Navigation

rlmesh.spaces.MultiDiscrete

classExperimentalfrom rlmesh.spaces import MultiDiscrete[source]

Vector of discrete dimensions.

MultiDiscrete(
    nvec: Sequence[int] | SpaceSpec,
    dtype: IntDTypeLike = 'int64',
)

Parameters

NameTypeDescription
nvecSequence[int] | SpaceSpecPer-dimension counts, or an existing native SpaceSpec.
dtypeIntDTypeLikeInteger dtype name.

Attributes

Methods

nvec

attribute[source]
nvec: list[int] | None

spec

property[source]
property spec: SpaceSpec

Native space specification backing this wrapper.

kind

property[source]
property kind: str

Native RLMesh space kind.

shape

property[source]
property shape: list[int]

Native shape for tensor-like spaces.

dtype

property[source]
property dtype: str

Element dtype reported by the native spec.

seed

function[source]
seed(seed: int | None = None) -> int | None

Seed the native sampler for this space.

sample

function[source]
sample() -> OutputT

Sample a value from this space.

contains

function[source]
contains(value: object) -> bool

Return whether a value is contained in this space.

to_gymnasium_space

function[source]
to_gymnasium_space() -> object

Convert this wrapper into a Gymnasium space.