Draw a risk map

To draw a choropleth map, fetch the outlines once, fetch the scores for the month you want, and join the two on the area code.

What you need

A key with the Risk scores and LGA boundaries scopes.

Join scores to shapes

Map	Shapes	Scores	Join
National, by state	State boundaries	State risk	properties.code = state_code
One state, by LGA	LGA boundaries	LGA risk	properties.lgaCode = lga_code

Example with GeoPandas

import os
import geopandas as gpd
import pandas as pd
from ai4m_sdk import AI4MClient

client = AI4MClient(api_key=os.environ["AI4M_API_KEY"])

shapes = gpd.GeoDataFrame.from_features(client.get_lga_boundaries("KN"), crs="EPSG:4326")
report = client.get_lga_risk("KN")
scores = pd.DataFrame([vars(lga) for lga in report.lgas])

kano = shapes.merge(scores, left_on="lgaCode", right_on="lga_code")
kano.plot(column="score", cmap="YlOrRd", vmin=0, vmax=1, legend=True, edgecolor="white")

Colours

The AI4M dashboard uses these colours for the five levels. Using the same ones keeps your map consistent with it.

Level	Colour
very_low	#FFF3B0
low	#FDD35C
moderate	#F9A13B
high	#E0412B
very_high	#9E0B2B

Good practice

• Fetch boundaries once and keep them. They do not change from month to month.
• The outlines are simplified for drawing, so do not use them to measure area or to decide which LGA a point falls in.
• Add a legend with the score bands, and say whether the month is an estimate or a forecast.
• Credit the boundaries: GADM, version 4.1.