UK food hygiene authority intelligence (weekly)

Weekly local-authority food-hygiene signals for the UK, aggregated from the Food Standards Agency's official keyless open-data bulk registers (~613k establishments). Each of the 363 local authorities carries its rating mix (FHRS 0-5 counts for England/Wales/Northern Ireland; Pass / Improvement Required for Scotland under FHIS), a documented 0-100 hygiene_score = 100 * achieved/maximum score (FHRS: rating points over 5*n_rated; FHIS: pass share), unified share_top/share_low for cross-scheme comparison, coverage and freshness (share of rated establishments inspected within the trailing 365 days, median days since rating), and a deterministic hygiene_rank (1 = best; ties: share_top desc, n_rated desc, authority_code asc). low_rating_flag marks authorities with >= 5% of rated establishments at 0-1 stars or Improvement Required. From the second weekly snapshot on, score_change_pts and n_rated_change report week-over-week movement versus the prior snapshot (null on the baseline). Columns: ISO week, as-of date, authority code (primary key), authority name, nation, scheme, establishment and rated counts, per-rating counts, coverage %, top/low rating shares, mean 0-5 rating (FHRS only), hygiene score, freshness metrics, hygiene rank, low-rating flag, the authority bulk file's extract date, row hash. Nullability: FHRS per-rating counts and mean_rating are null for Scottish (FHIS) authorities and vice versa; share_top/share_low/hygiene_score are unified across schemes. Caveats: hygiene_score is a within-week relative measure of the *mix* (an authority can score 100 with one rated establishment); schemes differ, so compare FHRS with FHRS and FHIS with FHIS for like-for-like analysis; bulk files republish on each authority's own schedule, so extract_date varies by row. No business names, addresses, or phone numbers are stored — aggregates only. Sample use: order by hygiene_rank for the cleanest and dirtiest-scoring authorities, or filter low_rating_flag for the week's enforcement hotspots.

Les titres et les descriptions proviennent des sources de données, en anglais.