CODEWITHSUNDEEP
pythonarraysnumpyrandom
OHHH
OHHH

Reputation: 1051

How to use NumPy to create two subarrays from the randomized contents of a source array?

I have an array A, I want to create arrays B and C.

Basically, I want to randomly find 100 elements in A, put them in B and the rest in C.

If I use numpy.random.choice I can easily create B by just extracting all the elements in A that match the indices in the random list, but I would have to go through A again in order to find all the values that are not in B and put them in C. This works, but maybe there is a built-in function that can do this for me.

Is there a cheaper way?

Upvotes: 1

Views: 99

Answers (2)

arajshree
arajshree

Reputation: 624

You can use this, numpy inbuilt function

import numpy as np
import random
a = np.array([1,2,3,4,5,6,7,8,9,10]);
b = np.array(random.sample(a.values,3));
c = np.setdiff1d(a,b)

Upvotes: 0

user2314737
user2314737

Reputation: 29337

You could select elements from the array using indices and create a boolean mask like this:

indices = np.random.choice(len(a),3) # pick indices at random from array a
mask = np.ones(a.shape,dtype=bool)   # create boolean mask
mask[indices] = False                
a[indices]     # random elements from array 
a[mask]        # rest of array

Upvotes: 3

Related Questions