使用卡方找到拟合优度

Question

I'm attempting to use scipy.stats chisquare tool to be able to determine goodness of fit between two arrays. I have used the np.polyfit tool with a fit of 10 to be able to find a fit between two arrays, now I'm just trying to figure out how to use the chisquare method to determine how good that fit is. How do I correctly use scipy's chisquare tool in this situation?

data = np.loadtxt("location of data")

x = data[:,0] #defining the first column as x
y = data[:,1] #defining the second column as y

fit = np.polyfit(x, y, 10)
p = np.poly1d(fit)

import scipy
from scipy.stats import chisquare

Answer 1

You can directly test the distributions by passing them in the chisquare module as group1 and group2. X being group1. Y being group2. Here group1 represents freq_observed and group2 represents freq_expected. Set degrees of freedom as optional argument ddof or you can let the model select it from the number of elements

import scipy.stats as stat
stat.chisquare(np.array(group1), np.array(group2))

is the way to use it. Now you can use your fitted data against both of these columns for the test individually as group1 being the fitted data and group2 being the independent X, Y.