data example : variable a{1, 0, 1, 0, 1, 0, 1, 0} variable b{0, 1, 0, 1, 0, 1, 0, 1}
can I use spearman method for this analysis?
Asked
Active
Viewed 260 times
0
-
Correlations are used for numeric values, you variables are binary, have a look at https://en.wikipedia.org/wiki/Phi_coefficient. – user2974951 Oct 01 '18 at 06:18
1 Answers
0
Yes, you can apply both Spearman's and Pearson's correlations on the data. I went on and run both in Python using your data:
from scipy.stats.stats import pearsonr
from scipy.stats import spearmanr
a = [1, 0, 1, 0, 1, 0, 1, 0]
b = [0, 1, 0, 1, 0, 1, 0, 1]
print("PEARSON'S: ", pearsonr(a, b))
print("SPEARMAN'S: ", spearmanr(a, b))
The results were:
PEARSON'S: (-1.0, 0.0)
SPEARMAN'S: SpearmanrResult(correlation=-0.9999999999999998, pvalue=2.7369110631344156e-47)
Mohammad Sohaib
- 133
-
thx for your comment, but I'm afraid that it wouldn't be proper for binary variables. What I'm required now I think is advice of statistics – Jungho Kim Oct 01 '18 at 06:38
-
In that case, @user2974951 correctly states that you should take a look at "Phi Coefficient". – Mohammad Sohaib Oct 01 '18 at 06:42
-
Thx again! do you know which package support a phi coeffiecient function? phi() just takes a square matrix. But I should deal with 2 by 1000 matrix. – Jungho Kim Oct 01 '18 at 06:52
-
You can use Matthews Correlation Coefficient available in sklearn in python. – Mohammad Sohaib Oct 02 '18 at 09:59