Try our new documentation site.
sensitivity.py
#!/usr/bin/python
# Copyright 2013, Gurobi Optimization, Inc.
# Simple MIP sensitivity analysis example.
# For each integer variable, fix it to its lower and upper bound
# and check the impact on the objective.
import sys
from gurobipy import *
if len(sys.argv) < 2:
print('Usage: sensitivity.py filename')
quit()
a = gurobi.read(sys.argv[1])
a.optimize()
a.params.outputFlag = 0
# Extract variables from model
avars = a.getVars()
# Iterate through binary variables in model
for i in range(len(avars)):
v = avars[i]
if v.vType == GRB.BINARY:
# Create clone and fix variable
b = a.copy()
bv = b.getVars()[i]
if v.x - v.lb < 0.5:
bv.lb = bv.ub
else:
bv.ub = bv.lb
b.optimize()
if b.status == GRB.status.OPTIMAL:
objchg = b.objVal - a.objVal
if objchg < 0:
objchg = 0
print('Objective sensitivity for variable %s is %g' % \
(v.varName, objchg))
else:
print('Objective sensitivity for variable %s is infinite' % \
v.varName)
