Kepler.gl: Time-series choropleth by combining GeoJSON with secondary data

[brylie]

We have timeseries data containing a column for country name. We also have a GeoJSON of world countries. However, when creating a polygon layer for the GeoJSON file, it does not allow us to color the countries based on data in a second data set.

I would like to create a choropleth map by providing a GeoJSON file for geometry and then styling the GeoJSON polygons based on fields in a second data 'layer'. This could additionally allow the secondary data set to be time filtered, so we could see change over time:

Describe alternatives you've considered
I have considered creating the interactive, time-series choropleth with Bokeh.

[macrigiuseppe]

Hello @brylie,
in order to perform reach the result you are looking for Kepler.gl should be able to join datasets and perform operations on the join result. Said feature is part of the out roadmap and we are going to work on it next year.

As workaround, I suggest to merge the two datasets into in order achieve the result you want.

[brylie]

Alright, I can give it a try. As the data are time-series measurements binned by country and date, do you recommend any resources or tutorials for joining such data? I will search around and ask on Stack Overflow.

[brylie]

Also, wouldn't a join creat a lot of duplicated geometry data? Geopolitics aside, the geometries do not change as often as the measurements, so might be efficient to reference only once. Relatedly, in our data some countries have no measurements in a given period or at all.

[heshan0131]

@brylie you are right, you will have to create duplicate geometry while joining. The duplicated geometry will be filtered out once the time filter is enabled.

[heshan0131]

Closing it for now, please re-open with more questions

[damnko]

Hi, Is there any update on this feature? Thanks