Cubed

What's Changed
* Improve separation between layers by tomwhite in https://github.com/tomwhite/cubed/pull/314
* Fix Modal deprecation of `f.call(...)` by tomwhite in https://github.com/tomwhite/cubed/pull/320
* Disable the Hypthosis deadline when running Array API tests. by tomwhite in https://github.com/tomwhite/cubed/pull/325
* Allow dtype to be passed to blockwise and reduction functions by tomwhite in https://github.com/tomwhite/cubed/pull/321
* Array API backend by tomwhite in https://github.com/tomwhite/cubed/pull/317
* Update examples to use Python 3.9 by tomwhite in https://github.com/tomwhite/cubed/pull/330
* Re-write `arange` to use `map_direct`, which avoids writing a Zarr array by tomwhite in https://github.com/tomwhite/cubed/pull/332


**Full Changelog**: https://github.com/tomwhite/cubed/compare/0.11.0...0.12.0

What's Changed
* Introduce VirtualEmptyArray and hide virtual arrays in the DAG visualization by tomwhite in https://github.com/tomwhite/cubed/pull/290
* Introduce `RetryingFunctionExecutor` by tomwhite in https://github.com/tomwhite/cubed/pull/292
* Add support for batching inputs to avoid overwhelming the backend service by tomwhite in https://github.com/tomwhite/cubed/pull/291
* Support version 2022.12 of the Array API by tomwhite in https://github.com/tomwhite/cubed/pull/293
* Fix array `__abs__` method for complex inputs by tomwhite in https://github.com/tomwhite/cubed/pull/294
* Be explicit when opening Zarr arrays by tomwhite in https://github.com/tomwhite/cubed/pull/296
* Fix bug when running with Kerchunk by tomwhite in https://github.com/tomwhite/cubed/pull/297
* Remove limited implementation of `__setitem__` from `Array` by tomwhite in https://github.com/tomwhite/cubed/pull/302
* Introduce `VirtualFullArray` by tomwhite in https://github.com/tomwhite/cubed/pull/303
* [pre-commit.ci] pre-commit autoupdate by pre-commit-ci in https://github.com/tomwhite/cubed/pull/305
* Fix case where `shape` is an int for `empty` by tomwhite in https://github.com/tomwhite/cubed/pull/306
* Update Coiled example by tomwhite in https://github.com/tomwhite/cubed/pull/309
* Add `nansum` and `nanmean` functions by tomwhite in https://github.com/tomwhite/cubed/pull/308
* Add `moveaxis` by tomwhite in https://github.com/tomwhite/cubed/pull/313
* Update minimum Python version to 3.9 by tomwhite in https://github.com/tomwhite/cubed/pull/298


**Full Changelog**: https://github.com/tomwhite/cubed/compare/0.10.0...0.11.0

What's Changed
* Scaling docs by TomNicholas in https://github.com/tomwhite/cubed/pull/252
* Remove PipelineExecutor classes by tomwhite in https://github.com/tomwhite/cubed/pull/253
* Surface rechunk intermediates in DAG by tomwhite in https://github.com/tomwhite/cubed/pull/257
* Add type hints for runtime by tomwhite in https://github.com/tomwhite/cubed/pull/258
* Fix modal import by tomwhite in https://github.com/tomwhite/cubed/pull/261
* Add `compute_arrays_in_parallel` option to Modal and Python async executors by tomwhite in https://github.com/tomwhite/cubed/pull/263
* Use new `merge_chunks` function in `reduction` by tomwhite in https://github.com/tomwhite/cubed/pull/256
* Remove vendored rechunker API, since we only use the algorithm by tomwhite in https://github.com/tomwhite/cubed/pull/264
* Add unit tests for `virtual_offsets` function by tomwhite in https://github.com/tomwhite/cubed/pull/265
* [pre-commit.ci] pre-commit autoupdate by pre-commit-ci in https://github.com/tomwhite/cubed/pull/266
* Lithops compute arrays in parallel by tomwhite in https://github.com/tomwhite/cubed/pull/259
* Remove duplicate copy of `projected_mem` and `num_tasks` from DAG by tomwhite in https://github.com/tomwhite/cubed/pull/271
* Allow step size > 1 by tomwhite in https://github.com/tomwhite/cubed/pull/270
* Fix `reserved_mem` in `create_lazy_zarr_arrays` by tomwhite in https://github.com/tomwhite/cubed/pull/273
* Dask tests by tomwhite in https://github.com/tomwhite/cubed/pull/269
* Coiled Functions executor by TomNicholas in https://github.com/tomwhite/cubed/pull/260
* Ensure runtime memory is not less than allowed by tomwhite in https://github.com/tomwhite/cubed/pull/274
* In-process Python executor should not retry by tomwhite in https://github.com/tomwhite/cubed/pull/275
* Refactor pipeline mappable by tomwhite in https://github.com/tomwhite/cubed/pull/272
* Generalise async executor code by tomwhite in https://github.com/tomwhite/cubed/pull/278
* Dask distributed async executor by tomwhite in https://github.com/tomwhite/cubed/pull/279
* [pre-commit.ci] pre-commit autoupdate by pre-commit-ci in https://github.com/tomwhite/cubed/pull/283
* Dask check_runtime_memory by tomwhite in https://github.com/tomwhite/cubed/pull/281
* Fixes for converting strings to bytes by TomNicholas in https://github.com/tomwhite/cubed/pull/289


**Full Changelog**: https://github.com/tomwhite/cubed/compare/0.9.0...0.10.0

What's Changed
* Fix Modal deprecated API by tomwhite in https://github.com/tomwhite/cubed/pull/214
* Modal GCP by tomwhite in https://github.com/tomwhite/cubed/pull/215
* Add html array repr by TomNicholas in https://github.com/tomwhite/cubed/pull/216
* Add to pre-commit.ci by tomwhite in https://github.com/tomwhite/cubed/pull/218
* Skip brew update when setting up GraphViz on mac CI to improve run time by tomwhite in https://github.com/tomwhite/cubed/pull/227
* Set 'reraise' to True for retries by tomwhite in https://github.com/tomwhite/cubed/pull/225
* Run mypy in CI by tomwhite in https://github.com/tomwhite/cubed/pull/229
* Print irregular chunking pattern when error raised by TomNicholas in https://github.com/tomwhite/cubed/pull/231
* [pre-commit.ci] pre-commit autoupdate by pre-commit-ci in https://github.com/tomwhite/cubed/pull/232
* Allow rechunk to take a dict for chunks by tomwhite in https://github.com/tomwhite/cubed/pull/226
* Improve Modal backups implementation by tomwhite in https://github.com/tomwhite/cubed/pull/238
* Test that plan generation scales to very large workloads by tomwhite in https://github.com/tomwhite/cubed/pull/240
* Don't materialize input iterables to `map_unordered` by tomwhite in https://github.com/tomwhite/cubed/pull/242
* Lithops retries and backups by tomwhite in https://github.com/tomwhite/cubed/pull/217
* Add type hints for primitive functions by tomwhite in https://github.com/tomwhite/cubed/pull/236
* Add links to Xarray integration and blog post to docs by tomwhite in https://github.com/tomwhite/cubed/pull/243
* Account for reserved_mem when creating arrays by tomwhite in https://github.com/tomwhite/cubed/pull/245
* Use np.unravel_index rather than iterating over np.ndindex by tomwhite in https://github.com/tomwhite/cubed/pull/246
* Add matmul example that validates result by tomwhite in https://github.com/tomwhite/cubed/pull/244
* Document strong consistency by tomwhite in https://github.com/tomwhite/cubed/pull/250
* Re-enable rechunker's `consolidate_reads` by tomwhite in https://github.com/tomwhite/cubed/pull/249
* Introduce VirtualOffsetsArray that is never materialized by tomwhite in https://github.com/tomwhite/cubed/pull/248
* Avoid divide by zero from projected mem by tomwhite in https://github.com/tomwhite/cubed/pull/251

New Contributors
* pre-commit-ci made their first contribution in https://github.com/tomwhite/cubed/pull/232

**Full Changelog**: https://github.com/tomwhite/cubed/compare/0.8.0...0.9.0

What's Changed
* Lithops gcf example by TomNicholas in https://github.com/tomwhite/cubed/pull/183
* Name Modal Stub objects to avoid warning by tomwhite in https://github.com/tomwhite/cubed/pull/189
* Convert str to bytes by TomNicholas in https://github.com/tomwhite/cubed/pull/188
* Reinstate Lithops AWS examples that were accidently removed in 174 by tomwhite in https://github.com/tomwhite/cubed/pull/185
* Improvements and fixes to Modal examples by tomwhite in https://github.com/tomwhite/cubed/pull/190
* More examples on GCF by tomwhite in https://github.com/tomwhite/cubed/pull/198
* Fix measure reserved by tomwhite in https://github.com/tomwhite/cubed/pull/200
* Include 'cubed' in the directories created in the working directory by tomwhite in https://github.com/tomwhite/cubed/pull/202
* Add user guide and break out getting started into separate pages by tomwhite in https://github.com/tomwhite/cubed/pull/203
* Show array variable name in visualization by tomwhite in https://github.com/tomwhite/cubed/pull/192
* Runtime refactoring by tomwhite in https://github.com/tomwhite/cubed/pull/206
* Make Zarr array creation lazy by tomwhite in https://github.com/tomwhite/cubed/pull/207
* Make executors create Zarr arrays by tomwhite in https://github.com/tomwhite/cubed/pull/209
* Make Lithops retries apply per-task by tomwhite in https://github.com/tomwhite/cubed/pull/211
* Move Modal examples to 'aws' directory by tomwhite in https://github.com/tomwhite/cubed/pull/212
* Publish packages to PyPI using GitHub Actions by tomwhite in https://github.com/tomwhite/cubed/pull/213


**Full Changelog**: https://github.com/tomwhite/cubed/compare/0.7.0...0.8.0

What's Changed
* History analysis by tomwhite in https://github.com/tomwhite/cubed/pull/170
* Rename memory settings by tomwhite in https://github.com/tomwhite/cubed/pull/180
* Docs on installing from conda + optional dependencies by TomNicholas in https://github.com/tomwhite/cubed/pull/178
* Rename `peak_memory` to `peak_measured_mem` in line with other memory settings by tomwhite in https://github.com/tomwhite/cubed/pull/182


**Full Changelog**: https://github.com/tomwhite/cubed/compare/0.6.3...0.7.0