"""Watch the closest-rule policy in MuJoCo's viewer window, five targets in a row.

Run: python3 site/ml/foundations/examples/gymnasium_watch.py
render_mode="human" makes the task open MuJoCo's viewer and show every step. Close the window to stop early.
"""
import gymnasium as gym

import gymnasium_basics as G

env = gym.make("ArmReach-v0", render_mode="human")
for seed in range(5):
    steps = G.run_episode(env, G.closest_policy, seed)
    print(f"seed {seed}: {len(steps)} commands, {'reached' if steps[-1][3] else 'time limit'}")
env.close()
