For now the downsample_zarr notebook downgrades hvplot to work around the bug. And in #139 a different workaround is introduced of just passing a cartopy crs object directly.
But once the bug is resolved and a new version of hvplot is released and available in the pangeo image, we should be able to remove the workaround.
Based off the report in https://github.com/NASA-IMPACT/veda-docs/pull/119#issue-2136845617 there was a bug uncovered in hvplot > 0.8.4: https://github.com/holoviz/hvplot/issues/1329.
For now the
downsample_zarrnotebook downgrades hvplot to work around the bug. And in #139 a different workaround is introduced of just passing a cartopy crs object directly.But once the bug is resolved and a new version of hvplot is released and available in the pangeo image, we should be able to remove the workaround.