Closed ianthomas23 closed 11 months ago
ianthomas23
commented
11 months ago Looks good! Can you include a sentence or two in the new notebook about the tradeoffs between GeoPandas and SpatialPandas? Presumably something about one being faster, but the other being more convenient if you are already have your data in GeoPandas?
Yes, I'll do that.
ianthomas23
commented
11 months ago The new page is at https://holoviz-dev.github.io/datashader/user_guide/Geopandas.html in the dev docs.
jbednar
commented
11 months ago Looks good, thanks! I forget -- do we have antialiasing supported for polygons? If so it would be good to add that to the example.
ianthomas23
commented
11 months ago Looks good, thanks! I forget -- do we have antialiasing supported for polygons?
No, there is only antialiasing support for lines.
jbednar
commented
11 months ago Oops; I wasn't clear. I meant, antialiasing for polygon outlines. Those are just polylines, so I'd think they should be supported?
ianthomas23
commented
11 months ago Oops; I wasn't clear. I meant, antialiasing for polygon outlines. Those are just polylines, so I'd think they should be supported?
Yes, any use of Canvas.line() supports antialiasing regardless of the type of the source. Here it is with line_width=2:
This adds a new page to the User Guide in the docs to show use of GeoPandas. It uses one of the
geodatasetsand shows three images, one of which is (population density per US county in 1990):It is minimalistic at this stage as it uses pure Datashader. When we have support in HoloViews for this, we can add an interactive section to the end and hence show interactive zooming and panning, and show colorbars as well, to make it more interesting.