I am optimizing f(x) using scipy.optimize.minimize and the value of f(x) cannot be negative but scipy.optimize returns a min value as negative

Question

The minimized function value of f(x) should be greater than or equal to 0. However, scipy.optimize.minimize provides a negative value of f(x). Is there any way we could put a constraint on the function value?

If your function can be negative, then I don't see why it's a problem that a *minimizer* finds a negative value. It's unclear why or what you want to constrain exactly: the function parameters, the function value (in which case: change the function to always return a positive value), or something else. Can you provide a (simplified) concrete example of your problem? — 9769953, Mar 15 '23 at 12:35
Please provide enough code so others can better understand or reproduce the problem. — Community, Mar 16 '23 at 01:01

score 0 · Answer 1 · answered Mar 15 '23 at 15:17

You could try to solve:

 min z
 z = f(x)
 z >= 0

However, nonlinearly constrained problems are often more difficult to solve than unconstrained or linearly constrained problems.

If the model is essentially a root finding problem (i.e. we can get very close to zero), you can try:

  min f(x)^2

I am optimizing f(x) using scipy.optimize.minimize and the value of f(x) cannot be negative but scipy.optimize returns a min value as negative

1 Answers1