I just went through the vectorization/rasterization tutorial.
What I found confusing is that it starts by claiming to convert between shapes and labels:
https://github.com/scverse/spatialdata-notebooks/blob/7ce65b2f8a291fa21fdb1542b91518378726b51d/notebooks/examples/labels_shapes_interchangeability.ipynb#L11
But then, later, shapes are not converted to labels, but to images:
https://github.com/scverse/spatialdata-notebooks/blob/7ce65b2f8a291fa21fdb1542b91518378726b51d/notebooks/examples/labels_shapes_interchangeability.ipynb#L265
I wanted to convert shapes to labels, but it took me a while to realize I needed to set return_regions_as_labels=True.
I imagine that converting from shape to label is needed more often, because then you can use the labels to aggregate. Would it make sense to update the example to conversion into labels instead of images?
I just went through the vectorization/rasterization tutorial.
What I found confusing is that it starts by claiming to convert between shapes and labels:
https://github.com/scverse/spatialdata-notebooks/blob/7ce65b2f8a291fa21fdb1542b91518378726b51d/notebooks/examples/labels_shapes_interchangeability.ipynb#L11
But then, later, shapes are not converted to labels, but to images:
https://github.com/scverse/spatialdata-notebooks/blob/7ce65b2f8a291fa21fdb1542b91518378726b51d/notebooks/examples/labels_shapes_interchangeability.ipynb#L265
I wanted to convert shapes to labels, but it took me a while to realize I needed to set
return_regions_as_labels=True.I imagine that converting from shape to label is needed more often, because then you can use the labels to aggregate. Would it make sense to update the example to conversion into labels instead of images?