We often need to rank data in a non-parametric test. We can use the rankdata() function from the scipy.stats module to rank data.

from scipy.stats import rankdata
 
data = [1, 3, 2, 2, 5, 8, 10, 3, 4, 3, 10]
 
ranks = rankdata(data)
 
print("Data: ", data)
print("Ranks: ", ranks)

The output of the above program will be like the following:

Data:  [1, 3, 2, 2, 5, 8, 10, 3, 4, 3, 10]
Ranks:  [ 1.   5.   2.5  2.5  8.   9.  10.5  5.   7.   5.  10.5]

Please note that if the data associated with two or more ranks are the same, we replace the ranks with the average rank. For example, the data associated with ranks 2 and 3 is 2. So, we replace the ranks 2 and 3 with the average ranks (2 + 3)/2 = 2.5.