Class SmoothMazeRLEnvironment

Smooth maze environment

Hierarchy (View Summary)

RLEnvironmentBase
- SmoothMazeRLEnvironment

Index

Constructors

Properties

__map _epoch _goal_size _height _map_resolution _max_step _orient _points _position _reward _rotate _velocity _width

Accessors

actions epoch map reward states

Methods

clone reset sample_action setState state step test

Constructors

constructor

new SmoothMazeRLEnvironment(
width: number,
height: number,
): SmoothMazeRLEnvironment
Parameters
- width: number
  Area width
- height: number
  Area height
Returns SmoothMazeRLEnvironment
Overrides RLEnvironmentBase.constructor
- Defined in lib/rl/maze.d.ts:9

Properties

__map

__map: any[][]

_epoch

_epoch: number

_goal_size

_goal_size: number[]

_height

_height: number

_map_resolution

_map_resolution: number[]

_max_step

_max_step: number

_orient

_orient: number

_points

_points: any[]

_position

_position: any[]

_reward

_reward: { fail: number; goal: number; step: number; wall: number }

_rotate

_rotate: number

_velocity

_velocity: number

_width

_width: number

Accessors

actions

get actions(): number[][]
Returns number[][]
- Defined in lib/rl/maze.d.ts:27

epoch

get epoch(): number
Epoch

Returns number
Inherited from RLEnvironmentBase.epoch
- Defined in lib/rl/base.d.ts:73

map

get map(): any[][]
Returns any[][]
- Defined in lib/rl/maze.d.ts:29

reward

set reward(value: any): void
Reward
Parameters
- value: any
  Reward object
Returns void
Inherited from RLEnvironmentBase.reward
- Defined in lib/rl/base.d.ts:78

states

get states(): (RLRealRange | RLIntRange)[]
Returns (RLRealRange | RLIntRange)[]
- Defined in lib/rl/maze.d.ts:28

Methods

clone

clone(): RLEnvironmentBase
Returns cloned environment.

Returns RLEnvironmentBase
Cloned environment
Inherited from RLEnvironmentBase.clone
- Defined in lib/rl/base.d.ts:83

reset

reset(): any[]
Reset environment.

Returns any[]
Overrides RLEnvironmentBase.reset
- Defined in lib/rl/maze.d.ts:30

sample_action

sample_action(agent: any): any[]
Sample an action.
Parameters
- agent: any
  Agent
Returns any[]
Sampled action
Inherited from RLEnvironmentBase.sample_action
- Defined in lib/rl/base.d.ts:130

setState

setState(state: any): void
Set new state.
Parameters
- state: any
  New state
Returns void
Overrides RLEnvironmentBase.setState
- Defined in lib/rl/maze.d.ts:32

state

state(): any[]
Returns current state.

Returns any[]
Current state
Overrides RLEnvironmentBase.state
- Defined in lib/rl/maze.d.ts:31

step

step(
action: any[],
agent: any,
): { done: boolean; invalid?: boolean; reward: number; state: any[] }
Do action and returns new state.
Parameters
- action: any[]
  Actions to be performed by the agent
- agent: any
  Agent
Returns { done: boolean; invalid?: boolean; reward: number; state: any[] }
state, reward, done
Inherited from RLEnvironmentBase.step
- Defined in lib/rl/base.d.ts:106

test

test(state: any, action: any): { done: boolean; reward: number; state: any[] }
Do actioin without changing environment and returns new state.
Parameters
- state: any
  Environment state
- action: any
  Actions to be performed by the agent
Returns { done: boolean; reward: number; state: any[] }
state, reward, done
Overrides RLEnvironmentBase.test
- Defined in lib/rl/maze.d.ts:33