v1.2.1

• [#96] Add support for numpy 1.25+ (as opposed to just 2.0+) and correct minimum supported scipy version (1.8→1.11) (@flying-sheep)

Dev changes:

• [#98] Derive test configs from hatch (@flying-sheep)

v1.2

• [#94] Add order parameter to to_dense (@flying-sheep)

v1.1.2

• [#91] Make array_type a session-scoped fixture (@flying-sheep)

v1.1.1

• [#86] Support older anndata versions (@flying-sheep)
• [#87] Fix Dask handling for chunks with 1-wide dimension (@flying-sheep)

v1.1

• [#84] Allow use on 1d arrays (@flying-sheep)

v1.0

The initial release has

• fast_array_utils.conv with to_dense.
• fast_array_utils.stats with is_constant, sum, mean, and mean_var.
• testing.fast_array_utils with SUPPORTED_TYPES, ArrayType, and Flags.
• testing.fast_array_utils.pytest with array_type and pytest.mark.array_type(select, skip).

Notable PRs:

• [#10] Zarr support (@flying-sheep)
• [#13] stats.sum (@flying-sheep)
• [#16] Sparse in dask (@flying-sheep)
• [#21] stats.is_constant (@flying-sheep)
• [#30] Refactor test utils to use a class (@flying-sheep)
• [#44] stats.mean (@flying-sheep)
• [#46] stats.mean_var (@flying-sheep)
• [#50] anndata.abc.CS{RC}Dataset support (@flying-sheep)
• [#51] CuPy support (@flying-sheep)
• [#73] Add a scipy.sparse numba extension (@flying-sheep)