Feature request: reader plugin for zarr files output by empanada-napari #21
Description
It would be helpful if the empanada-napari plugin provided an inbuilt way to read back in prediction data generated by itself. The standard zarr reader plugins work in some, but not all cases for data generated by empanada-napari.
If there is just one dataset group in the zarr file, the regular zarr reader plugins can typically handle that. However, in cases where there are multiple datasets being saved by empanada-napari into a single zarr file (for example, when each orthoslice prediction output is saved individually, alongside the final combined predictions), that doesn't work.
Typically, those plugins find multiple datasets and attempt to parse it as a multi-resolution dataset, and fail because the data shape is the same for all of them (instead of steadily decreasing in size, like a multiscale zarr array).
I often want to re-load previous results from empanada-napari, in order to use the split and merge tools at a later date. Not being able to always read the data back in easily makes this workflow clunkier than is ideal.