
structures see chapter "Graph and Network API Routines" in a new
edition of the reference manual included in the distribution.
A minor change were made in the internal routine xputc. Thanks
to Luiz Bettoni <[email protected]> for suggestion.
A minor bug was fixed in the internal routine mpl_fn_time2str.
Thanks to Stefan Vigerske <[email protected]> for bug report.
GLPK 4.34 (release date: Dec 04, 2008)
The GNU MathProg modeling language was supplemented with three
new built-in functions:
gmtime obtaining current calendar time
str2time converting character string to calendar time
time2str converting calendar time to character string
(Thanks to Xypron <[email protected]>.)
For detailed description of these functions see Appendix A in
the document "Modeling Language GNU MathProg", a new edition of
which was included in the distribution.
A bug was fixed in the MIP solver. Thanks to Nigel Galloway
<[email protected]> for bug report.
A new makefile was added to build the GLPK DLL with Microsoft
Visual Studio Express 2008 for 64-bit Windows. Thanks to Xypron
<[email protected]> for contribution and testing.
GLPK 4.33 (release date: Oct 30, 2008)
The following new API routines were added to the package:
glp_copy_prob copy problem object content
glp_exact solve LP in exact arithmetic
(makes lpx_exact deprecated)
glp_get_unbnd_ray determine variable causing unboundedness
(makes lpx_get_ray_info deprecated)
glp_interior solve LP with interior-point method
(makes lpx_interior deprecated)
The following new API routines for processing models written in
the GNU Mathprog language were added to the package:
glp_mpl_alloc_wksp allocate the translator workspace
glp_mpl_read_model read and translate model section
glp_mpl_read_data read and translate data section
glp_mpl_generate generate the model
glp_mpl_build_prob build LP/MIP instance from the model
glp_mpl_postsolve postsolve the model
glp_mpl_free_wksp deallocate the translator workspace
(These routines make lpx_read_model deprecated.)
For description of all these new API routines see the reference
manual included in the distribution.
A crude implementation of CPLEX-like interface to GLPK API was
added to the package. Currently it allows using GLPK as a core
LP solver for Concorde, a well known computer code for solving
the symmetric TSP. For details see examples/cplex/README.
Some bugs were fixed in the SQL table driver. Thanks to Xypron
GLPK 4.32 (release date: Oct 03, 2008)
The following new features were included in the MIP solver
(the API routine glp_intopt):
* MIP presolver
* mixed cover cut generator
* clique cut generator
* Euclidean reduction of the objective value
Due to changes the routine glp_intopt may additionally return
GLP_ENOPFS, GLP_ENODFS, and GLP_EMIPGAP.
The API routines lpx_integer are lpx_intopt are deprecated,
since they are completely superseded by glp_intopt.
The following new branch-and-cut API routines were added:
glp_ios_row_attr determine additional row attributes
glp_ios_pool_size determine current size of the cut pool
glp_ios_add_row add constraint to the cut pool
glp_ios_del_row delete constraint from the cut pool
glp_ios_clear_pool delete all constraints from the cut pool
For description of these new routines see the reference manual
included in the distribution.
The stand-alone solver glpsol was changed to allow multiple
data files.
A new edition of the supplement "Using Data Tables in the GNU
MathProg Modeling Language" was included.
As usual, some bugs were fixed (in the MathProg translator).
Thanks to Xypron <[email protected]>.
GLPK 4.31 (release date: Sep 02, 2008)
The core LP solver based on the dual simplex method was
re-implemented and now it provides both phases I and II.
The following new API routines were added:
glp_scale_prob automatic scaling of problem data
glp_std_basis construct standard initial LP basis
glp_adv_basis construct advanced initial LP basis
glp_cpx_basis construct Bixby's initial LP basis
For description of these new routines see the reference manual
included in the distribution.
The following API routines are deprecated:
lpx_scale_prob, lpx_std_basis, lpx_adv_basis, lpx_cpx_basis.
Necessary changes were made in memory allocation routines to
resolve portability issues for 64-bit platforms.
New version of the routine lpx_write_pb to write problem data
in OPB (pseudo boolean format) was added to the package. Thanks
to Oscar Gustafsson <[email protected]> for the contribution.
Two new makefiles were added to build the package for 32- and
64-bit Windows with Microsoft Visual Studio Express 2008.
Thanks to Heinrich Schuchardt <[email protected]> (aka
Xypron) for the contribution and testing.
Two new makefiles were added to build the package with Digital
Mars C/C++ 8.50 and Open Watcom C/C++ 1.6 (for 32-bit Windows).
GLPK 4.30 (release date: Aug 13, 2008)
The core LP solver based on the primal simplex method was
re-implemented to allow its further improvements. Currently the
new version provides the same features as the old one, however,
it is a bit faster and more numerically stable.
Some changes were made in the MathProg translator to allow <,
<=, >=, and > on comparing symbolic values. Thanks to Heinrich
Schuchardt <[email protected]> for patches.
Internal routine set_d_eps in the exact LP solver was changed
to prevent approximation errors in case of integral data.
Thanks to Markus Pilz <[email protected]> for bug report.
GLPK 4.29 (release date: Jul 06, 2008)
The configure script was changed to disable all optional
features by default. For details please see file INSTALL.
The following new API routines were added:
glp_erase_prob erase problem object content
glp_read_mps read problem data in MPS format
glp_write_mps write problem data in MPS format
glp_read_lp read problem data in CPLEX LP format
glp_write_lp write problem data in CPLEX LP format
For description of these new routines see the reference manual
included in the distribution.
The following API routines are deprecated:
lpx_read_mps, lpx_read_freemps, lpx_write_mps,
lpx_write_freemps, lpx_read_cpxlp, and lpx_write_cpxlp.
Two bugs were fixed. Thanks to
Anne-Laurence Putz <[email protected]> and
Xypron <[email protected]> for bug report.
GLPK 4.28 (release date: Mar 25, 2008)
The iODBC and MySQL table drivers, which allows transmitting
data between MathProg model objects and relational databases,
were re-implemented to replace a static linking by a dynamic
linking to corresponding shared libraries.
Many thanks to Heinrich Schuchardt <[email protected]>
for the contribution, Rafael Laboissiere <[email protected]>
for useful advices concerning the shared library support under
GNU/Linux, and Vijay Patil <[email protected]> for testing
this feature under Windows XP.
A new optional feature was added to the package. This feature
is based on the zlib data compression library and allows GLPK
API routines and the stand-alone solver to read and write
compressed data files performing compression/decompression "on
the fly" (compressed data files are recognized by suffix `.gz'
in the file name). It may be useful in case of large MPS files
to save the disk space (up to ten times).
The `configure' script was re-implemented. Now it supports the
following specific options:
--with-gmp Enable using the GNU MP bignum library
--without-gmp Disable using the GNU MP bignum library
--with-zlib Enable using the zlib data compression
library
--without-zlib Disable using the zlib data compression
library
--enable-dl Enable shared library support (auto check)
--enable-dl=ltdl Enable shared library support (GNU)
--enable-dl=dlfcn Enable shared library support (POSIX)
--disable-dl Disable shared library support
--enable-odbc Enable using ODBC table driver
--disable-odbc Disable using ODBC table driver
--enable-mysql Enable using MySQL table driver
--disable-mysql Disable using MySQL table driver
For more details please see file INSTALL.
GLPK 4.27 (release date: Mar 02, 2008)
Three new table drivers were added to the MathProg translator:
xBASE built-in table driver, which allows reading and writing
data in .dbf format (only C and N fields are supported);
MySQL table driver, which provides connection to a MySQL
database;
iODBC table driver, which provides connection to a database
through ODBC.
The MySQL and iODBC table drivers were contributed to GLPK by
Heinrich Schuchardt <[email protected]>.
The table driver is a program module which allows transmitting
data between MathProg model objects and external data tables.
For detailed description of the table statement and table
drivers see the document "Using Data Tables in the GNU MathProg
Modeling Language" (file doc/tables.txt) included in the
distribution. Some examples which demonstrate using MySQL and
iODBC table drivers can be found in subdirectory examples/sql.
GLPK 4.26 (release date: Feb 17, 2008)
The table statement was implemented in the GNU MathProg
modeling language. This new feature allows reading data from
external tables into model objects such as sets and parameters
as well as writing results of computations to external tables.
A table is a (unordered) set of records, where each record
consists of the same number of fields, and each field is
provided with a unique symbolic name called the field name.
Currently the GLPK package has the only built-in table driver,
which supports tables in the CSV (comma-separated values) file
format. This format is very simple and supported by almost all
spreadsheets and database management systems.
Detailed description of the table statement and CSV format can
be found in file doc/tables.txt, included in the distribution.
GLPK 4.25 (release date: Dec 19, 2007)
A tentative implementation of Gomory's mixed integer cuts was
included in the branch-and-cut solver. To enable generating
Gomory's cuts the control parameter gmi_cuts passed to the
routine glp_intopt should be set to GLP_ON. This feature is
also available in the solver glpsol through command-line option
'--gomory'. For more details please see the reference manual
included in the distribution.
GLPK 4.24 (release date: Nov 21, 2007)
A tentative implementation of MIR (mixed integer rounding) cuts
was included in the MIP solver. To enable generating MIR cuts
the control parameter mir_cuts passed to the routine glp_intopt
should be set to GLP_ON. This feature is also available in the
stand-alone solver glpsol via command-line option '--mir'. For
more details please see the reference manual included in the
distribution.
The implementation is mainly based on the following two papers:
1. H. Marchand and L. A. Wolsey. Aggregation and mixed integer
rounding to solve MIPs. CORE discussion paper 9839, CORE,
Universite catholique de Louvain, June 1998.
2. G. Andreello, A. Caprara, and M. Fischetti. Embedding cuts
in a Branch&Cut framework. Preliminary draft, October 2003.
MIR cuts can be generated on any level of the search tree that
makes the GLPK MIP solver to be a real branch-and-cut solver.
A bug was fixed in the routine lpx_write_cpxlp. If a variable
x has upper bound and no lower bound, it should appear in the
bounds section as "-inf <= x <= u", not as "x <= u". Thanks to
Enric Rodriguez <[email protected]> for the bug report.
GLPK 4.23 (release date: Oct 28, 2007)
The following new API routines were added:
glp_read_sol read basic solution from text file
glp_write_sol write basic solution to text file
glp_read_ipt read interior-point solution from text file
glp_write_ipt write interior-point solution to text file
glp_read_mip read MIP solution from text file
glp_write_mip write MIP solution to text file
For description of these routines and corresponding file
formats see Chapter "API Routines", Section "Utility routines"
in the reference manual included in the distribution.
Advanced API routine glp_free_env was added. It may be used by
the application program to free all resources allocated by GLPK
routines.
The following three new command-line options were added to the
solver glpsol:
--mipgap tol set relative MIP gap tolerance
-r filename read solution from filename
-w filename write solution to filename
GLPK 4.22 (release date: Sep 19, 2007)
This is a maintainer release.
A bug was fixed in the MIP preprocessor (ios_preprocess_node).
Thanks to Roberto Bagnara <[email protected]> (Department of
Mathematics, University of Parma, Italy) for the bug report.
A bug was fixed in the MIP preprocessor (col_implied_bounds),
due to which constraint coefficients with small magnitude could
lead to wrong implied bounds of structural variables.
A similar bug was fixed in the routine reduce_bounds.
A bug was fixed in the routines glp_set_mat_row and
glp_set_mat_col. (The bug appeared due to incorrect removing
zero elements from the row/column lists.)
A bug was fixed in the API routines lpx_read_mps and
lpx_read_freemps, due to which bounds of type LI specified in
BOUNDS section were incorrectly processed.
A call to standard function vsprintf was replaced by a call to
vsnprintf for security reasons. Many thanks to Peter T. Breuer
<[email protected]> and Rafael Laboissiere <[email protected]>.
GLPK 4.21 (release date: Aug 28, 2007)
Additional reasons for calling the callback routine used in the
MIP solver (glp_intopt) were introduced. Currently the following
reasons are supported:
* request for subproblem selection
* request for preprocessing
* request for row generation
* request for heuristic solution
* request for cut generation
* request for branching
* better integer solution found
A basic preprocessing component used to improve subproblem
formulations by tightening bounds of variables was included in
the MIP solver. Depending on the control parameter pp_tech
passed to the routine glp_intopt the preprocessing can be
performed either on the root level or on all levels (default)
or can be disabled.
Backtracking heuristic used by default in the MIP solver was
changed to the "best local bound".
For more details see Chapter "Advanced API routines", Section
"Branch-and-bound interface routines" in a new edition of the
reference manual included in the distribution.
GLPK 4.20 (release date: Jul 26, 2007)
API routine lpx_integer was replaced by API routine glp_intopt,
which provides equivalent functionality and additionally allows
the application to control the solution process by means of the
user-written callback routine, which is called by the solver at
various points of the branch-and-bound algorithm. Besides, the
new MIP solver allows generating "lazy" constraints and cutting
planes on all levels of the branch-and-bound tree, not only on
the root level. The routine lpx_integer is also still available
for the backward compatibility.
The following new advanced API routines, which may be called
from the B&B callback routine, were included in the package:
glp_ios_reason determine reason for calling callback
routine
glp_ios_get_prob access the problem object
glp_ios_tree_size determine size of the branch-and-bound tree
glp_ios_curr_node determine current active subproblem
glp_ios_next_node determine next active subproblem
glp_ios_prev_node determine previous active subproblem
glp_ios_up_node determine parent subproblem
glp_ios_node_level determine subproblem level
glp_ios_node_bound determine subproblem local bound
glp_ios_mip_gap compute relative MIP gap
glp_ios_heur_sol provide solution found by heuristic
glp_ios_terminate terminate the solution process
For description of these routines see Chapter "Advanced API
routines", Section "Branch-and-bound interface routines" in a
new edition of the reference manual, which was included in the
distribution.
Old version of the integer optimization suite (IOS) as well as
TSP solver tspsol based on it are no longer supported and were
removed from the package.
A minor error in the MIP presolver was fixed; thanks to Graham
Rockwell <[email protected]> for the bug report.
GLPK 4.19 (release date: Jul 05, 2007)
The principal change is upgrading to GPLv3.
A serious bug in the routine glp_del_cols was fixed; thanks to
Cedric[FR] <[email protected]> for the bug report. The bug
appeared because on deleting non-basic columns the basis header
remained valid, however, contained invalid (old) column ordinal
numbers.
A new advanced API routine glp_mem_limit was added.
The case GLP_EBOUND was added to the routine lpx_simplex.
Thanks to Cameron Kellough <[email protected]> for the
bug report.
An API routine lpx_write_pb to write the problem instance in
OPB (pseudo boolean) format format was added. Thanks to Oscar
Gustafsson <[email protected]> for the contribution.
Two new options --wpb and --wnpb were added to glpsol to write
the problem instance in OPB format.
GLPK 4.18 (release date: Jun 25, 2007)
The following new API routines were added:
glp_set_rii set (change) row scale factor
glp_set_sjj set (change) column scale factor
glp_get_rii retrieve row scale factor
glp_get_sjj retrieve column scale factor
glp_simplex solve LP problem with the simplex method
(this routine replaces lpx_simplex, which is
also available for backward compatibility)
glp_init_smcp initialize simplex method control params
glp_bf_exists check if the basis factorization exists
glp_factorize compute the basis factorization
glp_bf_updated check if the basis factorization has been
updated
glp_get_bfcp retrieve basis factorization control params
glp_set_bfcp change basis factorization control params
glp_get_bhead retrieve the basis header information
glp_get_row_bind retrieve row index in the basis header
glp_get_col_bind retrieve column index in the basis header
glp_ftran perform forward transformation
glp_btran perform backward transformation
For description of all these routines see a new edition of the
reference manual included in the distribution.
Type names ulong_t and uldiv_t were changed to glp_ulong and
glp_uldiv to avoid conflicts with standard type names on some
platforms. Thanks to Boris Wirtz <[email protected]>
for the bug report.
Some new examples in the MathProg language were added. Thanks
to Sebastian Nowozin <[email protected]>.
GLPK 4.17 (release date: May 26, 2007)
API routines glp_set_mat_row, glp_set_mat_col, and glp_load_mat
were modified to allow zero constraint coefficients (which are
not stored in the constraint matrix). Note that constraint
coefficients with duplicate row/column indices are not allowed.
Another form of LP basis factorization was implemented in the
package. It is based on LU-factorization of an initial basis
and Schur complement to reflect changes in the basis. Currently
the implementation is incomplete and provides only updating the
factorization on replacing a column of the basis matrix. On API
level the user can set the control parameter LPX_K_BFTYPE to
choose between the folloiwng forms of LP basis factorization to
be used in the simplex method routines:
1) LU + Forrest-Tomlin update;
2) LU + Schur complement + Bartels-Golub update;
3) LU + Schur complement + Givens rotation update.
The GLPK implementation is similar to LUSOL/LUMOD developed by
Michael A. Saunders.
The user can choose the form of LP basis factorzation used by
the simplex method routines by specifying the folloiwng options
of glpsol: --luf, --cbg, --cgr.
GLPK 4.16 (release date: May 05, 2007)
A number of basic GLPK API routines, which now are in the
stable stable, were renamed to be prefixed with 'glp_'. Note
that all these routines are available via their old names
prefixed with 'lpx_' that keeps the downward compatibility with
older versions of the package.
Three new GLPK API routines were added to the package:
glp_version, glp_term_hook, and glp_mem_usage; for more details
see a new edition of the GLPK reference manual included in the
distribution. The routine glp_version reports the actual version
of the GLPK library and also can be used (along with the header
glpk.h) in Autotools specification files to check if the GLPK
library has been installed.