The KROEG Index

A "third place" is a place outside home and work where people spend time — cafés, shops, community centers, sports clubs, and the like. This index ranks every PC4 area (a 4-digit Dutch postal code area) by how many third places it has, relative to all other PC4 areas, using company data from FirmBackbone. The index is a percentile-style score from 0 to 100. A score of 30 means 70% of PC4 areas have more third places of that type, and 30% have fewer. Third places are split into six categories: Code Category a Food and beverage establishments b Retail and services c Community spaces (noncommercial) d Cultural venues e Sport and leisure venues f Medical services Three aggregate scores are also provided: a total across all six categories, a non-medical total (a–e), and a non-medical, non-retail total (a, c, d, e). Columns Column Description PC4 4-digit postal code area Stedelijkheid Urbanisation level of the area, based on address density (levels below) geometry Polygon geometry of the PC4 area, used to draw the map Every ranking measure comes in two versions: a national _ranked column (ranked against all PC4 areas) and a _ranked_by_urbanisation column (ranked only against PC4 areas in the same urbanisation class). The measures are: Ranking column (national version shown) Measures third_places_total_ranked All third places third_places_non_medical_total_ranked Non-medical third places (a–e) third_places_medical_ranked Medical services (f) third_places_food_beverage_ranked Food and beverage establishments (a) third_places_retail_ranked Retail and services (b) third_places_community_ranked Community spaces (c) third_places_cultural_ranked Cultural venues (d) third_places_sport_leisure_ranked Sport and leisure venues (e) third_places_non_medical_non_retail_total_ranked Non-medical, non-retail third places (a, c, d, e) Each row above also exists with a _by_urbanisation suffix (e.g. third_places_total_ranked_by_urbanisation). Stedelijkheid (urbanisation) levels: Level Description Address density 1 Extremely urbanised 2,500+ per km² 2 Strongly urbanised 1,500–2,500 per km² 3 Moderately urbanised 1,000–1,500 per km² 4 Hardly urbanised 500–1,000 per km² 5 Not urbanised Fewer than 500 per km² Data pipeline available only inside SANE secure computing environment of ODISSEI Nederland The maps used by the dashboard are produced by four notebooks, run in order: 1_data_processing_sbi_codes.ipynb — Loads 2024 KVK (Dutch business registry) data and a codebook mapping each SBI code (industry classification) to a third-place category, then counts, per PC4 area, how many businesses fall into each category. 2_kvk_cbs_linking.ipynb — Loads CBS open geospatial PC4 data, keeps the geometry, PC4, and Stedelijkheid columns, merges it with the category counts from step 1 on PC4, and saves the result as a .gpkg file. 3_index_ranking.ipynb — Loads that .gpkg file and, for every third-place column, ranks PC4 areas by count and rescales the ranks to run from 0 to 100, then repeats the same ranking within each urbanisation category. The raw counts are dropped afterward, leaving only the ranking columns described above, and the result is saved as the final .gpkg. 4_visualization.ipynb — Loads the final .gpkg, checks that it loaded correctly and that the coordinate reference system is set as expected, builds an example map with matplotlib and geopandas, and then plots a color-coded map for every index. These are the maps that get exported into plots/ for the dashboard below. A future iteration may add Folium for additional plotting, pending approval of offline_Folium for use in SANE.