Re: [Help-glpk] non-official updated version of glpk (4.63 pre-release)
< On Mon, 2017-08-07 at 10:18 -0500, Michael Hennebry wrote: I get a zero-byte document. Sorry. I don't know why that url doesn't work now. Try http://www.ia.pw.edu.pl/~wogrycza/publikacje/artykuly/mymps87.pdf It should work. It does. -- Michael henne...@web.cs.ndsu.nodak.edu "Sorry but your password must contain an uppercase letter, a number, a haiku, a gang sign, a heiroglyph, and the blood of a virgin." -- someeecards ___ Help-glpk mailing list Help-glpk@gnu.org https://lists.gnu.org/mailman/listinfo/help-glpk
Re: [Help-glpk] non-official updated version of glpk (4.63 pre-release)
On Mon, 2017-08-07 at 10:18 -0500, Michael Hennebry wrote: > > I get a zero-byte document. Sorry. I don't know why that url doesn't work now. Try http://www.ia.pw.edu.pl/~wogrycza/publikacje/artykuly/mymps87.pdf It should work. ___ Help-glpk mailing list Help-glpk@gnu.org https://lists.gnu.org/mailman/listinfo/help-glpk
Re: [Help-glpk] non-official updated version of glpk (4.63 pre-release)
On Sun, 6 Aug 2017, Andrew Makhorin wrote: On Sun, 2017-08-06 at 12:13 +0300, Andrew Makhorin wrote: Interesting to note that exactly the same picture (cycling on two bases near the optimum on solving LP with MPSX/370) is described in the paper: W.Ogryczak, On practical stopping rules for the simplex method, Math.Prog.Study 31 (1987), pp.167-174. BTW, this article is freely available at: http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.703.6576=rep1=pdf I get a zero-byte document. -- Michael henne...@web.cs.ndsu.nodak.edu "Sorry but your password must contain an uppercase letter, a number, a haiku, a gang sign, a heiroglyph, and the blood of a virgin." -- someeecards ___ Help-glpk mailing list Help-glpk@gnu.org https://lists.gnu.org/mailman/listinfo/help-glpk
Re: [Help-glpk] non-official updated version of glpk (4.63 pre-release)
On Sun, 2017-08-06 at 12:13 +0300, Andrew Makhorin wrote: > Hi Chris, > > > > > I did some additional checking and it seems that spx_chuzc_sel finds > > an eligible variable with d[j] ~3 times the eps value but no progress > > is made, ending up cycling between two variables. > > > > Interesting to note that exactly the same picture (cycling on two bases > near the optimum on solving LP with MPSX/370) is described in the paper: > W.Ogryczak, On practical stopping rules for the simplex method, > Math.Prog.Study 31 (1987), pp.167-174. BTW, this article is freely available at: http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.703.6576=rep1=pdf > > Andrew Makhorin ___ Help-glpk mailing list Help-glpk@gnu.org https://lists.gnu.org/mailman/listinfo/help-glpk
Re: [Help-glpk] non-official updated version of glpk (4.63 pre-release)
A "smart" LP perturbation was also implemented in the dual simplex solver. This feature is similar to the one implemented in the primal simplex solver (see below). Andrew Makhorin On Sat, 2017-07-01 at 14:58 +0300, Andrew Makhorin wrote: > Please see an updated version of glpk here: > http://sourceforge.net/projects/noumenon/files/tmp/ > (Note that this is *not* an official release.) > > In this release: > > 1. See http://lists.gnu.org/archive/html/help-glpk/2017-06/msg00032.html > > 2. A "smart" LP perturbation was implemented in the primal simplex > solver. Now LP is perturbed only if it is necessary, and by default > perturbation is not activated. > > 3. The sum of primal infeasibilites that appears in the terminal output > of the primal simplex solver (as "inf = ...") now corresponds to the > original bounds of variables. This allows to see how much perturbed LP > solution violates the original bounds. > > 4. The Posix version of glp_time (see glpk/src/env/time.c) was changed > to resolve time_t issue on msys2. > Thanks to Rob Schroederfor bug report. > > > Andrew Makhorin > > > > ___ > Help-glpk mailing list > Help-glpk@gnu.org > https://lists.gnu.org/mailman/listinfo/help-glpk > ___ Help-glpk mailing list Help-glpk@gnu.org https://lists.gnu.org/mailman/listinfo/help-glpk
Re: [Help-glpk] non-official updated version of glpk (4.63 pre-release)
On 25/06/2017 11:06, Andrew Makhorin wrote: Please see an updated version of glpk here: http://sourceforge.net/projects/noumenon/files/tmp/ (Note that this is *not* an official release.) The long-step technique was implemented for phase I of the primal simplex solver. This feature can be enabled by specifying --flip option for glpsol or by specifying glp_smcp.r_test = GLP_RT_FLIP on api level. on cygwin 64 bit, it works fine. $ ./glpsol.exe --flip /pub/devel/glpk/mps/maros-r7.mps GLPSOL: GLPK LP/MIP Solver, v4.63 Parameter(s) specified in the command line: --flip /pub/devel/glpk/mps/maros-r7.mps Reading problem data from '/pub/devel/glpk/mps/maros-r7.mps'... Problem: MAROS-R7 Objective: R0 3137 rows, 9408 columns, 151120 non-zeros 80307 records were read One free row was removed GLPK Simplex Optimizer, v4.63 3136 rows, 9408 columns, 144848 non-zeros Preprocessing... 2152 rows, 5288 columns, 98334 non-zeros Scaling... A: min|aij| = 1.672e-03 max|aij| = 1.000e+00 ratio = 5.981e+02 GM: min|aij| = 4.089e-02 max|aij| = 2.446e+01 ratio = 5.981e+02 EQ: min|aij| = 1.672e-03 max|aij| = 1.000e+00 ratio = 5.981e+02 Constructing initial basis... Size of triangular part is 2152 0: obj = -8.590082000e+06 inf = 9.937e+06 (2152) 255: obj = 3.760574391e+08 inf = 0.000e+00 (0) 52% * 500: obj = 3.331710763e+08 inf = 0.000e+00 (2106) 2 * 1000: obj = 2.857528650e+08 inf = 0.000e+00 (1834) 5 * 1500: obj = 1.687226691e+07 inf = 0.000e+00 (1769) 5 * 2000: obj = 4.909614786e+06 inf = 0.000e+00 (1467) 5 * 2500: obj = 2.780511274e+06 inf = 0.000e+00 (1526) 5 * 3000: obj = 1.699791861e+06 inf = 0.000e+00 (645) 5 * 3500: obj = 1.497291523e+06 inf = 0.000e+00 (28) 4 * 3528: obj = 1.497185166e+06 inf = 0.000e+00 (0) 1 OPTIMAL LP SOLUTION FOUND Time used: 1.6 secs Memory used: 28.6 Mb (29973184 bytes) ___ Help-glpk mailing list Help-glpk@gnu.org https://lists.gnu.org/mailman/listinfo/help-glpk