Map Data Preprocessing

Use preprocess_map_data.py to convert one wide CBS/export CSV into the long CSV and filtered GeoJSON expected by Map Explorer.

Input

The current map uses wijk-level output. The raw CSV must contain:

• wijk_code, for example WK034400
• n_sample
• profile_1, profile_2, profile_3, profile_4, profile_5

The same command also supports buurt-level output by using --geo-level buurt and a buurt_code column.

Optional column:

• gender

Profile values may be proportions (0 to 1) or percentages (0 to 100). The script writes percentages.

Command

uv run python scripts/preprocess_map_data.py \
  --input data/raw/profile_export.csv \
  --geo-level wijk \
  --geo-year 2026 \
  --geojson data/geo/wijk_2026.geojson \
  --output-public public

If --geojson is omitted, the script looks for data/geo/<geo-level>_<geo-year>.geojson.

Output

For --geo-level wijk, the script writes:

• public/wijk_5_processed.csv
• public/wijk_2026.geojson
• public/preprocess_report.json

The processed wijk CSV schema is:

• profile
• wijk_code
• wijk (compatibility alias used by the earlier vaxprofiles fork)
• value
• n_sample
• optional filter columns such as gender

For --geo-level buurt, the equivalent files are public/buurt_5_processed.csv and public/buurt_2026.geojson, with buren as the compatibility alias.

Checks

The script fails when required columns are missing, profile values do not sum to about 100, duplicates exist for a region/filter combination, values fall outside the expected range, or more than 5% of input region codes do not join to the GeoJSON.