Solution: Geometric
Assuming you have a flipping function like:
def flip(p):
"""Return True with probability p"""
from random import random
return random() < p
we can define a recursive function like:
def geometric(p):
return 1 if flip(p) else 1 + geometric(p)
Let’s play a little with our creation:
def test_geometric(p, n):
from matplotlib import pyplot as plt
import numpy as np
n = int(n)
# k is the number of experiments you want to make
results = [geometric(p) for _ in range(n)]
# the bins you want in your histogram
bins = np.arange(1, max(results) + 2)
_, ax = plt.subplots()
ax.hist(results, bins=bins, density=True, edgecolor="black", align="left")
ax.set_xlabel("Number of tosses until first success (k)")
ax.set_ylabel("Estimated probability")
ax.set_title(f"Geometric distribution samples (p = {p}, n = {n})")
plt.show()
return None
test_geometric(0.3, 1e6)
