olfactory-search

A simple implementation of the olfactory search problem as a Gymnasium environment.

Installation

git clone https://github.com/kevinmcareavey/olfactory-search.git
cd olfactory-search
python3 -m pip install .

Getting started

import gymnasium
from olfactory_search.envs import SMALLER_ISOTROPIC_DOMAIN, SMALLER_WINDY_DOMAIN

seed = None # if seed is not None then episodes will always start in the same initial state
parameters = SMALLER_WINDY_DOMAIN
env = gymnasium.make("olfactory_search/Windy2D-v0", parameters=parameters, max_episode_steps=parameters.T_max)

# Uncomment the following if you specify the original location
# options = {
#     'agent_location': [4, 2],
#     'source_location': [2, 4],
# }
options = None
observation, info = env.reset(seed=seed, options=options)
print(f"initial observation = {observation}, info = {info}")
for _ in range(1000):
    action = env.action_space.sample()
    observation, reward, terminated, truncated, info = env.step(action)
    print(f"action = {action}, observation = {observation}, reward = {reward}, done = {terminated or truncated}, info = {info}")

    if terminated or truncated:
        print(f"episode done")
        observation, info = env.reset(seed=seed, options=options)
        print(f"\ninitial observation = {observation}, info = {info}")

print("simulation done")
env.close()