Options
All
  • Public
  • Public/Protected
  • All
Menu

Class TapeAlgorithmicEnv

A 1 dimensional algorithmic env.

Environment observations wrap around tape with an additional observation for the out-of-bounds case.

The MIN_LENGTH of the tape is 5 plus a random number between 0 and 2.

Hierarchy

Implements

Index

Constructors

Properties

Methods

Constructors

constructor

Properties

MIN_LENGTH

MIN_LENGTH: number

MOVEMENTS

MOVEMENTS: string[]

action_space

action_space: DiscreteTuple

agentActions

agentActions: number[]

base

base: number

charmap

charmap: string[]

cursor

cursor: number

done

done: boolean

inputData

inputData: any

observation_space

observation_space: Discrete

reward

reward: number

reward_range

reward_range: Discrete

target

target: []

targetLength

targetLength: number

Methods

close

  • close(): void

genInputData

  • genInputData(size: number): any

getStrObs

  • getStrObs(): string

move

render

  • render(): void

renderHTML

  • renderHTML(): string

reset

  • reset(): tf.Tensor

seed

  • seed(seed: number): void

  • Parameters

    • seed: number

    Returns void

Abstract setTarget

  • setTarget(input_data: any): void

  • Parameters

    • input_data: any

    Returns void

step

  • step(action: actionSpace): [tf.Tensor, number, boolean, __type]

  • Parameters

    Returns [tf.Tensor, number, boolean, __type]

toObs

  • toObs(): any

  • Returns any

Generated using TypeDoc