Randomly picking Items with Given Probabilities

Question

Currently i'm working on a project where I need to code something in Python but stumbled upon some obstacles. I have an excel sheet with 692 items (stock keeping units) with their corresponding sales numbers. The items are in sequence from most sold till least sold. I calculated both the probability and cumulative probability for each item as can be seen in images 1 and 2 below:

excel sheet

excel sheet 2

Based on the cumulative probability and following formula: =INDEX(A$2:A$721;COUNTIF(H$2:H$721;"<="&RAND())+1)

randomly 20 items are selected from the 692 items to generate an order picking list, which can be seen in the following image:

generated pick list

Basically i want to do the same in Python. Thus, select a number of items from a large list based on their probability.

I found the following code, but not sure how to use it:

    import random
def random_pick(some_list, probabilities):
    x = random.uniform(0, 1)
    cumulative_probability = 0.0
    for item, item_probability in zip(some_list, probabilities):
        cumulative_probability += item_probability
        if x < cumulative_probability: break
    return item

It would be great if someone could give me some directions in how to achieve this. Thanks in advance!

You can provide normal weights and cumulated weights via `random.choice` - no need to prog something yourself: [random.choices](https://docs.python.org/dev/library/random.html#random.choices) — Patrick Artner, May 10 '18 at 20:19

score 2 · Answer 1 · answered May 10 '18 at 20:22

Use random.choices and give it some probabilities:

from collections import Counter
from random import choices


data = [1,2,3,4]
prob = [1,1,8,1] # 3 about 8 times more likely to being choosen then 1,2 or 4

c = Counter(choices(data,prob,k=1000)) # generate 1000 and count

print(c)

Output:

Counter({3: 717, 1: 101, 2: 92, 4: 90})  # 3 being taken far more often - about 8:1:1:1 if
                                         # you close both eyes and squint really hard ;)

Randomly picking Items with Given Probabilities

1 Answers1