Selman Özleyen

@ivirshup @rahulbshrestha, I did some changes that enables the Zarr support. I still don't know what encoding-type and encoding-version refer to, so there is probably a cleaner way to do...

Added Hdf5 and Zarr support for read and write operations as array. Optional dependency issue is also solved. I will now be looking to the index operations. I also added...

@ivirshup @rahulbshrestha, I added DaskArray to the view dispatch and added the `matrix_type`. But the thing is, this doesn't cover all the tests. Should I create more tests or just...

> I would probably make objects returned by `gen_adata` not contain dask arrays by default. So tests would have to opt in to getting a dask array. I was planning...

Added new cases, having errors like these: ``` FAILED anndata/tests/test_views.py::test_view_delattr[array_subset-X] - NotImplementedError: Don't yet support nd fancy indexing ... FAILED anndata/tests/test_views.py::test_view_delattr[array_subset-varp] - NotImplementedError: Don't yet support nd fancy indexing ```...

While looking for the concatenation support, I ran into some errors. My guess is that it has something to do with the indexing in dask. I don't know if it...

Hi @ivirshup, I will focus on these two, while @rahulbshrestha will do the other two. > * [ ] `adata.to_memory` should load dask arrays into memory > > * [...

Hi @ivirshup, While, waiting for clarifications on my questions above, I added dask array to `test_concatenate.py` functions. There are errors coming from this function in `anndata/__core/merge.py` https://github.com/scverse/anndata/blob/919d34cfe08faf54daa9f9caa76090739fd66262/anndata/_core/merge.py#L367-L381 This is probably...

Apart from this, I fixed another compatibility issue on concatenation. Traced the merge functions and found out there was a bug because we didn't have equal implemented for **DaskArray**. So...

> I think we can leave np.nan to be not equal to np.nan as it is part of the definition. But in anndata we don't want this behavior. Please see...