Can not read freemps file

[mingodad]

Hello ! Cbc can not read this mps file from miplib2017 https://miplib.zib.de/instance_details_maritime-jg3d9.html, it seems that it's in freemps format but the CoinMpsIO::readMps parser can not detect it as freemps and give this error:

cbc maritime-jg3d9.mps.gz solve
Welcome to the CBC MILP Solver 
Version: 2.10.5 
Build Date: Jul 28 2020 

command line - cbc maritime-jg3d9.mps.gz solve (default strategy 1)
At line 17 NAME maritime-jg3d9
At line 18 ROWS
At line 126641 COLUMNS
At line 754614 RHS
At line 790363 BOUNDS
No match for column   at line 790364 <  UP bnd b1 1 >
At line 827336 ENDATA
Problem maritime-jg3d9 has 126621 rows, 92261 columns and 627969 elements
Coin0008I maritime-jg3d9 read with 1 errors
There were 1 errors on input
** Current model not valid
Total time (CPU seconds):       0.21   (Wallclock seconds):       0.22

GLPK, HiGHS can read it without problem. Cheers !

[mingodad]

Also from netlib2017 this mps files fail for a similar reason neos-5044663-wairoa.mps.gz, rvb-sub.mps.gz

[jjhforrest]

I have no intention of changing the code to allow neos-5044663-wairoa.mps.gz to be read

The image in error is

 C0063540  LB_full              0   $ empty column

and there is a comment at beginning of file

• Format: Fixed MPS

so it should adhere to the original MPS format and not allow $ to start a comment.

maritime-jg3d9.mps.gz is fine if defined as in free format with FREE on name card. Fixed MPS format allows blanks in names, but I think it safe to go to free format after COLUMNS section if there are no blanks in names. So I will make that modification in master.

John Forrest

On 07/10/2020 12:38, Domingo Alvarez Duarte wrote:

Also from netlib2017 this mps files fail for a similar reason neos-5044663-wairoa.mps.gz, rvb-sub.mps.gz

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/coin-or/Cbc/issues/342#issuecomment-704877083, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABWJYHGFYSV6D5D746ZOU6TSJRHM5ANCNFSM4SHGR2QA.

[mingodad]

Thank you John ! I'm not sure if also it's possible to look/fix for this (cplex can handle then) https://www.gams.com/latest/docs/UG_LanguageFeatures.html#UG_LanguageFeatures_IndicatorConstraints

Bad image at line xxxx < INDICATORS > allcolor58.mps.gz amaze22012-06-28i.mps.gz amaze22012-07-04i.mps.gz cvrpa-n64k9vrpi.mps.gz cvrpb-n45k5vrpi.mps.gz cvrpp-n16k8vrpi.mps.gz cvrpsimple2i.mps.gz elitserienhandball11i.mps.gz elitserienhandball13i.mps.gz elitserienhandball14i.mps.gz elitserienhandball3i.mps.gz fillomino7x7-0i.mps.gz fjspeasy01i.mps.gz gfd-schedulen180f7d50m30k18-16i.mps.gz gfd-schedulen25f5d20m10k3i.mps.gz gfd-schedulen55f2d50m30k3i.mps.gz ghoulomb4-9-10i.mps.gz k1mushroomi.mps.gz l2p2i.mps.gz mappingmesh3x3mpeg2i.mps.gz mario-t-hard5i.mps.gz mrcpspj30-15-5i.mps.gz mrcpspj30-17-10i.mps.gz mrcpspj30-53-3i.mps.gz mspsphard01i.mps.gz mspsphard03i.mps.gz oocsp-racks030e6cci.mps.gz oocsp-racks030f7cci.mps.gz pizza27i.mps.gz pizza78i.mps.gz rpp22falsei.mps.gz shipschedule3shipsi.mps.gz shipschedule6shipsmixi.mps.gz shipschedule8shipsmixuci.mps.gz splice1k1i.mps.gz stoch-vrpvrp-s5v2c8vrp-v2c8i.mps.gz supportcase21i.mps.gz supportcase27i.mps.gz supportcase28i.mps.gz

Bad image at line xxxx < LAZYCONS > diameterc-mstc-v20a190d5i.mps.gz diameterc-msts-v40a100d5i.mps.gz

Bad image at line xxx < xx xxx > ivu59.mps.gz rvb-sub.mps.gz

Cheers !

[tkralphs]

I'll take this up with the MIPLIB 2017 committee, as there was quite a bit of effort put into ensuring adherence of the distributed files to an agreed-upon standard.

[jjhforrest]

Domingo,

INDICATORS is not just a format issue - I will look and see if it is useful for Cbc.

Ihave put into master (should be safe in stable as well) changes to fix the free format issue. The code had already recognized that the file was free format - but had to make sure bounds section treated correctly.

Not sure about $ as a comment. So put it in conditionally if COIN_ALLOW_DOLLAR_AS_COMMENT defined

John Forrest

On 07/10/2020 16:34, Domingo Alvarez Duarte wrote:

Thank you John ! I'm not sure if also it's possible to look/fix for this (cplex can handle then) https://www.gams.com/latest/docs/UG_LanguageFeatures.html#UG_LanguageFeatures_IndicatorConstraints

Bad image at line xxxx < INDICATORS > allcolor58.mps.gz amaze22012-06-28i.mps.gz amaze22012-07-04i.mps.gz cvrpa-n64k9vrpi.mps.gz cvrpb-n45k5vrpi.mps.gz cvrpp-n16k8vrpi.mps.gz cvrpsimple2i.mps.gz elitserienhandball11i.mps.gz elitserienhandball13i.mps.gz elitserienhandball14i.mps.gz elitserienhandball3i.mps.gz fillomino7x7-0i.mps.gz fjspeasy01i.mps.gz gfd-schedulen180f7d50m30k18-16i.mps.gz gfd-schedulen25f5d20m10k3i.mps.gz gfd-schedulen55f2d50m30k3i.mps.gz ghoulomb4-9-10i.mps.gz k1mushroomi.mps.gz l2p1i.mps.gz l2p2i.mps.gz mappingmesh3x3mpeg2i.mps.gz mario-t-hard5i.mps.gz mrcpspj30-15-5i.mps.gz mrcpspj30-17-10i.mps.gz mrcpspj30-53-3i.mps.gz mspsphard01i.mps.gz mspsphard03i.mps.gz oocsp-racks030e6cci.mps.gz oocsp-racks030f7cci.mps.gz pizza27i.mps.gz pizza78i.mps.gz rpp22falsei.mps.gz shipschedule3shipsi.mps.gz shipschedule6shipsmixi.mps.gz shipschedule8shipsmixuci.mps.gz splice1k1i.mps.gz stoch-vrpvrp-s5v2c8vrp-v2c8i.mps.gz supportcase21i.mps.gz supportcase27i.mps.gz supportcase28i.mps.gz

Bad image at line xxxx < LAZYCONS > diameterc-mstc-v20a190d5i.mps.gz diameterc-msts-v40a100d5i.mps.gz

Bad image at line xxx < xx xxx > ivu59.mps.gz rvb-sub.mps.gz

Cheers !

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/coin-or/Cbc/issues/342#issuecomment-705018161, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABWJYHBBHLB2KY6D2QGNXXDSJSDCBANCNFSM4SHGR2QA.