CODEWITHSUNDEEP
pythonnumpy
SnG
SnG

Reputation: 402

how to randomly sample from a numpy array of probabilities?

I have a numpy array like this:

arr = [.2 , .1, .2, .3, .2 ]

Now what I want to do is sample from this array, such that the probability i get a certain index is dependent on the probability at the index. So for example, the probability i get index 3 is .3

Anyone know any nifty ways to do this?

Upvotes: 7

Views: 6314

Answers (1)

Alex
Alex

Reputation: 239

You can do weighted sampling with a discrete probability distribution using np.random.choice by providing the sampling distribution as a parameter p:

import numpy as np

x = [.2 , .1, .2, .3, .2 ]

# sample from `x` 100 times according to `x`
n_samples = 100
samples = np.random.choice(x, n_samples, p=x)

Upvotes: 6

Related Questions