1ec5 / osmcounts Goto Github PK

In 2018, when I last worked on this project, I kept a running to-do list for this project. There’s obviously still much more to do, and it may be necessary to rethink some of the goals.

Milestones:

• Webpage with sortable tables of POI counts by state
• …plus sortable tables of road lengths by state — comparable to ImproveOSM
• …plus completeness and rank — ready to commit
• …on a choropleth
• …plus all statistics by core-based statistical area (CBSA) on a bubble map
• …plus all statistics by county — bloggable
• …plus island areas, Canada, and Mexico
• Running weekly
• Easy to set up — ready for SotMUS

Backend:

• Develop PGSQL query for POIs
• Load Economic Census (ECN) into Postgres
• Load OSM–NAICS correspondence into Postgres
• Fix Alaska
• Performantly query polygons
• Develop PGSQL query for all POIs at once (~ 15 min → 7.6 min; 16 min)
• Load Census state boundaries into Postgres
• Join POIs w/ ECN
• Get all-POI query down to 10 minutes
• Performantly cluster nearby duplicate features
• Develop PGSQL query for pathways
• Generalize dual carriageways
• Load Highway Statistics into Postgres
• Load Census CBSA boundaries into Postgres
• Load Economic Census for CBSAs
• Load Economic Census of Island Areas (IA)
• Load Economic Census County Business Patterns (BP)
• Load Census county boundaries into Postgres
• Break down queries by county
• Load HPMS shapefiles into Postgres, especially for turn lanes and lane counts
• Load Ag Census into Postgres
• Load National Bridge Inventory into Postgres
• Research Canada NAICS-based statistics
• Research Mexico NAICS-based statistics

Frontend:

• Generate TSV files
• Write simple webpage displaying TSV in tabular format
• Set up cron job to regenerate statistics
• Write webpage listing sources and caveats
• Add state-by-state choropleth map using d3
• Add county-by-county choropleth map
• Add toggle between choropleth and bubble map

Logistics:

• Fully automate data loading
• Check in scripts
• Investigate dockerization