Python数学函数优化及Float Division By Zero错误排查求助

Hey there! Let's break down why you're hitting that frustrating "Float Division By Zero" error and get your optimization working smoothly.

Why the Error Pops Up #

Your cost function minimF(n) includes the term 10 ** 9 / n — so if n ever equals 0 (or gets extremely close to it), Python will throw that division-by-zero error. The root issue is that minimize_scalar (using its default Brent method) doesn't know your n should be a positive value. It might explore values near or equal to 0 during its search for the minimum, triggering the error.

The Fix: Restrict the Search to Valid Values #

Since n is almost certainly a positive number (I assume it represents something like production quantity or a count that can't be zero), we just need to tell minimize_scalar to only search in a positive range. Also, let's make your code more flexible by using the parameters from your function instead of hardcoding values.

Here's the revised, working code:

import math
from scipy.optimize import minimize_scalar

def function(N, ca, cs, theta):
    # Define the cost function inside to use the passed parameters directly
    def minimF(n):
        return (N / n) * ca + (0.5 * n * cs * theta)
    
    # Minimize only over positive values of n (avoid zero entirely)
    C0_Min = minimize_scalar(minimF, bounds=(1e-6, 1e12), method='bounded')
    print(C0_Min)

# Call the function with your original input values
function(10**9, 9*10**6, 1/25, 180)

Key Changes Explained #

• Bounded Search: We added bounds=(1e-6, 1e12) and set method='bounded' to lock the optimizer into positive n values. The lower bound 1e-6 avoids zero while still allowing very small positive numbers, and the upper bound 1e12 gives enough room for large values if needed.
• Parameterized Cost Function: By nesting minimF inside function, we use the input parameters N, ca, cs, theta directly instead of hardcoding them. This makes your code reusable for different input values.
• Cleaner Arithmetic: I simplified the calculation a bit (e.g., 0.5 instead of 1/2) for readability — the math is exactly the same, just easier to follow.

When you run this revised code, it should successfully find the minimum of your cost function without hitting the division-by-zero error.

内容的提问来源于stack exchange，提问作者Andrei Rus