I am currently co-supervising a high school student on a research project and she is using PySCIPOpt. We would like to use PySCIPOpt to implement some machine learning method on branching.
We are using the problem here https://miplib.zib.de/instance_details_milo-v13-4-3d-3-0.html. We would like to know if there is a function we can call on PySCIPOpt that gives us the coefficient matrix and RHS vector of this problem, to which we can modify some numbers, and resend it through PySCIPOpt to optimize. The purpose of doing this is to generate more training data to be used on a package such as Scikit-learn.
I have looked through the source code and could only find functions such as chgLhs and chgRhs, but this seems more difficult to use than just editing the entries of the coefficient matrix and RHS vector directly.
Thank you for your help!