scipy.optimize.fmin error: setting an array element with a sequence

Question

I tried to fit a sin-function by finding the parameter, which has lowest error-value.

Below is my code:

import numpy as np
import scipy.optimize as opt
from scipy.optimize import leastsq
import matplotlib.pyplot as plt

def func_model(x, para):
    ''' Model: y = a*sin(2*k*pi*x+theta)'''
    a, k, theta = para
    return a*np.sin(2*k*np.pi*x+theta)

def func_noise(x, para):
    a, k, theta = para
    return a*np.sin(2*k*np.pi*x+theta) + np.random.randn(100)

def func_error(para_guess):
    '''error_func'''
    error_sum = 0
    x_seq = np.linspace(-2*np.pi, 0, 100)
    para_fact = [10, 0.34, np.pi/6]
    for x in x_seq:
        error_value = (func_noise(x, para_fact)-func_model(x, para_guess))**2
        error_sum = error_sum + error_value
    return error_sum

para_guess_init = np.array([7, 0.2, 0])
solution = opt.fmin(func_error, para_guess_init) 
print(solution)

But it doesn't work, and said the error: setting an array with a sequence

Traceback:

  File "", line 26, in <module>
    solution = opt.fmin(func_error, para_guess_init)
  File "C:\Users\sun\AppData\Local\Continuum\anaconda3\lib\site-packages\scipy\optimize\optimize.py", line 408, in fmin
    res = _minimize_neldermead(func, x0, args, callback=callback, **opts)
  File "C:\Users\sun\AppData\Local\Continuum\anaconda3\lib\site-packages\scipy\optimize\optimize.py", line 532, in _minimize_neldermead
    fsim[k] = func(sim[k])
ValueError: setting an array element with a sequence.

Can someone help me, thanks in advance

Answer 1

This minimizer expects a scalar function-evaluation to minimize.

Your function func_error returns a vector of size (100,) .

Compare your line:

error_value = (func_noise(x, para_fact)-func_model(x, para_guess))**2

with for example:

error_value = np.sum(np.square(
                               func_noise(x, para_fact)-func_model(x, para_guess)))

although i would prefer (objective changes!):

error_value = np.linalg.norm(
                             func_noise(x, para_fact)-func_model(x, para_guess))