Add Zarr to benchmarks#446
Conversation
…dataR into keller-mark/zarr
…milar to HDF5AnnData
Simplify how obs and var names handled in ZarrAnnData
More zarr-related changes
Update comments
…ome documentation
|
Bencher Report| Branch | add-zarr-benchmarks |
| Testbed | ubuntu-latest |
🐰 View full continuous benchmarking report in Bencher
⚠️ WARNING: Truncated view!The full continuous benchmarking report exceeds the maximum length allowed on this platform.
|
@lazappi Aside from the linting issue, is this PR ready for review? |
|
I think so, now that the main PR is merged |
|
Out of curiosity, would I be able to observe the impact of some changes in Rarr here? I have my own continuous benchmarking suite but anndata(R) uses zarr in a different way which made me discover more performance issues last week. So I'd be interested in keeping an eye on your benchmarks. |
Hmm.. This might not be straight forward since you'll only know the impact of the changes you made in Rarr a new release was created. I see Rarr has touchtone set -- does this not do something similar? |
It does. But when I worked with some anndata datasets last month, I discovered some inefficiencies I didn't suspect. In particular, because my current set of benchmarks only tests single arrays, instead of a collection of arrays as in anndata. But based on your answer, the simplest solution is maybe to add an anndata dataset to my benchmarks then. |
5 participants
Description
Add the Zarr backend to the benchmark tests
Needs #190
Checklist
Before review
Before merge
NEWS