Jussi Kujala jukujala

jukujala / moneyh.py

Created May 16, 2018 19:46

calculate annualised return R on financial investment that has returned 11% over 28 monthly investements of $1

	import numpy as np
	from scipy.optimize import minimize_scalar

	def moneyh(R):
	""" save 1$ every month for 28 month

	Args:
	R: annualised yearly interest rate

	Returns:

jukujala / python_blackjack_simulator.py

Last active February 20, 2023 10:09

Blackjack simulator written in Python. API can act as environment for RL agents by passing custom play strategy function to `game(...)`

	""" Blackjack simulator, works for RL
	* Copied and modified from:
	https://gist.github.com/mjhea0/5680216
	* See function game(get_player_action) for instructions.
	* Requires Python 3.7
	* Handles 10M simulations in < 30 mins at Macbook Pro
	* Interactive play: python blackjack_simu.py
	* Might not implement blackjack rules perfectly
	"""
	import random

jukujala / fit_lognorm_vs_many_data_distributions.py

Last active June 26, 2024 11:13

Generate data from different distributions, fit a log-normal and show how much fitted mean overvaluates data mean.

	"""

	Generate data from different distributions, fit a log-normal and
	show how much fitted mean overvaluates data mean.
	Overvaluation 0.05 = 5% higher fitted log-norm mean than data mean.

	Copy-paste of output:

	* log normal data
	- mean(empirical data): 4.22, mean(fitted long-norm): 4.26, overvaluation: 0.01