On Wednesday, June 10, 2015 at 2:37:00 PM UTC+3, M.H. wrote: > > 1. Do you know if octave and GLPK run smoothly on windows? > > Yes, both of them run on Windows, however I experienced several > 'corner-cases'. E.g., I am not aware if modern Octave GUI (introduced in > 3.8.x) is already available for Windows without Cygwin. I also had some > problems with running it on Windows 8, but I must say, I wasn't too > persistent to solve that problems. On the other hand, I worked with stable > enough configuration of Octave 3.6.4 on Windows 7 - everything was fine. GLPK > is both a standalone program for solving LP/MILP and a dynamic library for > that purpose. Both of its forms are available on Windows. GLPK is actually > used by Octave to solve LP/MILP in the form of glpk() function. So, you don't > have to think about installing GLPK separately, if you have a working Octave > installation, you already have glpk() function in it which uses GLPK dynamic > library under the hood. However, you may prefer to use GLPK without Octave, > it may be beneficial in some cases. So, to clarify my original proposal, > Octave OR (not "+") GLPK may be used. (I'am also aware that there are plenty > of other LP/MILP solvers with permissive licenses - brightest minds can > actually brew them in the contest time frame ;) - I just did not happen to > use them.) > > > > 2. Is it possible to call such methods from another language? > 1. Octave. Well, I didn't use anything of the kind, so my browsing of Octave > documentation won't be more informative than anyone else's. However, there is > always quick-and-dirty solution, that is to generate ".m" file, run it by > subprocess.call (if it is Python) and parse the output. > 2. GLPK. Yes, you can certainly use the mentioned quick-and-dirty way with > the standalone form of GLPK, or use dll, or use java wrapper for that dll, or > even Python module (PULP, as far as I remember - but that's quite another > story). > > > > 3. Is it possible to have complete accuracy in octave/glpk when handling > > fractions? > A very good question! As far as I understand, it is impossible with Octave's > glpk() function. But if you use GLPK API directly, there is a function that > solves LP in rational numbers - I didn't use it and the documentation says > that it is "tentative implementation". > > > Kind regards, > Andrew >
Great! Thanks for all the useful info. I think I'll try out GLPK in the near future. -- You received this message because you are subscribed to the Google Groups "Google Code Jam" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/google-code/3d73fc4e-e7a1-4f0e-a120-64fa1bdd1003%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
