Describes the storage layout of a tensorstore.TensorStore.

Constructors¶

ChunkLayout(json: Any)

Constructs from the JSON representation.

ChunkLayout(*, rank: int | None = None, ...)

Constructs from component parts.

Classes¶

class Grid

Describes a regular grid layout for write/read/codec chunks.

Accessors¶

to_json() → Any

Converts to the JSON representation.

rank : int

Number of dimensions in the index space.

ndim : int

Alias for rank.

inner_order : tuple[int, ...] | None

Permutation specifying the element storage order within the innermost chunks.

inner_order_soft_constraint : tuple[int, ...] | None

Permutation specifying soft constraint on the element storage order.

grid_origin : tuple[int | None, ...] | None

Hard constraints on the grid origin.

grid_origin_soft_constraint : tuple[int | None, ...] | None

Soft constraints on the grid origin.

write_chunk : ChunkLayout.Grid

Chunk grid for efficient writes.

read_chunk : ChunkLayout.Grid

Chunk grid for efficient reads.

codec_chunk : ChunkLayout.Grid

Chunk grid used by the codec.

Setters¶

update(*, rank: int | None = None, ...) → None

Adds additional constraints.

Chunk templates¶

write_chunk_template : IndexDomain

Chunk offset and shape for efficient writes.

read_chunk_template : IndexDomain

Chunk offset and shape for efficient reads.

Comparison operators¶

__eq__(other: ChunkLayout) → bool

Compares two chunk layouts for equality.