Memoisation demo in python

memoisation_demo.py

# This is the memoization demo.
# First a wrapper is created, which can be used for each call of the fibonacci function.
# The wrapper will save (or cache) the result of each function call - index for a given set of arguments and keyword arguments.
from functools import wraps
from time import perf_counter


def memoize(func):  # The memoisation needs to take a function, since it's a decorator.
	cache = {}

	@wraps(func)
	def wrapper(*args, **kwargs):  # Needs the arguments and keyword arguments - see info about making custom decorators.
		key = str(args) + str(kwargs)  # To make the result identifiable from its arguments - we want to store the results by the arguments given.

		if key not in cache:  # Add to cache, if it's not already in there.
			cache[key] = func(*args, **kwargs)

		return cache[key]

	return wrapper


@memoize
def memoized_fibo(n) -> int:  # This is the function that uses the memoisation.
	if n < 2:
		return n
	return memoized_fibo(n - 1) + memoized_fibo(n - 2)



def non_memoized_fibo(n) -> int:  # This is the function that doesn't use the memoisation.
	if n < 2:
		return n
	return non_memoized_fibo(n - 1) + non_memoized_fibo(n - 2)


if __name__ == '__main__':
	n = 35  # We'll try to calculate the n'th fibonacci number
	print(f"Will try to find the {n}'th Fibonacci number.")
	# This is the memoized run
	start = perf_counter()
	print(f"Result, memoized:\t\t\t{memoized_fibo(n)}")
	end = perf_counter()
	print(f"Duration, memoized:\t\t\t{end - start:.20f} seconds")

	# This is the non-memoized run
	start = perf_counter()
	print(f"Result, non-memoized:\t\t{non_memoized_fibo(n)}")
	end = perf_counter()
	print(f"Duration, non-memoized:\t\t{end - start:.20f} seconds")