Reason or Problem
docs/source/user_guide/multispectral.ipynb imports datashader, uses Canvas().raster() to resample bands, and renders with the transfer functions and the Images helper.
Canvas().raster() is raster resampling, not vector rasterization, so arr.xrs.rasterize does not replace it directly.
Proposal
- Remove the datashader imports.
- Replace the
Canvas().raster() resampling step with a plain xarray/numpy resample, or drop it if it was only cosmetic.
- Replace the transfer functions,
Images, and datashader colormaps with matplotlib.
- Update the prose cells that describe the datashader workflow.
- Re-execute the notebook so outputs and the preview image regenerate cleanly.
Done when
- The notebook runs top to bottom with no datashader import and produces the expected plots.
Part of removing datashader as a dependency; final packaging removal is tracked in #3450.
Reason or Problem
docs/source/user_guide/multispectral.ipynbimports datashader, usesCanvas().raster()to resample bands, and renders with the transfer functions and theImageshelper.Canvas().raster()is raster resampling, not vector rasterization, soarr.xrs.rasterizedoes not replace it directly.Proposal
Canvas().raster()resampling step with a plain xarray/numpy resample, or drop it if it was only cosmetic.Images, and datashader colormaps with matplotlib.Done when
Part of removing datashader as a dependency; final packaging removal is tracked in #3450.