Closed tibuch closed 6 months ago
Moving towards a dask array supported (#42) io implementation adds another layer of complexity. In light of this I would perform the histogram computation as an additional step once the ome-zarr is fully written. This would mean that the histogram computation becomes our first processing step working on an ome-zarr file.
We can use dask-arrays as input to the histogram class see: https://github.com/fmi-faim/faim-ipa/blob/6a1aa95def0dd513a24102c02abf1e607830c194/src/faim_ipa/UIntHistogram.py#L39.
The current code keeps the histogram computation separate from the ome-zarr creation, which seems to work fine so far.
We can re-open this issue if it becomes a problem in the future.
A remaining question is how to deal with the histograms (
UIntHistogram
): should we still read all pixel values to create them immediately when processing/stacking the image, or rather delay creation until we actually write the well/field into the zarr store?Originally posted by @imagejan in https://github.com/fmi-faim/faim-hcs/issues/42#issuecomment-1739095300