A very simple MILP via CasADi

casadi_milp.py

from casadi import *
import numpy as np


# optimization variables
x1 = SX.sym('x1')
x2 = SX.sym('x2')
w = [x1, x2]
w0 = [0, 0]
lbw = [-np.inf, -np.inf]
ubw = [np.inf, np.inf]
discrete = [False, True]

# objective function
J = 8.0 * x1 + x2

# constraints
g = []
lbg = []
ubg = []

g += [x1 + 2 * x2]
lbg += [-14.0]
ubg += [np.inf]

g += [-4 * x1 - x2]
lbg += [-np.inf]
ubg += [-33.0]

g += [2 * x1 + x2]
lbg += [-np.inf]
ubg += [20.0]

# Concatenate decision variables and constraint terms
w = vertcat(*w)
g = vertcat(*g)

prob = {'f': J, 'x': w, 'g': g}
solver = nlpsol('solver', 'bonmin', prob, {"discrete": discrete})
# solver = nlpsol('solver', 'ipopt', prob)

sol = solver(x0=vertcat(*w0), lbx=lbw, ubx=ubw, lbg=lbg, ubg=ubg)
print('solution: ', sol['x'])