You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Currently, arrays for labels, images, etc. are required to be dask arrays. This makes sense when loading data from a store, but doesn't make sense to me when creating in memory (e.g. spatialdata.models.Labels2DModel.parse(np.array([[0, 0], [0, 0]]), dims=["y", "x"]))
Hi Jousha, thanks for the feeback! I agree with you. We are planning to work on the support both dask and non-dask objects in a new refactoring of the data model that we plan to do in a hackathon in 2 weeks. Now that this issue has been reported we'll try to prioritize it. CC @giovp
Activity
LucaMarconato commentedon May 9, 2023
Hi Jousha, thanks for the feeback! I agree with you. We are planning to work on the support both dask and non-dask objects in a new refactoring of the data model that we plan to do in a hackathon in 2 weeks. Now that this issue has been reported we'll try to prioritize it. CC @giovp
lopollar commentedon May 17, 2023
Hi both, I do have the same issues, both for my points layer as for my segmentation masks. It would be great if this can be solved.