warning lp file contains small (< 1e-13) coefficients, ignored

65 views
Skip to first unread message

moati

unread,
Jan 17, 2017, 4:18:25 PM1/17/17
to Gurobi Optimization
Hi,

When I read the attached model using gurobi 7.0.0 I get the above mentioned error!

Any idea why is that although I don't have any such coefficients!

Thanks,

Model_warning.lp

Daniel Espinoza

unread,
Jan 18, 2017, 8:56:30 AM1/18/17
to Gurobi Optimization
Moati,

You do have a couple of `0` coefficients in the constraint matrix, and then the warning is correct, those coefficients will be dropped from the constraints.

The warning is generic, any coefficient with absolute value under 1e-13 are ignored.

Best
Daniel

moati

unread,
Jan 18, 2017, 10:08:58 AM1/18/17
to Gurobi Optimization
Hi Daniel,

Thanks for your reply.

If you check the attached model, you can find '0' coefficients in the constraint matrix as well but the same version of Gurobi does not give any warning!

Is there a way to know which coefficient the optimizer is complaining of?

Regards,
No warning.lp

Renan Garcia

unread,
Jan 18, 2017, 11:40:52 AM1/18/17
to gur...@googlegroups.com
Moati,

Thanks for reporting this issue, as it helped us find a bug in our LP parser. The warning is being issued because the same variable appears multiple times in some constraints. For example, the first constraint was:

  x_00_00 + ... + x_00_62 + x_00_63 + x_00_63 >= 1

Gurobi can handle this case just fine, but the warning was issued incorrectly. This will be fixed in a future release.

Thanks again,
Renan

--

---
You received this message because you are subscribed to the Google Groups "Gurobi Optimization" group.
To unsubscribe from this group and stop receiving emails from it, send an email to gurobi+un...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
<No warning.lp>

moati

unread,
Jan 18, 2017, 2:28:19 PM1/18/17
to Gurobi Optimization
Thanks for pointing out the problem and glad that it was of help to you :)

The warning disappeared once I fixed that constraint.

Thanks once again,
Moati
Reply all
Reply to author
Forward
0 new messages