MIKRIPOLI

Three small cities of Rio Grande do Sul, Brazil, in the 2010 census. A reproducible look at the population, household size, income and neighborhoods of Sobradinho, Rio Pardo and Encantado, the case-study towns of the comparative morphology study of the Mikripoli research network on small cities.

It rebuilds, in Python, the census maps of Morfologia urbana, pequenas cidades e desenvolvimento regional na Região Intermediária de Santa Cruz do Sul e Lajeado-RS-Brasil (Silveira, Faccin, Detoni & Silva, 2024) from the IBGE census tracts of 2010.

Key results

• The center is where income is. In all three towns the household heads of the central neighborhood earn the most: on average 3.7 minimum wages in Sobradinho, 4.8 in Rio Pardo and 3.8 in Encantado (2010), against 0.9 to 1.5 in the poorest neighborhoods.
• Larger households at the edges. Tracts with 3.5 or more residents per dwelling sit on the periphery and in the rural surroundings; the centers have the smallest households.
• Different degrees of urbanization. 87% of Encantado's residents lived in urban tracts in 2010, 79% in Sobradinho and 68% in Rio Pardo.
• Small, but not simple. The three seats have 11 to 15 neighborhoods each, with the center, the riverside and the highway accesses as their structuring elements.

Figures

How it works

flowchart LR
    A["IBGE census tracts 2010<br/>Rio Grande do Sul"] --> C
    B["Census 2010 universe results<br/>Basico file, by tract"] --> C
    D["OpenStreetMap roads<br/>IBGE water bodies"] --> C
    C["pipeline.py"] --> E["tables (CSV)"]
    C --> F["figures (PNG)"]

Loading

1. Tracts. The 154 census tracts of the three municipalities, joined to the universe results: residents in permanent private dwellings (V002), mean residents per dwelling (V003) and mean monthly income of the household heads (V005), in minimum wages of 2010 (R$ 510).
2. Neighborhoods. The urban tracts of each municipal seat dissolved by their IBGE neighborhood code.
3. Context. Roads from OpenStreetMap (downloaded once and cached) and water bodies from IBGE.

Run it

python -m venv venv && source venv/bin/activate
pip install -r requirements.txt
cp config/config.local.json.example config/config.local.json   # set sources_dir and outputs_dir
python pipeline.py                               # tables + figures + README images
pytest                                           # unit tests

config/config.local.json (gitignored) sets two folders:

Key	Purpose
sources_dir	Shared raw-data catalog. Reads the 2010 tracts from ibge/censo/2010/setores_censitarios_rs/t0/, the universe results from ibge/censo/2010/universo_rs/t0/CSV/Basico_RS.csv and the water bodies from _projetos/rio_pardo_enchentes_2024/shp/
outputs_dir	This project's outputs: tables/, figures/, cache/

Outputs (outputs_dir)

File	Content
tables/tracts_2010.csv	The 154 tracts with residents, residents per dwelling and income
tables/towns_2010.csv	Residents, urban share and mean income by town
tables/neighborhoods_2010.csv	Residents, residents per dwelling and income by neighborhood (means weighted by residents)
tables/comparison_with_paper.csv	Value ranges vs. the class limits of the published map
figures/*.png	All figures (copied to docs/img/ for this README)

Check against the paper

The class limits printed on the published map match the data: residents per tract up to 1,229 and residents per dwelling from 2.0 to 4.5. The smallest tract has 2 residents; the map's first class starts at 3.

Notes on the data

• Census 2010. The paper used the 2010 census, the latest with income by tract when it was written.
• Residents count people in permanent private dwellings, so they add up to slightly less than the total population.
• Neighborhoods are those coded by IBGE in 2010; municipal laws may delimit them differently today.

Repository layout

pipeline.py            orchestrator (tables, figures, docs)
src/mikripoli/
  config.py            paths from config/config.local.json
  data.py              tracts, census results, neighborhoods, roads, water
  metrics.py           tables and the check against the paper
  osm.py               OpenStreetMap roads (Overpass API, cached)
  figures.py           figures
  style.py             figure style: source line and repository name
  brand.py             visual identity (colors, palettes, Source Code Pro, layout); copied
                       from the author's brand repository, do not edit here
tests/                 synthetic unit tests
assets/fonts/          Source Code Pro (SIL OFL)

Credits

Paper: Silveira, R. L. L.; Faccin, C. R.; Detoni, L. P.; Silva, P. J. R. (2024). Morfologia urbana, pequenas cidades e desenvolvimento regional na Região Intermediária de Santa Cruz do Sul e Lajeado-RS-Brasil. Redes, 29. Part of the work of the Mikripoli research network on small cities.

Data: IBGE Census 2010 and OpenStreetMap contributors (ODbL). Figures use the Source Code Pro typeface (SIL Open Font License).

License

GNU General Public License v3.0, see LICENSE.