atarihns 0.0.1

Sum tree and min tree implementation used particularly in reinforcement learning algorithms.

Atari Human Normalized Score

Lightweight helper for computing Atari human-normalized scores (HNS) using published human and random baselines.

Installation

• From PyPI (once published): pip install atarihns
• From source: clone the repo and run pip install .

Usage

from atarihns.helpers import calculate_hns, get_human_score, get_random_score

env = "Pong-v5"
agent_score = 15.0

human = get_human_score(env)
random = get_random_score(env)
hns = calculate_hns(env, agent_score)

print(f"{env} human: {human}, random: {random}, hns: {hns:.3f}")

Notes

• Baseline scores for ALE environments are defined in atarihns.constants.ATARI_SCORES as (random, human).
• Helper functions raise KeyError if an environment name is missing from the table.