Problems with scipy minimize

Question

I’m trying to use the minimize function to optimize two parameters for a power law fit. I am getting these errors so far. I already obtained variables that I wanted using curve fit, but I specifically wanted to use the minimize function on a chi_square function.

Here is the code:

import matplotlib.pyplot as plt
import math
import scipy.optimize as scp
from scipy.optimize import curve_fit
import numpy as np

x = []
y1 = []
y2 = []
y3 = []
y4 = []
y5 = []
y6 = []
y7 = []
y8 = []
y9 = []
y10 = []
y_av = []
unc = []
logx = []
logy = []


with open('data1.txt', 'r') as data_file:
    for line in data_file:
        lines = [i for i in line.split()]
        x.append(float(lines[0]))
        y1.append(float(lines[1]))
        y2.append(float(lines[2]))
        y3.append(float(lines[3]))
        y4.append(float(lines[4]))
        y5.append(float(lines[5]))
        y6.append(float(lines[6]))
        y7.append(float(lines[7]))
        y8.append(float(lines[8]))
        y9.append(float(lines[9]))
        y10.append(float(lines[10]))

#print(x)
#print(x)
#print(y1)
#print(y2)
#print(y3)
#print(y4)
#print(y5)
#print(y6)
#print(y7)
#print(y8)
#print(y9)
#print(y10)

length = len(x)
dof = length-2

for i in range(0, length,1):
    y_av.append(float((y1[i]+y2[i]+y3[i]+y4[i]+y5[i]+y6[i]+y7[i]+y8[i]+y9[i]+y10[i])/10)) #Need to find way to generalize # of y data sets
    

#print(length)
#print(y_av)
    
plt.title("Test")
plt.xlabel('xdata')
plt.ylabel('ydata')
plt.yticks(y_av)
plt.plot(x, y_av, marker = 'o', c = 'g')
    
plt.show

for i in range(0, length,1):
    unc.append(math.sqrt(((y1[i]-y_av[i])**2 + (y2[i]-y_av[i])**2 + (y3[i]-y_av[i])**2 + (y4[i]-y_av[i])**2 + (y5[i]-y_av[i])**2 + (y6[i]-y_av[i])**2 + (y7[i]-y_av[i])**2 + (y8[i]-y_av[i])**2 + (y9[i]-y_av[i])**2 + (y10[i]-y_av[i])**2)/10))
    
    
#print(unc)

def power_fit(x,A,p):
         return A*x**p

plt.errorbar(x, y_av, yerr = unc, fmt = 'ko', label = 'data')

(params,cov)=scp.curve_fit(power_fit,x,y_av,sigma = unc)

#print(params)
#print(cov)

param_err = np.sqrt(np.diagonal(cov))
#print(param_err)

A1 = params[0]
a1_err = param_err[0]
p1 = params[1]
p1_err = param_err[1]

#print(A1)
#print(A1_err)
#print(p1)
#print(p1_err)


plt.plot(x, power_fit(x,*params), 'r--')

chi_sq = np.sum(((y_av-A1*x**p1)/unc)**2)
#print(chi_sq)
chi_sq_red = chi_sq/dof
#print(chi_sq_red)

def chi_square_ols(A, p):
    model = A*np.power(x,p)
    return np.sum((np.power(y_av-model,2)))

x0 = [1.00,1.00]`your text`
args = [x,y_av]
min_chi = scp.minimize(chi_square_ols(), x0=x0, args = args)

print(min_chi)

Running that, I get an error: TypeError: chi_square_ols() missing 2 required positional arguments: ‘A’ and ‘p’

So I switch the minimize function to : min_chi = scp.minimize(chi_square_ols(A,p), x0=x0, args = args)

and get this error: NameError: name ‘A’ is not defined

I am new-ish to python, and am not sure if I should declare these variables outside of the function to call them?

Any help debugging this would be much appreciated. Thanks!

Asked By: eird19

||

Source

Answer 1

Change this:

def chi_square_ols(A, p):
    model = A*np.power(x,p)
    return np.sum((np.power(y_av-model,2)))

x0 = [1.00,1.00]`your text`
args = [x,y_av]
min_chi = scp.minimize(chi_square_ols(), x0=x0, args = args)

print(min_chi)

to this:

def chi_square_ols(params, x, y):
    A, p = params
    model = A*np.power(x, p)
    return np.sum((np.power(y - model, 2)))

x0 = [1.0, 1.0]
args = (x, y_av)

min_chi = scp.minimize(chi_square_ols, x0=x0, args=args)

print(min_chi)

Answered By: Alon Alush

Problems with scipy minimize

Question:

Answers: