method: SciPy.optimize

user_16824394 0
Viewed 61

I wrote the following code to minimize objective where the variables x1, x2, x3 are between 0.8-1.2 and the sum of x1, x2, x3 is a number.

import numpy as np
from scipy.optimize import minimize 


def objective(x):
    x1 = x[0]
    x2 = x[1]
    x3 = x[2]
    return 1/x1 +2/x3+ 0.5/x2

def constraint1(x):
    return 7*x[0]+8*x[1]+4*x[2]-5

x0 = [1,0.8, 1.1]
b = (0.8, 1.2)
bnds = (b,b,b)
con1 = {'type':'eq', 'fun' :constraint1}

solution = minimize(objective, x0, method ='', bounds= bnds, constraints= con1)

print (solution)

I used the BFGS function which gave me an array that is outside the range of 0.8-1.2 while SLSQP gave me success=false. Is there something wrong with my code or am I using the wrong method?

scipy optimization python
1 Answers
  • BFGS is not for constrained problems
  • With SLSQP I saw: message: 'Iteration limit exceeded'
  • After increasing the iteration limit: message: 'Inequality constraints incompatible'
  • That makes sense. How can 7*x[0]+8*x[1]+4*x[2]=5 if the lower bounds on the variables are 0.8? I.e. we know in advance that 7*x[0]+8*x[1]+4*x[2]>=15.2.

I suggest providing this method with feasible problems.

user_56255340
Related