Joris Van den Bossche

Note that the "accepted" answer (not the first one) on that SO question is outdated (it has an updated note, but the starting paragraph about saying there is no standard...

`bounds` and `unary_union` both correctly read only the geometry column (if doing `gdf.geometry.bounds/unary_union` of course). So not really sure what's special about `total_bounds`. One difference is that it returns an...

See https://github.com/dask/dask/issues/7885 for an analysis of the issue. The way to solve it (with current dask at least) is to return a Series/DataFrame object instead of an array. I was...

> A Series would be okay but to have it compatible in the way you mention `gdf.total_bounds[0]` it needs to have a range index while it would make more sense...

Also on the main development branch, this is still returning "geometry" (but it _does_ give a clear error message when accesing `.geometry`now). I don't think this would be an easy...

There was a basic `sjoin` implementation in the old repo: https://github.com/mrocklin/dask-geopandas/blob/8133969bf03d158f51faf85d020641e86c9a7e28/dask_geopandas/core.py#L409-L430. I didn't yet port that over, although it should be possible now basic spatial partition information is available (https://github.com/geopandas/dask-geopandas/pull/14)....

If I remove the line `ddf = ddf.set_crs(pyproj.CRS(4326))` from your example, it still runs fine for me (except that `sjoin` raises a warning about mismatching CRS). Do you know what...

Although this is indeed the geojson spec, IMO we at least should give the option to not do this conversion. For example, I myself have used geojson in my own...

Another aspect: you can also write GeoJSON files with `to_file(.., driver="GeoJSON")`, and there, such a reprojection does not happen. So that also introduces an inconsistency ..

I am not aware of any way to do that currently. But it would certainly be great to have something like this in the future!