Python - Estimate Distribution of a Population Category through Randomly Sampling From it and find the number of samples to arrive at a decent approximation of the pop distribution.

python_sample_size_estimator_simulation.py

# How to determine sample size through simulation
import random
from collections import Counter

things = ['camel', 'horse', 'donkey', 'mule', 'bulldozer', 'chain', 'machete', 'tool' ,'border collie', 'widget']
weights = [random.random() * 1000 for _ in range(10)]

assert(len(things) == len(weights))

universe = random.choices(things, weights=weights, k=100000)

uni_counts = Counter(universe).most_common(10)
for x, counts in uni_counts:
    print(x, counts/100000)
    
    
border collie 0.21839
tool 0.16326
horse 0.14853
mule 0.11849
machete 0.10348
bulldozer 0.09127
chain 0.06403
donkey 0.05163
camel 0.02267
widget 0.01825


for item, frequency in Counter(random.sample(universe, k=400)).most_common(10):
    print(item, frequency/400)
    
border collie 0.2125
tool 0.1625
horse 0.1525
machete 0.1175
mule 0.1125
bulldozer 0.0825
chain 0.065
donkey 0.0425
camel 0.0425
widget 0.01