Datasets:

cmg777
/

project2026e

This directory contains the administrative boundary polygons for Bolivia's 339 municipalities, used by the spatial/ESDA notebooks (Moran's I, LISA, choropleths). Every feature is keyed on asdf_id, the universal join key used across all datasets in this project, so any tabular file in data/ can be merged onto these geometries for mapping.

Files

File	Features	Properties	Description
`bolivia339geoqueryOpt.geojson`	339	`asdf_id`, `shapeName`, `COORD_X`, `COORD_Y`	Web-optimized boundaries (simplified geometry + centroid coordinates). Preferred for interactive maps and most analysis.
`bolivia339geoqueryFull.geojson`	339	`asdf_id`	Full-resolution boundaries (detailed geometry, larger file). Use when maximum geometric precision is required.
`archive/`	—	—	Superseded boundary versions, kept for provenance. Local-only — not mirrored to Hugging Face.

Variable Dictionary

Property	Description
asdf_id	Unique spatial identifier (`0`–`338`) for joining to every other dataset in `data/`.
shapeName	Municipality name (present in the optimized file).
COORD_X	Municipality centroid longitude (present in the optimized file).
COORD_Y	Municipality centroid latitude (present in the optimized file).

Example Code

import geopandas as gpd
import pandas as pd

# Load the (web-optimized) municipality boundaries
gdf = gpd.read_file("bolivia339geoqueryOpt.geojson")

# Attach any indicator and make a choropleth
sdg = pd.read_csv("../sdg/sdg.csv")
gdf = gdf.merge(sdg[["asdf_id", "index_sdg1"]], on="asdf_id")
gdf.plot(column="index_sdg1", legend=True)

Access via Hugging Face

The two GeoJSON boundary files are mirrored to the public Hugging Face dataset cmg777/project2026e. (The archive/ folder is local-only and is not on the Hub.)

Browse / download in a browser: https://huggingface.co/datasets/cmg777/project2026e/tree/main/maps

Load in Python (pick any one):

# 1) Project helper — local-first, falls back to the Hub (code/hf_data.py)
import sys, geopandas as gpd
sys.path.append("code")            # add the repo-root "code" folder to the path
from hf_data import data_path
gdf = gpd.read_file(data_path("maps/bolivia339geoqueryOpt.geojson"))

# 2) Direct download from the Hub
from huggingface_hub import hf_hub_download
path = hf_hub_download("cmg777/project2026e", repo_type="dataset",
                       filename="maps/bolivia339geoqueryOpt.geojson")
gdf = gpd.read_file(path)

The datasets library's load_dataset() targets the tabular CSV files in this repository; for GeoJSON geometry use geopandas with either helper above.

Join Key

Use asdf_id to join these geometries with any dataset in data/ (e.g. sdg/sdg.csv, sdgVariables/sdgVariables.csv, satelliteEmbeddings/, nighttimeLights/).