Choropleth map using folium and pandas

Question

I am using folium to create a choropleth map of a set of countries. I am following the documentation. However. for some reason the map doesn't show any shades. I am using the world geojson from natural earth (see the gist).

My dataframe looks like:

>>> spatial_scores.head()

Out[1]:
id  Country Score
PER Peru    2.810300
HND Honduras    2.734521
GUF French Guiana   2.730886
SLV El Salvador 2.473134
CRI Costa Rica  2.454963

The world geojson looks like:

>>> world_json['features'][0]['id']

Out [2]:
u'AFG'

The relevant portions of the choropleth codes are as below:

map1 = folium.Map(location=[-15., -60], zoom_start=4)

map1.geo_json(geo_path=world_json_path,
              data_out='data.json',
              data=spatial_scores,
              columns=['id', 'Score'],
              threshold_scale=[0, 1, 2, 3, 4],
              key_on='features.id',
              fill_color='BuPu', fill_opacity=0.7, line_opacity=0.5,
              legend_name='Score')

map1.create_map('./Scores.html')

However, am not getting any choropleth result and left with just the base country map as below

Is there something I am doing wrong?

[Edit]

I figured out the problem. To plot the choropleth I needed to keep only those keys in the geojson which were also in my data frame.

merged = gdf.merge(spatial_scores, left_on='name', right_on='Country')
spatial_gdf = gpd.GeoDataFrame(merged.iloc[:, [0, 1]])
data_df = merged.iloc[:, [2, 3, 4]]

hi I actually solved it myself.. Will update the gist tonight. The dead simple answer is that in the geojson you need to put only those keys for which the data is available in the corresponding table — goofd, Feb 09 '16 at 21:23
See the updated question and also if you click the linked gist you will get your answer — goofd, Feb 11 '16 at 01:07
If this library is giving you trouble, you could try Plotly instead. It also supports choropleth plots and hovering for exact data. — Jacob K, Aug 18 '20 at 02:59
@goofd, go ahead and write an answer to your own question -- that way it can be closed out — Schalton, Oct 06 '21 at 18:27

score 1 · Answer 1 · answered Dec 15 '21 at 20:00

1

To plot the choropleth OP needed to keep only those keys in the geojson which were also in the data frame.

merged = gdf.merge(spatial_scores, left_on='name', right_on='Country')
spatial_gdf = gpd.GeoDataFrame(merged.iloc[:, [0, 1]])
data_df = merged.iloc[:, [2, 3, 4]]

answered Dec 15 '21 at 20:00

0x263A

1,807
9
22

1

@goofd I was looking at the unaswered questions in the Python tag and this was the top question. I found it a bit silly for the top unanswered question to have an answer that was added in 2016 so I looked [at the meta](https://meta.stackoverflow.com/a/267439/16450169) and followed their advice by simply adding your answer as a community answer. – 0x263A Dec 15 '21 at 20:06

Choropleth map using folium and pandas

1 Answers1

Linked