Any performance difference of CP solver with Python compared to Java/C++ ?

[Nicolas Blandamour]

Hello,

I would like to find realisable solutions of a MIP problem using the constraint solver of OR-TOOLS, but before coding my problem, I'm wondering concerning the langage I should use.

Is there any chance that the CP solver could be slower if I modelize my problem with Python API contrary to the same modelisation coded with Java or C++ ?

I guess that both use the underlying C++ engine of the solver, but I wonder if the ease of modeling with the Python API (in particular for constraint declaration, summing over a dimension without having to create auxiliary variables, overloaded operators,...) could result in a slower resolution compared to a low-level implementation in Java or C++.

Thanks

[Jing Jiang]

I also have a related question: the reference page has a C++ API list. Is there a Python API list?

在 2017年6月22日星期四 UTC-5上午3:01:29，Nicolas Blandamour写道：

[Laurent Perron]

I do not expect any difference in the solving time.

And usually, the time to build the model is negligible.

So, no.

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-discuss+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.