Skip to content

README.md

Latest commit

 

History

History
47 lines (35 loc) · 2.1 KB

README.md

File metadata and controls

47 lines (35 loc) · 2.1 KB
Raw

Name ground-truth data

Real first/last name data powering the gender-aware, country-aware faker providers (faker2.realnames in Python, the real-names feature in Rust).

Files

File Rows Notes
first_names.parquet 1,433,891 139 countries, gender (m/f), relative frequency, country_rank, continent
last_names.parquet 30 small sample only — surnames are not comprehensive

Format: Parquet, zstd-compressed, dictionary-encoded low-cardinality columns. first_names.parquet is 11.5 MB and loads fully in ~0.1 s.

Schema (first_names)

name, name_ascii, country_code, country_name, continent, gender, unisex, country_rank, frequency, country_share, phonetic

  • frequency — relative share within (country_code, gender) (weighted sampling).
  • country_share — relative share within country_code (cross-gender comparable; powers detect_country / homophones probabilities).
  • phonetic — double-metaphone key (homophones method="metaphone").
  • ipa — IPA transcription (homophones method="ipa").

frequency is a relative share (0..1) within each (country_code, gender) group — it preserves the weighting used for sampling but deliberately carries no raw population counts. Absolute headcounts are not exposed.

Provenance

Extracted from a PostgreSQL 17 custom dump (pg_dump -Fc, 3.57 GB) restored locally with PG 18 client + pgvector. Source table also carried phonetic columns (ipa, metaphone, double_metaphone, soundex, panphon_vector) dropped here as irrelevant to generation.

Regenerating

-- frequency normalized to a relative share; raw counts never leave the DB
\copy (SELECT name, name_ascii, country_code, country_name, continent, gender,
              unisex, country_rank,
              (frequency::float8 / NULLIF(SUM(frequency)
                 OVER (PARTITION BY country_code, gender),0))::real AS frequency
       FROM first_names) TO 'first_names.csv' WITH (FORMAT csv, HEADER true)

then convert CSV -> parquet (pyarrow, zstd, dictionary-encode country_code/country_name/continent/gender).