gym_wrappers

Source code: tianshou/env/gym_wrappers.py

---

class ContinuousToDiscrete(env: Env, action_per_dim: int | list[int])

Bases: ActionWrapper

Gym environment wrapper to take discrete action in a continuous environment.

Parameters:

• env (gym.Env) – gym environment with continuous action space.

• action_per_dim – number of discrete actions in each dimension of the action space.

Constructor for the action wrapper.

action(act: ndarray) → ndarray

Returns a modified action before env.step() is called.

Args:

action: The original step() actions

Returns:

The modified actions

class MultiDiscreteToDiscrete(env: Env)

Bases: ActionWrapper

Gym environment wrapper to take discrete action in multidiscrete environment.

Parameters:

env (gym.Env) – gym environment with multidiscrete action space.

Constructor for the action wrapper.

action(act: ndarray) → ndarray

Returns a modified action before env.step() is called.

Args:

action: The original step() actions

Returns:

The modified actions

class TruncatedAsTerminated(env: Env)

Bases: Wrapper

A wrapper that set terminated = terminated or truncated for step().

It’s intended to use with gym.wrappers.TimeLimit.

Parameters:

env (gym.Env) – gym environment.

Wraps an environment to allow a modular transformation of the step() and reset() methods.

Args:

env: The environment to wrap

step(act: ndarray) → tuple[Any, SupportsFloat, bool, bool, dict[str, Any]]

Uses the step() of the env that can be overwritten to change the returned data.