Can you save a CP-SAT model as an MPS file?

[Raphael C]

I have a CP-SAT model I make in Python using 

from ortools.sat.python import cp_model

model = cp_model.CpModel()

for i in range(1000):
        cpVars += [model.NewBoolVar('v%d' % i)]

[...]

model.Maximize(sum(cpVars))

I would like to also run this with gurobi but that isn't working with the current version of ortools.  Is it possible to export the model to an MPS file instead?

[Laurent Perron]

No. CP-SAT supports much more complex constraints than linear constraints.

Laurent Perron | Operations Research | lpe...@google.com | (33) 1 42 68 53 00

--
You received this message because you are subscribed to the Google Groups "or-tools-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to or-tools-discu...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/or-tools-discuss/9847e9be-46b1-41ca-8fb6-0016d2add9dbn%40googlegroups.com.

[Laurent Perron]

Actually, In some really limited cases (if you have only non-reified linear constraint and simple Boolean constraint), using the flag --cp_model_dump_models, we will dump a version of the model in the MpModelProto format that you can later convert to mps.

to enable it, you need to set the cp_model_dump_models C++ flag. See https://github.com/google/or-tools/blob/49b6301e1e1e231d654d79b6032e79809868a70e/examples/tests/pywrapinit_test.py#L26 on how to bridge the python code with the C++ flags.

Once this is done, it will write in /tmp/ a MpModelProto. You can load this in linear solver, and then export to MPS.

As you can imagine, this is only used to debug currently.

--Laurent

Laurent Perron | Operations Research | lpe...@google.com | (33) 1 42 68 53 00

[Stuart Rogers]

Actually, In some really limited cases (if you have only non-reified linear constraint and simple Boolean constraint)

Do these cases include mixed integer linear programs such as a simple facility location problem like the p-median problem?

[Laurent Perron]

It does not care about the type of problems, just the type of constraints. 

To view this discussion on the web visit https://groups.google.com/d/msgid/or-tools-discuss/838289e2-07d6-45ac-a1bc-f4c767daeb1cn%40googlegroups.com.