Where have all the post offices gone? Discovering neighborhood facilities with Python and OSM
summary
Uses Python with PyOsmium, GeoPandas, Bokeh and Overpass/Geofabrik extracts to parse OSM .pbf snapshots, spatially join post office nodes to admin polygons, enrich with Wikidata population, and build an interactive time-slider map that links back to OSM objects. Finds specific changes: a home district loses two post offices between 2024 and 2025, some districts show large net losses by population, and data artifacts appear (mass oak imports, duplicated names, highway cameras). The CLI workflow and visualization work and expose deletions and tagging issues.
AI-written summary · about
chapters (21)
- 0:00Introduction
- 1:14Why tracking post office closures
- 2:15OpenStreetMap data structure
- 4:43Finding post offices in OSM
- 7:09Downloading OSM data extracts
- 8:56Geospatial analysis workflow
- 10:01DINER OSM project setup
- 10:44Processing OSM data with Pyosmium
- 13:15Joining post offices with neighborhoods
- 14:41Calculating areas and filtering
- 15:44Visualizing results with Bokeh
- 16:47Post offices over time map
- 18:11Exploring other resources
- 19:54Where post offices went
- 21:41Summary
- 22:07Q&A
- 23:35OSM and Python resources
- 24:42Mapping population to neighborhoods
- 25:44OSM data accuracy
- 26:46Spatial distribution of post offices
- 27:48OSM data quality in dumps
related talks
01h 41mOpenStreetWhat? Mapping The World With Open Data
33mAdd geocoordinates to your addresses in PostgreSQL
50mMaking use of OpenStreetMap data with Python
11mHow we use OSM data for news mapping
40mGIS capability of PostgreSql using POSTGIS & OSM DATA
05mUsing OpenStreetMap with the Python OpenPoliceData package
22mUnderstanding our cities with FOSS4G
03h 59mGeospatial Public Policy Analysis with GeoPandas