I wanted to quickly grab a random sample of values from a larger array to perform some testing. I didn’t need to worry about the distribution of the sampled data: I just needed n values drawn at random from the array. I used shuffle from numpy.random to randomise the values in my input array. shuffle operates in place, so I made a copy of the array to avoid future problems. I can then simply take a slice of the first n values in the array, which will be a random selection of n values drawn from the original data:

def sample_from_population(a,n):
    """
    Get a random sample of length n from array a
    """
    from numpy.random import shuffle
        
    a_new = a.copy()
    shuffle(a_new)
    
    return a_new[:n]

This can be tested using some example data:

import numpy as np        

data = np.arange(0,10)

print data
print sample_from_population(data,5)

Which yields:

[0 1 2 3 4 5 6 7 8 9]
[5 2 8 1 0]