basicgym.envs.simulator.function#

Mathematical Functions used in Synthetic System .

Classes

RewardFunction

Class to define the reward function.

StateTransitionFunction

Class to define the state transition function.
class basicgym.envs.simulator.function.StateTransitionFunction(state_dim, action_dim, random_state=None)[source]#

Class to define the state transition function.

Bases: basicgym.BaseStateTransitionFunction

Imported as: basicgym.envs.simulator.StateTransitionFunction

Tip

Use BaseStateTransitionFunction to define a custom StateTransitionFunction.

Parameters:

  • state_dim (int) – Dimension of the state.

  • action_dim (int) – Dimension of the action (context).

  • random_state (int, default=None (>= 0)) – Random state.

Attributes:
random_state

Methods

step(state, action)

Update the state based on the presented action.
step(state, action)[source]#

Update the state based on the presented action.

Parameters:

  • state (array-like of shape (state_dim, )) – Current state.

  • action (array-like of shape (action_dim, )) – Indicating the action chosen by the agent.

Returns:

state – Next state.

Return type:

array-like of shape (state_dim, )

class basicgym.envs.simulator.function.RewardFunction(state_dim, action_dim, reward_type='continuous', reward_std=0.0, random_state=None)[source]#

Class to define the reward function.

Bases: basicgym.BaseRewardFunction

Imported as: basicgym.envs.simulator.RewardFunction

Tip

Use BaseRewardFunction to define a custom RewardFunction.

Parameters:

  • state_dim (int) – Dimension of the state.

  • action_dim (int) – Dimension of the action (context).

  • reward_type ({"continuous", "binary"}, default="continuous") – Reward type.

  • reward_std (float, default=0.0 (>=0)) – Noise level of the reward. Applicable only when reward_type is “continuous”.

  • random_state (int, default=None (>= 0)) – Random state.

Attributes:
random_state

Methods

mean_reward_function(state, action)

Linear expected immediate reward function.

sample_reward(state, action)

Sample reward.
mean_reward_function(state, action)[source]#

Linear expected immediate reward function.

Parameters:

  • state (array-like of shape (state_dim, )) – State in the RL environment.

  • action (array-like of shape (action_dim, )) – Indicating the action chosen by the agent.

Returns:

mean_reward_function – Expected immediate reward function conditioned on the state and action.

Return type:

float