I am familiar with the ability to load a starting solution for a MIP model by using the mipstart 1 command. This is an attractive feature if you are running multiple solves of the same model with different run tuning options like branchdir. The transition to a second run does port over the prior runs best MIP solution, lets say after a timeout, however the Best Bd value is not preserved and the model starts again with the root relaxation objective value rather than the bestbd found in the first model run.
Therefore, my question is whether there is a way to also import the best bound value with the best MIP solution found in the prior run which may have timed out to avoid the redundant evaluation to get to a solution that meets the tolerance gap specified.
–
To unsubscribe from this group and stop receiving emails from it, send an email to gamsworld+unsubscribe@googlegroups.com.
To post to this group, send email to gamsworld@googlegroups.com.
Visit this group at http://groups.google.com/group/gamsworld.
For more options, visit https://groups.google.com/d/optout.
Ending up solving this problem with utilizing the starting positions of the first model run with either the .fx or .lo commands prior to the model equations in the second run. So, basically consider this, however it would have been nice to import all the variables of the best bound as the starting point, but in most models it is really very little work to define the variables that matter and set them as starting points or answers in the next model run.
Regards,
John A. Ryan john.ryan@orm-tech.com
Cell: 214-878-0728
From: gamsworld@googlegroups.com [mailto:gamsworld@googlegroups.com] On Behalf Of John Ryan
Sent: Sunday, April 27, 2014 11:27 PM
To: gamsworld@googlegroups.com
Subject: Loading a best bound in a second run
I am familiar with the ability to load a starting solution for a MIP model by using the mipstart 1 command. This is an attractive feature if you are running multiple solves of the same model with different run tuning options like branchdir. The transition to a second run does port over the prior runs best MIP solution, lets say after a timeout, however the Best Bd value is not preserved and the model starts again with the root relaxation objective value rather than the bestbd found in the first model run.
Therefore, my question is whether there is a way to also import the best bound value with the best MIP solution found in the prior run which may have timed out to avoid the redundant evaluation to get to a solution that meets the tolerance gap specified.
–
To unsubscribe from this group and stop receiving emails from it, send an email to gamsworld+unsubscribe@googlegroups.com.
To post to this group, send email to gamsworld@googlegroups.com.
Visit this group at http://groups.google.com/group/gamsworld.
For more options, visit https://groups.google.com/d/optout.
–
To unsubscribe from this group and stop receiving emails from it, send an email to gamsworld+unsubscribe@googlegroups.com.
To post to this group, send email to gamsworld@googlegroups.com.
Visit this group at http://groups.google.com/group/gamsworld.
For more options, visit https://groups.google.com/d/optout.