All issues

write_zarr with dask array should use dask chunks by default

#2627

· joshua-gould opened

on Sep 1, 2026

Side effect when writing string columns to zarr

#2610

· ajkswamy opened

on Aug 19, 2026

Disable zarr auto sharding with dask as it's much slower

#2598

· joshua-gould opened

on Aug 10, 2026

No partial read for nullable-string-array

#2595

· Koncopd opened

on Aug 7, 2026

read_zarr fails with obs columns that differ only by case

#2576

· cvduffy opened

on Jul 28, 2026

Migrate to `cs{r,c}_array` internally/externally

type: sparse 🫥

#2477

· ilan-gold opened

on Jun 3, 2026

`pandas>=3` breaks with `xarray.Dataset` in `obsm`/`varm` when using `anndata.concat` and `load_annotation_index`

#2475

· ilan-gold opened

on Jun 2, 2026

make ref path available in `write_dispatched` and `read_dispatched` callback

#2457

· flying-sheep opened

on May 19, 2026

Allow any unicode incl. `/` and `\` in `obsm` keys, `obs` columns, `uns` keys, …

#2450

· flying-sheep opened

on May 18, 2026

·

write_zarr fails with LocalCUDACluster (dask-cuda) when adata.X has been persisted

#2444

· Intron7 opened

on May 13, 2026

`layers` can be 3D on-disk

#2430

· ilan-gold opened

on May 8, 2026

·

make write_zarr respect the public chunks parameter when the input is sparse

#2415

· mkarikom opened

on May 2, 2026