Performing the non-paired t-test column-wise in my data

Question

When we have two 1-D arrays:

import numpy as np
import scipy.stats as stats    
a=np.array([0.36619718309859156,
     0.32558139534883723,
     0.3333333333333333,
     0.3333333333333333,
     0.2549019607843137,
     0.3695652173913043,
     0.3157894736842105,
     0.3625])

and

b=np.array([0.938456,
 0.3239485723,
 0.300,
 0.8658,
 1.254901137,
 2.3695,
 0.75,
 1.3625])

we can perform the t-test by the following:

stats.ttest_ind(a=a, b=b, equal_var=np.amax([np.var(a),np.var(b)])/np.amin([np.var(a),np.var(b)])<4)

However, I would like to compare the columns of A and B where A and B are 2-D arrays:

A=np.array([[0, 0.375, 0.5, 0.3917],
 [0, 0.333, 0.4, 0.4285],
 [0, 0.27272727, 0.0, 0.2],
 [0.0, 0.25, 0.36365, 0.272],
 [0, 0.285857, 0.4, 0.25],
 [0, 0.416667, 0.33, 0.375],
 [0, 0.28, 0.083, 0.41667],
 [0, 0.2858, 0.25, 0.41666]])

B=np.array([[0, 0.4, 0.333, 0.142],
 [0, 0.33333, 0.4, 0.1111111],
 [0, 0.25, 0.285, 0.333333],
 [0.0, 0.5, 0.380, 0.333],
 [0.0, 0.5, 0.33, 0.375],
 [0, 0.25, 0.294, 0.5],
 [0.0, 0.5, 0.333, 0.2068965],
 [0, 0.5, 0.3846, 0.2]])

ie. I would like to perform t-test on, and compare, the first column of A and the first column of B, then the second column of A and the second column of B, and so on. (I tried specifying the axes but I think I am not sure how to correctly incorporate the equal_var < 4 property in this case.)

score 1 · Accepted Answer · answered Jan 25 '23 at 01:49

You can transpose the data and then go through both at the same time:

def non_paired_t_test(a, b):
    return stats.ttest_ind(a=a, b=b, equal_var=np.amax([np.var(a),np.var(b)])/np.amin([np.var(a),np.var(b)])<4)

for a, b in zip(A.transpose(), B.transpose()):
    print(non_paired_t_test(a, b))

S C · Answer 2 · 2023-01-30T21:56:02.010

1

You can use A[:,i] which denotes all the rows in the i^th column:

for i in range(0,4):
    print(stats.ttest_ind(a=A[:,i], b=B[:,i], equal_var=np.amax([np.var(A[:,i]),np.var(B[:,i])])/np.amin([np.var(A[:,i]),np.var(B[:,i])])<4))

edited Jan 30 '23 at 21:56

answered Jan 25 '23 at 02:00

S C

284
2
9

1

While this code snippet may be the solution, including an explanation really helps to improve the quality of your post. Remember that you are answering the question for readers in the future, and those people might not know the reasons for your code suggestion. – Shawn Hemelstrand Jan 25 '23 at 02:17

Performing the non-paired t-test column-wise in my data

2 Answers2