Images, points, shapes etc coordinates are no synced in a 1-to-1. E.g. a transcripts with coordinate (x, y) may not fall onto the pixel y, x of an image. We could make it easier to have this (we already can do this with transform_to_data_extent() , but this function is not “advertised”.
- (Luca’s and Marvin’s favorite): push forward with ngff-transformations. Ergonomic library for dealing with transformation
- (Wouter): we can introduce a “prealigned=True”, is_prealigned: bool
Images, points, shapes etc coordinates are no synced in a 1-to-1. E.g. a transcripts with coordinate (x, y) may not fall onto the pixel y, x of an image. We could make it easier to have this (we already can do this with
transform_to_data_extent(), but this function is not “advertised”.