Python Native Implementation of Mixed Integer Linear Programming

37 Views Asked by At

Is it possible to have pure Python implementation of Mixed Integer Linear Programming, something similar to mip, pulp, cvxpy, etc. -

but such simple as https://github.com/ispaneli/lippy - it is Simplex-method based implementations of such methods as Branch and bound, Cutting-plane and another methods. Those methods are full-integer solvers, which does not work for bigger number of variables - some methods go to infinite loops, another - returns zeroes-only solve.

There is a requirement to solve a problem, which returns BOTH fractional and integer variables.

Probably it is something simplex-like problem, where some of solved variables should be integers.

The thing is at this case it's impossible to import c-based libs like mip.

So question is: Are there similar solutions as github.com/ispaneli/lippy , but for problems where some of solved variables should be integers?

Those lippy lib methods do not work, return zeroes or stacks infinite loops