mtenv.wrappers package

Submodules

mtenv.wrappers.env_to_mtenv module

Wrapper to convert an environment into multitask environment.

class mtenv.wrappers.env_to_mtenv.EnvToMTEnv(env: gym.core.Env, task_observation_space: gym.spaces.space.Space)

Bases: mtenv.core.MTEnv

Wrapper to convert an environment into a multitak environment.

Parameters

• env (Env) – Environment to wrap over.

• task_observation_space (Space) – Task observation space for the resulting multitask environment.

classmethod class_name() → str

close() → Any

Override close in your subclass to perform any necessary cleanup.

Environments will automatically close() themselves when garbage collected or when the program exits.

get_task_obs() → Union[str, int, float, numpy.ndarray]

Get the current value of task observation.

Environment returns task observation everytime we call step or reset. This function is useful when the user wants to access the task observation without acting in (or resetting) the environment.

Returns

Return type

TaskObsType

get_task_state() → Any

Return all the information needed to execute the current task again.

This function is useful when we want to set the environment to a previous task.

Returns

For more information on task_state, refer Task State.

Return type

TaskStateType

render(mode: str = 'human', **kwargs: Dict[str, Any]) → Any

Renders the environment.

reset(**kwargs: Dict[str, Any]) → Dict[str, Union[numpy.ndarray, str, int, float]]

Reset the environment to some initial state and return the observation in the new state.

The subclasses, extending this class, should ensure that the environment seed is set (by calling seed(int)) before invoking this method (for reproducibility). It can be done by invoking self.assert_env_seed_is_set().

Returns

For more information on multitask observation returned by the environment, refer MultiTask Observation.

Return type

ObsType

reset_task_state() → None

Sample a new task_state and set the environment to that task_state.

For more information on task_state, refer Task State.

sample_task_state() → Any

Sample a task_state.

task_state contains all the information that the environment needs to switch to any other task.

The subclasses, extending this class, should ensure that the task seed is set (by calling seed(int)) before invoking this method (for reproducibility). It can be done by invoking self.assert_task_seed_is_set().

Returns

For more information on task_state, refer Task State.

Return type

TaskStateType

seed(seed: Optional[int] = None) → List[int]

Set the seed for the environment’s random number generator.

Invoke seed_task to set the seed for the task’s random number generator.

Parameters

seed (Optional[int], optional) – Defaults to None.

Returns

Returns the list of seeds used in the environment’s random number generator. The first value in the list should be the seed that should be passed to this method for reproducibility.

Return type

List[int]

set_task_state(task_state: Any) → None

Reset the environment to a particular task.

task_state contains all the information that the environment needs to switch to any other task.

Parameters

task_state (TaskStateType) – For more information on task_state, refer Task State.

property spec

step(action: Union[str, int, float, numpy.ndarray]) → Tuple[Dict[str, Union[numpy.ndarray, str, int, float]], float, bool, Dict[str, Any]]

Execute the action in the environment.

Parameters

action (ActionType) –

Returns

Tuple of multitask observation, reward, done, and info. For more information on multitask observation returned by the environment, refer MultiTask Observation.

Return type

StepReturnType

property unwrapped

Completely unwrap this env.

Returns

The base non-wrapped gym.Env instance

Return type

gym.Env

mtenv.wrappers.multitask module

Wrapper to change the behaviour of an existing multitask environment.

class mtenv.wrappers.multitask.MultiTask(env: mtenv.core.MTEnv)

Bases: mtenv.core.MTEnv

Wrapper to change the behaviour of an existing multitask environment

Parameters

env (MTEnv) – Multitask environment to wrap over.

assert_env_seed_is_set() → None

Check that the env seed is set.

assert_task_seed_is_set() → None

Check that the task seed is set.

get_task_obs() → Union[str, int, float, numpy.ndarray]

Get the current value of task observation.

Environment returns task observation everytime we call step or reset. This function is useful when the user wants to access the task observation without acting in (or resetting) the environment.

Returns

Return type

TaskObsType

get_task_state() → Any

Return all the information needed to execute the current task again.

This function is useful when we want to set the environment to a previous task.

Returns

For more information on task_state, refer Task State.

Return type

TaskStateType

reset() → Dict[str, Union[numpy.ndarray, str, int, float]]

Reset the environment to some initial state and return the observation in the new state.

The subclasses, extending this class, should ensure that the environment seed is set (by calling seed(int)) before invoking this method (for reproducibility). It can be done by invoking self.assert_env_seed_is_set().

Returns

For more information on multitask observation returned by the environment, refer MultiTask Observation.

Return type

ObsType

reset_task_state() → None

Sample a new task_state and set the environment to that task_state.

For more information on task_state, refer Task State.

sample_task_state() → Any

Sample a task_state.

task_state contains all the information that the environment needs to switch to any other task.

The subclasses, extending this class, should ensure that the task seed is set (by calling seed(int)) before invoking this method (for reproducibility). It can be done by invoking self.assert_task_seed_is_set().

Returns

For more information on task_state, refer Task State.

Return type

TaskStateType

seed(seed: Optional[int] = None) → List[int]

Set the seed for the environment’s random number generator.

Invoke seed_task to set the seed for the task’s random number generator.

Parameters

seed (Optional[int], optional) – Defaults to None.

Returns

Returns the list of seeds used in the environment’s random number generator. The first value in the list should be the seed that should be passed to this method for reproducibility.

Return type

List[int]

seed_task(seed: Optional[int] = None) → List[int]

Set the seed for the task’s random number generator.

Invoke seed to set the seed for the environment’s random number generator.

Parameters

seed (Optional[int], optional) – Defaults to None.

Returns

Returns the list of seeds used in the task’s random number generator. The first value in the list should be the seed that should be passed to this method for reproducibility.

Return type

List[int]

set_task_state(task_state: Any) → None

Reset the environment to a particular task.

task_state contains all the information that the environment needs to switch to any other task.

Parameters

task_state (TaskStateType) – For more information on task_state, refer Task State.

step(action: Union[str, int, float, numpy.ndarray]) → Tuple[Dict[str, Union[numpy.ndarray, str, int, float]], float, bool, Dict[str, Any]]

Execute the action in the environment.

Parameters

action (ActionType) –

Returns

Tuple of multitask observation, reward, done, and info. For more information on multitask observation returned by the environment, refer MultiTask Observation.

Return type

StepReturnType

mtenv.wrappers.ntasks module

Wrapper to fix the number of tasks in an existing multitask environment.

class mtenv.wrappers.ntasks.NTasks(env: mtenv.core.MTEnv, n_tasks: int)

Bases: mtenv.wrappers.multitask.MultiTask

Wrapper to fix the number of tasks in an existing multitask environment to n_tasks.

Each task is sampled in this fixed set of n_tasks.

Parameters

• env (MTEnv) – Multitask environment to wrap over.

• n_tasks (int) – Number of tasks to sample.

reset_task_state() → None

Sample a new task_state from the set of n_tasks tasks and set the environment to that task_state.

For more information on task_state, refer Task State.

sample_task_state() → Any

Sample a task_state from the set of n_tasks tasks.

task_state contains all the information that the environment needs to switch to any other task.

The subclasses, extending this class, should ensure that the task seed is set (by calling seed(int)) before invoking this method (for reproducibility). It can be done by invoking self.assert_task_seed_is_set().

Returns

For more information on task_state, refer Task State.

Return type

TaskStateType

mtenv.wrappers.ntasks_id module

Wrapper to fix the number of tasks in an existing multitask environment and return the id of the task as part of the observation.

class mtenv.wrappers.ntasks_id.NTasksId(env: mtenv.core.MTEnv, n_tasks: int)

Bases: mtenv.wrappers.ntasks.NTasks

Wrapper to fix the number of tasks in an existing multitask environment to n_tasks.

Each task is sampled in this fixed set of n_tasks. The agent observes the id of the task.

Parameters

• env (MTEnv) – Multitask environment to wrap over.

• n_tasks (int) – Number of tasks to sample.

get_task_obs() → Any

Get the current value of task observation.

Environment returns task observation everytime we call step or reset. This function is useful when the user wants to access the task observation without acting in (or resetting) the environment.

Returns

Return type

TaskObsType

get_task_state() → Any

Return all the information needed to execute the current task again.

This function is useful when we want to set the environment to a previous task.

Returns

For more information on task_state, refer Task State.

Return type

TaskStateType

reset() → Dict[str, Union[numpy.ndarray, str, int, float]]

Reset the environment to some initial state and return the observation in the new state.

The subclasses, extending this class, should ensure that the environment seed is set (by calling seed(int)) before invoking this method (for reproducibility). It can be done by invoking self.assert_env_seed_is_set().

Returns

For more information on multitask observation returned by the environment, refer MultiTask Observation.

Return type

ObsType

sample_task_state() → Any

Sample a task_state from the set of n_tasks tasks.

task_state contains all the information that the environment needs to switch to any other task.

The subclasses, extending this class, should ensure that the task seed is set (by calling seed(int)) before invoking this method (for reproducibility). It can be done by invoking self.assert_task_seed_is_set().

Returns

For more information on task_state, refer Task State.

Return type

TaskStateType

set_task_state(task_state: Any) → None

Reset the environment to a particular task.

task_state contains all the information that the environment needs to switch to any other task.

Parameters

task_state (TaskStateType) – For more information on task_state, refer Task State.

step(action: Union[str, int, float, numpy.ndarray]) → Tuple[Dict[str, Union[numpy.ndarray, str, int, float]], float, bool, Dict[str, Any]]

Execute the action in the environment.

Parameters

action (ActionType) –

Returns

Tuple of multitask observation, reward, done, and info. For more information on multitask observation returned by the environment, refer MultiTask Observation.

Return type

StepReturnType

mtenv.wrappers.sample_random_task module

Wrapper that samples a new task everytime the environment is reset.

class mtenv.wrappers.sample_random_task.SampleRandomTask(env: mtenv.core.MTEnv)

Bases: mtenv.wrappers.multitask.MultiTask

Wrapper that samples a new task everytime the environment is reset.

Parameters

env (MTEnv) – Multitask environment to wrap over.

reset() → Dict[str, Union[numpy.ndarray, str, int, float]]

Reset the environment to some initial state and return the observation in the new state.

The subclasses, extending this class, should ensure that the environment seed is set (by calling seed(int)) before invoking this method (for reproducibility). It can be done by invoking self.assert_env_seed_is_set().

Returns

For more information on multitask observation returned by the environment, refer MultiTask Observation.

Return type

ObsType

Module contents