Exporting the Reduced (Pre-solved) model
AnsweredLet us call the model AFTER pre-solving as the "Reduced Model"... How can I get this reduced model? Is it possible to export this to a file from command line? Doesn't matter whether I can get this done with or without running the problem to completion.
I have seen this: https://support.gurobi.com/hc/en-us/articles/360024738352-How-does-presolve-work- and this: https://www.gurobi.com/documentation/9.5/refman/py_model_presolve.html
But unfortunately they don't say how to get the Reduced Model from the command line.
Why do I need the reduced model? So that I don't need to use the presolver repeatedly.
Thanks.
-
Hi Prabhu,
you could try:
reduced_model = model.presolve()
reduced_model.write("reduced_model.lp")Best regards
Jonasz0 -
Hi Jonasz, Thank you, but, as I asked in my previous post, is it possible to do all this from the command line?
0 -
Yes, exactly in the same fashion.
You first generate the presolved model:
> reduced_model = model.presolve()
and then you store it in your desired location:
> reduced_model.write("reduced_model.lp")
Best regards,
Jonasz1 -
Please also see the Knowledge Base article How does presolve work? for more information.
0 -
Please note that you need to be in an interactive Python shell to run those commands. Exporting the presolved model does not work directly through the command line with gurobi_cl.
Cheers,
Matthias1 -
Thanks Matthias.. Yes, I should have made it clear that I was talking about the LINUX shell command line (gurobi_cl) -- NOT the Python shell :-)
Anyway, I got the job done by writing Python code, by following this post by Glockner at Stack Overflow. Here's my modified code:
#!/opt/gurobi951/linux64/bin/python3.7
(Your python3.7 may be located in a different directory.
So replace the above line accordingly.)
import gurobipy as gp
from gurobipy import GRB
modello = gp.read("Original_Model.lp")
reduced_model = modello.presolve()
reduced_model.write("reduced_model.lp")Thanks everyone.
0
Please sign in to leave a comment.
Comments
6 comments