nyc_license is rebuilt from the October 2026 release of the NYC Dog Licensing Dataset. It now has 819,323 rows and adds the 2026 extract year.
nyc_license$breed_name has whitespace and capitalization standardized, so that e.g. "Beagle Mix", "BEAGLE MIX", and "beagle mix" are a single breed.
nyc_license$breed_rc is recoded much more thoroughly. Inverted names are reordered ("Bull Dog, French" is "French Bulldog"), varieties are collapsed to their breed ("Dachshund", "Poodle", "Collie"), and long, abbreviated, and misspelled names are standardized ("American Pit Bull Terrier/Pit Bull" is "Pit Bull"). Crossbreeds are now kept distinct from their breed and are consistently labeled "<Breed> Crossbreed". The former "Pit Bull (or Mix)" and "Labrador (or Crossbreed)" categories are gone.
nycdogs now depends on nycmaps rather than nyczips for zip code tables and maps. nyc_zips and nyc_zip_sf are still attached along with nycdogs.
nyc_bites is rebuilt from the October 2026 release of the DOHMH Dog Bite Data. It now has 39,082 rows and covers 2015 to 2025, up from 10,280 rows covering 2015 to 2017.
nyc_bites$breed_rc is now coded in the same way as nyc_license$breed_rc, so that breeds can be compared across the two tables. breed has whitespace and capitalization standardized.
nyc_bites has several breaking changes. zip_code is now character, not integer. unique_id and year are now integer. unique_id is no longer unique across the table, as the source restarts it in each release; it is unique within year. Columns are reordered, with year after date_of_bite and breed_rc after breed.
nyc_bites gains a zip column, identical to zip_code, to match nyc_license.