Favelas in Rio de Janeiro

This is another example in which we use data from the ArcGIS server of Rio de Janeiro city.

This time, we’ll try to visualize the distribution of slums or favelas in the city.

Let’s first try to get the map of the city and afterwards plot the favelas geometries.

import pandas as pd

pd.set_option("max_rows", 10)

from mapsbr import arcgis

districts = arcgis.get_map(
    service="Basicos/mapa_basico_UTM",
    layer=15,
    baseurl="https://pgeo3.rio.rj.gov.br/arcgis/rest/services/",
)

favelas = arcgis.get_map(
    service="SMH/Limites_de_Favelas",
    layer=0,
    baseurl="https://pgeo3.rio.rj.gov.br/arcgis/rest/services/",
)

Now, just to see what other interesting things we can do with GeoPandas and the Shapely library in general: let’s calculate the geometries’ areas.

This will give us a rough idea about which favela is the biggest:

favelas["area"] = favelas.geometry.area
favelas = favelas.sort_values("area", ascending=False)

Now, here is what I wanna do: highlight the five biggest ones.

This will be straight-forward to do with annotate method of matplotlib.

favelas["point"] = favelas.geometry.centroid  # calculate the geometry centroid

import matplotlib.pyplot as plt

fig, ax = plt.subplots(figsize=(10, 7), constrained_layout=True)
districts.plot(color="white", edgecolor="slategray", ax=ax)
favelas.plot(color="orange", ax=ax)
favelas.query("Nome in @favelas.Nome.head()").plot(color="gray", ax=ax)

ax.axis("off")
ax.set_title("Favelas in Rio\nBiggest 5 highlighted")

for _, row in favelas.head().iterrows():
    ax.annotate(
        row.Nome,
        xy=(row.point.x, row.point.y),
        xytext=(5, 10),
        textcoords="offset points",
        arrowprops=dict(arrowstyle="->", color="black", connectionstyle="angle3"),
        bbox=dict(boxstyle="round", alpha=0.7, facecolor="silver"),
        color="k",
    )

Estimated memory usage: 21 MB