Input an array of data for Scipy.optimize.minimize

Asked Mar 26 '21 at 22:52

Active Mar 27 '21 at 01:02

Viewed 231 times

import scipy.optimize as opt
    
def fun1(u,es):
    theta = u[0]
    ret = (np.fft.fft(es_storex.flatten()+1j*es_storey.flatten()))*np.exp(1j*theta)

    return (ret)

x = np.array([3.])
results = opt.minimize(fun1, x)

I'm trying to do a phase correction to my FFT signal. esx_store[0] and esy_store[0] are arrays of data. Is there a way I can use scipy.optimize.minimize to minimize these data, and get the optimal value for theta to perform phase correction to my data?

With flatten() it still gives the error " ValueError: The user-provided objective function must return a scalar value."

edited Mar 27 '21 at 01:02

asked Mar 26 '21 at 22:52

Dimuthu Kodituwakku

Does your `fun1` return a `float`? Try the function with some values check the `type`. – Max Pierini Mar 27 '21 at 17:50
@MaxPierini yes it returns float numbers. Among them, there are complex numbers too. Is that a problem? – Dimuthu Kodituwakku Mar 27 '21 at 18:03
Try `return ret.real` because complex numbers are a sort of "tuple", say `complex(2, 3)` is `(2+3j)` – Max Pierini Mar 27 '21 at 18:14
Thank you but it still returns the same error :( – Dimuthu Kodituwakku Mar 28 '21 at 18:01
The problem is much more complex, take a look to the basis of scipy optmization, you have to better define your function (it can't return complex number or arrays, but only scalars) https://realpython.com/python-scipy-cluster-optimize/#using-the-optimize-module-in-scipy – Max Pierini Mar 28 '21 at 19:31

Input an array of data for Scipy.optimize.minimize

0 Answers0