CLP Optimal vs GLOP Abnormal #2859
Unanswered
JinseGeerts
asked this question in
Linear Solver questions
Replies: 15 comments 1 reply
|
Usually, glop is much more pedantic with numerical errors. |
0 replies
|
This is what I get when running the solve binary |
0 replies
|
Thanks for checking. So the solver itself is acutally able to solve it? Is there anything in the post processing that can cause the abnormal, or is this an issue with the C# wrapper? |
0 replies
|
I do not know.
Are you sure you are picking the 9.1 assembly ?
Laurent Perron | Operations Research | ***@***.*** | (33) 1 42 68 53
00
Le mer. 20 oct. 2021 à 13:28, JinseGeerts ***@***.***> a
écrit :
… Thanks for checking.
So the solver itself is acutally able to solve it? Is there anything in
the post processing that can cause the abnormal, or is this an issue with
the C# wrapper?
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<#2859 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/ACUPL3K6R62T6HQM2LMZUUTUH2RUBANCNFSM5GJVLUHA>
.
Triage notifications on the go with GitHub Mobile for iOS
<https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675>
or Android
<https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>.
|
0 replies
|
I double checked. Yes, I'm using version 9.1.9490 |
0 replies
|
The exported model, from the internal representation is working fine. |
0 replies
|
Where should this extra log end up? Nothing extra appears in my debug window. Is there any config I can/need to set to get this log? |
0 replies
|
It should be printed on stdout
Laurent Perron | Operations Research | ***@***.*** | (33) 1 42 68 53
00
Le mer. 20 oct. 2021 à 14:23, JinseGeerts ***@***.***> a
écrit :
… Where should this extra log end up? Nothing extra appears in my debug
window. Is there any config I can/need to set to get this log?
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<#2859 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/ACUPL3OVLPCKJA2S2CZQZVLUH2YC7ANCNFSM5GJVLUHA>
.
Triage notifications on the go with GitHub Mobile for iOS
<https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675>
or Android
<https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>.
|
1 reply
|
Okay I'll try and capture this and come back to you |
|
Took me a while to figure out why I wasn't getting any output, but I fixed it. |
0 replies
|
Have you changed the tolerances of the solver ?
Laurent Perron | Operations Research | ***@***.*** | (33) 1 42 68 53
00
Le mer. 20 oct. 2021 à 17:14, JinseGeerts ***@***.***> a
écrit :
… Took me a while to figure out why I wasn't getting any output, but I fixed
it.
I've added my output as attachment.
solver-output.zip
<https://github.com/google/or-tools/files/7382370/solver-output.zip>
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<#2859 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/ACUPL3NGUCCIXZ75KRFK5UDUH3MGPANCNFSM5GJVLUHA>
.
Triage notifications on the go with GitHub Mobile for iOS
<https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675>
or Android
<https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>.
|
0 replies
|
No settings were adjusted. I only used solver.Solve() |
0 replies
|
Do you know the value the binary uses? Maybe I can set it to that and try again? |
0 replies
|
try 1e-5 for the primal tolerance. I do not think the binary sets anything.
Laurent Perron | Operations Research | ***@***.*** | (33) 1 42 68 53
00
Le mer. 20 oct. 2021 à 22:41, JinseGeerts ***@***.***> a
écrit :
… Do you know the value the binary uses? Maybe I can set it to that and try
again?
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<#2859 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/ACUPL3NPYPBY22MGE4I2LL3UH4SO3ANCNFSM5GJVLUHA>
.
Triage notifications on the go with GitHub Mobile for iOS
<https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675>
or Android
<https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>.
|
0 replies
|
That did not fix it, unfortunately |
0 replies
|
So, I've been looking around for a bit and I've encounter the following issue that seems to follow my symptoms. #1868 The issue is about the Java wrapper though, so the solution stated there doesn't seem to work. I see the following available in C# However after some quick research, this does not seem to be the same setting as 'SolutionFeasibilityTolerance' |
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
Hi guys
I've got another weird issue with a big difference between the CLP Solver and the GLOP Solver.
For the same model, CLP returns an Optimal result within a few seconds (~5), while GLOP returns an Abnormal result after a bit more time (~30 seconds).
As I understand it, abnormal means that the model is malformed somehow, right?
Is there a difference in how you should construct a model for the two solvers?
I'm running the latest available nuget package for the C# wrapper (9.1.9490 time of writing)
I've added the model as attachment. Any ideas would be greatly appreciated.
Kind regards
Jinse
Model.zip
All reactions