I am getting a problem where it says float is not iterable when trying to minimize the following functions in python. I have checked, and the problem seems to lie with how I am using this minimization. Just to clarify, the values being passed through for a and b, will be between 0.0 and 1.0. Any ideas are greatly appreciated!
I have put some of the code here just to show:
def func1(x):
return (x-1)\*\*2
def func2(x):
return math.pow(x-0.5,2)
p1 = 1.0 - a
p2 = 1.0 - b
bound1 = (0.0,p1)
bound2 = (0.0,p2)
x0 = 0.5
result1 = minimize(func1, x0, method='SLSQP', bounds = bound1)#, options={'maxiter':3})
result2 = minimize(func2, x0, method='SLSQP', bounds = bound2)#, options={'maxiter':3})
newX = result1.x
newY = result2.x
The issue is that bounds expects a list of tuples (or a Bounds objects), and here you are just providing a single tuple.
Since your cost functions only have a single argument to apply bounds to, you just need to change your current bounds to a list that contains a single tuple like so:
# bound1 = (0.0,p1)
bound1 = [(0.0,p1)]
# bound2 = (0.0,p2)
bound2 = [(0.0,p2)]
Below is the working code:
import math
from scipy.optimize import minimize
def func1(x):
return (x-1) ** 2
def func2(x):
return math.pow(x-0.5,2)
a, b = .2, .5
p1 = 1.0 - a
p2 = 1.0 - b
bound1 = [(0.0,p1)]
bound2 = [(0.0,p2)]
x0 = 0.5
result1 = minimize(func1, x0, method='SLSQP', bounds = bound1)#, options={'maxiter':3})
result2 = minimize(func2, x0, method='SLSQP', bounds = bound2)#, options={'maxiter':3})
newX = result1.x
newY = result2.x
print(newX, newY) # [0.8] [0.5]