Data

Every published series, in three formats, with the dictionary and the provenance record beside them. Nothing on this page is typed by hand — the counts and the bundle size are read from the package's own manifest at render time.

What the package contains

These are separate measures and they are deliberately not all equal. Forcing them to match would hide exactly the things a user needs to know.

Download everything

Download the data bundle (.zip) Download the site code (.zip)

37,403,606 bytes. The zip unpacks to one dated folder holding csv/, parquet/ and xlsx/, plus the data dictionary, the provenance record, the terms, the citation file and the registry under metadata/. Per-member byte counts and SHA-256 hashes are published in BUNDLE_MANIFEST.csv, so the size above can be checked against the file rather than taken on trust.

Formats

CSV, XLSX and Parquet. JSON is not offered as a data format — the registry is reachable as metadata at /metadata/series_registry.json, which is a description of the data, not the data.

• data_dictionary.csv — 62 rows, one per series: name, family, units, source, kind, row count, first and last real observation.
• PROVENANCE.csv — one row per series: who published it, how it was built, how far it really runs, and its triage verdict.
• TERMS.md — terms of use. Each upstream publisher keeps its own.
• CITATION.cff — how to cite this dataset.

Every series

“Through” is the last real observation in the file, not a declared end year. Where a source also publishes projections they are shown separately and are never counted as observations.