CODEWITHSUNDEEP
pythonstatisticssimulationdistributionsampling
Jacky
Jacky

Reputation: 9

How can I define a discrete distribution with values and possibilities I defined?

I have two lists, both the same size, let's call them values and possibility. I want to choose of the values list with discrete probability distribution given by possibility. possibility[i] corresponds to the probability of choosing value[i].

How can I do this in the simplest way in Python?

Upvotes: 0

Views: 1664

Answers (2)

user254873
user254873

Reputation: 176

Take a look into package SciPy, namely scipy.stats.rv_discrete (https://docs.scipy.org/doc/scipy/reference/generated/scipy.stats.rv_discrete.html).

>>> from scipy import stats
>>> import numpy as np
>>> xk = np.arange(7)
>>> pk = (0.1, 0.2, 0.3, 0.1, 0.1, 0.0, 0.2)
>>> myRV = stats.rv_discrete(name="MyRandomVar", values=(xk, pk))
>>> sample = myRV.rvs(size=5) # sample == array([2, 2, 0, 6, 1])

Upvotes: 1

Alistair Carscadden
Alistair Carscadden

Reputation: 1228

Here's a solution that generates a random number (0, 1) and then subtracts the probabilities to find which value to return.

import random

def choose_with_probability(probs, values):
    val = random.random()
    for x in enumerate(values):
        val -= probs[x[0]]
        if(val < 0):
            return x[1]

# probs = [0.5, 0.2, 0.1, 0.1, 0.05, 0.05]
# values = ['a', 'b', 'c', 'd', 'e', 'f']

The result is 'a' being returned 50% of the time, 'b' 20%, 'c' 10% etc.

random.random() returns a number from 0 to 1.

enumerate() creates a numbered list from the argument list. e.g. ['a', 'b', 'c'] becomes [[0, 'a'], [1, 'b'], [2, 'c']].

I used the enumeration index to find the probability of each value since both lists correspond 1:1.

Upvotes: 0

Related Questions