Numerical Computation

import epochly
import random
@epochly.optimize(level=2)
def monte_carlo_pi(iterations):
    inside = 0
    for _ in range(iterations):
        x = random.random()
        y = random.random()
        if x*x + y*y <= 1:
            inside += 1
    return 4 * inside / iterations
# Estimate pi with 100 million iterations
pi_estimate = monte_carlo_pi(100_000_000)
print(f"Pi ≈ {pi_estimate}")

import epochly
@epochly.optimize
def rolling_mean(data, window):
    results = []
    for i in range(len(data) - window + 1):
        window_sum = 0
        for j in range(window):
            window_sum += data[i + j]
        results.append(window_sum / window)
    return results
# 2M data points, window of 50
result = rolling_mean(data, 50)
# Faster with Epochly (Level 3 parallel or Level 4 GPU)