// the find
ResidentMario/geoplot
High-level geospatial data visualization library for Python.
geoplot is a high-level geospatial plotting library for Python, sitting on top of cartopy and matplotlib. It gives you choropleth, KDE, sankey, and other standard map types without needing to wrestle cartopy's low-level API directly. Think seaborn but for maps — good for data analysts who want publication-quality static maps fast.
1. The projection API is genuinely well thought out — you pass a cartopy CRS object and it handles the coordinate transforms, which is the part that normally eats an afternoon. 2. The gallery examples are concrete and runnable, covering real datasets (NYC collisions, LA flights, DC street network) rather than toy demos. 3. It stays in the matplotlib ecosystem, so anything you already know about axes, colormaps, and figure layout still applies — no new mental model required. 4. conda-forge packaging means geospatial dependency hell (GDAL, PROJ, GEOS) is handled for you.
1. The author explicitly calls it 'maintenance mode' — no new features coming, and the last meaningful activity was 2024. For anything beyond the existing plot types you're on your own. 2. It only produces static matplotlib figures; if you need interactive maps (zoom, hover, tooltips), this is the wrong tool entirely — use folium or plotly. 3. The entire library lives in one 2000-line file (geoplot/geoplot.py), which makes it hard to extend or contribute to. 4. cartopy itself has a complex install story on some platforms, and geoplot inherits all of that friction despite the conda-forge packaging.