[Octave-bug-tracker] [bug #66010] Function glpk produces incorrect output

[Hendrik K]

Follow-up Comment #14, bug #66010 (group octave):

I created a patch to improve the glpk behavior (bug #66010):


- align default parameter values to current glpk (version 5.0)
- implement support of msglev setting when calling glpk subroutines
- sort the constraints for better presolving analogue to the glpk provided
code sample

In contrast to my previous thoughts, I aligned the msg level behavior to the
core glpk solver, which does also scaling by default for most cases, but which
only shows the scaling results with msg-level 3. 

Note that it appears that the glpk internal message logging behavior is
somehow inconsistent. For instance a failed pre-solving ONLY shows the error
message "PROBLEM HAS NO PRIMAL FEASIBLE SOLUTION" with message level 3 (all)
and not with msg level 1 as I had expected when calling their main routine
glp_simplex... not much Octave can do about this...

(file #56321)

    _______________________________________________________

Additional Item Attachment:

File name: glpk.patch                     Size: 3KiB
    <https://file.savannah.gnu.org/file/glpk.patch?file_id=56321>


    AGPL NOTICE

These attachments are served by Savane. You can download the corresponding
source code of Savane at
https://git.savannah.nongnu.org/cgit/administration/savane.git/snapshot/savane-6bbbc948d69117a42771112bf22542263dbb32bb.tar.gz


    _______________________________________________________

Reply to this item at:

  <https://savannah.gnu.org/bugs/?66010>

_______________________________________________
Message sent via Savannah
https://savannah.gnu.org/

signature.asc
Description: PGP signature