API Reference

Array

A Cubed array can be created by from_array(), from_zarr(), or by one of the Python Array API Creation Functions.

Array
Array.compute	Compute this array, and any arrays that it depends on.
Array.rechunk	Change the chunking of this array without changing its shape or data.
Array.visualize	Produce a visualization of the computation graph for this array.
compute	Compute multiple arrays at once.
visualize	Produce a visualization of the computation graph for multiple arrays.

IO

from_array	Create a Cubed array from an array-like object.
from_zarr	Load an array from Zarr storage.
store	Save source arrays to array-like objects.
to_zarr	Save an array to Zarr storage.

Chunk-specific functions

apply_gufunc	Apply a generalized ufunc or similar python function to arrays.
map_blocks	Apply a function to corresponding blocks from multiple input arrays.

Non-standardised functions

nanmean	Compute the arithmetic mean along the specified axis, ignoring NaNs.
nansum	Return the sum of array elements over a given axis treating NaNs as zero.

Random number generation

random

Return random floats in the half-open interval [0.0, 1.0).

Runtime

Callback	Object to receive callback events during array computation.
Spec	Specification of resources available to run a computation.
TaskEndEvent	Callback information about a completed task (or tasks).
measure_reserved_mem	Measures the reserved memory use for a given executor runtime.

Executors

beam.BeamDagExecutor	An execution engine that uses Apache Beam.
lithops.LithopsDagExecutor	An execution engine that uses Lithops.
modal_async.AsyncModalDagExecutor	An execution engine that uses Modal's async API.
python.PythonDagExecutor	The default execution engine that runs tasks sequentially uses Python loops.