From what I’ve read in the docs and this site, it seems that there’s built-in support for global maps by country name or ISO code, and for US states. But, as far as I can tell, if you want a map with custom polygons, you can’t use any of the existing choropleth support. Rather, you have to roll your own from more primitive features.
It would seem feasible to extend the current implementation to add scope=“custom”, with a reference to geoJSON data or a geopandas object and an attribute to plot. Is any such enhancement in the works?
That’s correct. The plotly.js team is currently adding US counties to choropleths and, depending on the difficulty, may add features for custom polygons. There isn’t a PR yet but there will be one within a few weeks at https://github.com/plotly/plotly.js/pulls.
You can also currently add GeoJSON to scattermapbox (see e.g. https://plot.ly/python/county-level-choropleth/) however you have to map the data to colors yourself and there is no built-in hover interactions.
We would really like to expand our feature-set for custom shapefiles and choropleths but it’s a really big job. We are looking for a corporate sponsor to help make this possible.
There aren’t any updates on this yet. A few other company-sponsored projects have taken priority this Fall (if your company could help prioritize this work through a sponsorship, please reach out)