Closed constantinpape closed 2 months ago
If I run:
ome_zarr view Platynereis-H2B-TL.ome.zarr/c0-t0/
I see the following:
i.e. two separate issues:
The dimension separator is not a problem. However, it did complain if I didn't specify a shard size. This works...
$ ome2024-ngff-challenge resave --cc-by Platynereis-H2B-TL.ome.zarr/c0-t0 zarr_v3/Platynereis-H2B-TL.ome.zarr --output-overwrite --output-shards=512,512,512
Then view it with:
$ ome_zarr view zarr_v3/Platynereis-H2B-TL.ome.zarr
Chunks are loading OK etc:
Thanks for the replies @joshmoore and @will-moore
Regarding the points you raised:
I'm using the released version (or a git branch that is effectively the same).
Did you change your command to point at the lower directory /c0-t0
as input? ie. Platynereis-H2B-TL.ome.zarr/c0-t0
Thanks @will-moore . That did the trick! And I understand the issue now; there was still a conceptual misunderstanding from my side.
I want to convert data from ngff v0.4 to the challenge submission format, but the resave command only creates the output root directory without converting any data. It also throws a warning:
throws
I assume that is because the file is not fully spec compliant with the v0.4 spec (it comes from a custom pipeline), or that the converter makes some additional assumptions not part of the spec.
However, I can't determine what the issue is since the warning thrown here is quite unspecific and since I did not find another way to validate the file yet, see also https://github.com/ome/ome-zarr-py/issues/400.
The file in question is available here: https://drive.google.com/file/d/1WSHQWkOXUfSBahOJdVzrKbj91mczKhZP/view?usp=sharing