looking for some solutions? You are welcome.

SOLVED: Cplex performance difference in Python vs Matlab


I am developing an algorithm, which needs a QP solver. For that purpose, I use CPLEX solver in Python. I would like to ask you for help regarding the following few issues, which I met while using CPLEX.

  1. I am using CPLEX with Wrapper (CVXPY) and I noticed, that there is a significant performance difference between MATLAB and Python versions of CPLEX (I am handling matrix multiplication with numpy). Is there any way to solve this issue?

  2. I think Wrapper is causing this problem, that is why I am trying to use CPLEX directly without Wrapper. However, I could not find any example on IBM CPLEX website. Could anyone please provide an example or a direction on how to solve QP problem with CPLEX directly.

  3. According to IBM CPLEX manual, in order to set the QP matrices, I have to put them row by row in index and value pairs. However, since the matrices I have to put are quite big, it is hard to set all of them one by one. Is there any way to put matrices directly into the solver like in MATLAB?

Posted in S.E.F
via StackOverflow & StackExchange Atomic Web Robots

No comments: