Discrete Colors for Choroplethmapbox with plotly.graph_objects

Hi everyone !

I was wondering if it is possible to have discrete colors for go.Choroplethmapbox ?

So exactly like this example: https://plotly.com/python/mapbox-county-choropleth/#discrete-colors,
but with plotly.graph_objects instead of plotly.express ?

go.Choroplethmapbox doesn’t have a ‘color’ attribute like px.choropleth_mapbox.

When I use the ‘z’ attribute I get this item in the legend and since all my “polygons” have the same color, it makes no sense.


Thanks !

Hi @Ahmed3 the plotly express figure is built internally by creating a go.Choroplethmapbox trace of constant z, and a dedicated colorscale which has a constant color over the [0, 1] value range. By the way, you can learn this by doing print(fig) on the figure created by plotly express. Below is a code snippet reproducing the example figure

import plotly.express as px

df = px.data.election()
geojson = px.data.election_geojson()

colorscales = [
    ((0.0, '#636efa'), (1.0, '#636efa')),
    ((0.0, '#EF553B'), (1.0, '#EF553B')),
    ((0.0, '#00cc96'), (1.0, '#00cc96'))

fig = go.Figure()
for i, winner in enumerate(df['winner'].unique()):
    dfp = df[df['winner'] == winner]
    fig.add_choroplethmapbox(geojson=geojson, locations=dfp['district'],
                             z=[i,] * len(dfp), featureidkey="properties.district",
                             showlegend=True, name=winner,
                             colorscale=colorscales[i], showscale=False)

                  mapbox=dict(style="carto-positron", zoom=9,
                              center={"lat": 45.5517, "lon": -73.7073},))


Fantastic !

Thanks a lot Emmanuelle, solved my problem right away !