deemanone/padelnomics
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
453 Commits 1 Branch 4 Tags
9507767de1fc6af3f4773f44ab36c73c114d0fec
Commit Graph

2 Commits

Author SHA1 Message Date
Deeman
55f179ba54 fix(transform): increase geonames object size limit and remove stale column ref 
- stg_population_geonames: add maximum_object_size=40MB to read_json() call;
  geonames cities_global.json.gz is ~30MB, exceeding DuckDB's 16MB default
- dim_locations: remove stale 'population_year AS population_year' column ref;
  stg_population_geonames has ref_year, not population_year — caused BinderException

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
 2026-02-25 09:56:05 +01:00
Deeman
ebfdc84a94 feat(transform): add dim_locations + dual market scoring models 
dim_locations (foundation):
- Seeded from stg_population_geonames (all locations, not venue-dependent)
- Grain: (country_code, geoname_id)
- Enriched with: padel venues within 5km, nearest court distance (ST_Distance_Sphere),
  tennis courts within 25km, country income
- Covers zero-court Gemeinden for opportunity scoring

location_opportunity_profile (serving) — Padelnomics Marktpotenzial-Score:
- Answers "Where should I build?" — no padel_venue_count filter
- Formula: population (25) + income (20) + supply gap inverted (30) +
           catchment gap (15) + tennis culture (10) = 100pts
- Sorted by opportunity_score DESC

city_market_profile (serving) — Padelnomics Marktreife-Score:
- Add saturation discount (×0.85 when venues_per_100k > 8)
- Update header comment to reference Marktreife-Score branding
- Kept WHERE padel_venue_count > 0 (established markets only)
- column name market_score unchanged (avoids downstream breakage)

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
 2026-02-24 16:28:16 +01:00