0

Describe the bug Linear mixed-integer solver problem is failing with CVXOPT (or other solvers, same problems).

Execute "print(cp.installed_solvers())" show these solvers installed.

['CVXOPT', 'ECOS', 'ECOS_BB', 'GLPK', 'GLPK_MI', 'GUROBI', 'OSQP', 'SCIPY', 'SCS']

Execute code

order_thod = 500000
lower_budget = 0
upper_budget = 1.1
x = cp.Variable((M_order.shape[1],M_order.shape[0]), boolean=True)
constraits = [cp.trace(M_order.to_numpy() @ x) <= 
upper_budget*order_thod,cp.trace(M_order.to_numpy() @ x) >= 
lower_budget*order_thod, cp.sum(x, axis=0, keepdims=True) <= 1]
objective = cp.Maximize(cp.trace(M_roi.to_numpy() @ x))
prob = cp.Problem(objective, constraits)
prob.solve(solver=cp.CVXOPT, verbose=True)

Error info

SolverError: The solver CVXOPT is not installed.

Thank you!

1 Answers1

0

I had the same issue (using Anaconda3 and Jupyter Notebook). I restarted the Notebook along with Anaconda because I had pip installed cvxopt inside the Notebook. This fixed the problem. If anyone else has this problem, restart your runtimes!

Jop
  • 90
  • 11