CPP Data Release v2.0

This is the second public CPP data release. It was produced from internal data build 3.3.1; internal build numbers are retained only for technical provenance.

Highlights:

• restores documented no-prior-pregnancy codes to zero while preserving unknowns as missing;
• rebuilds the expanded obstetric and sibling-history fields using variable-specific sentinel rules;
• canonicalizes 2,421 source-proven alias groups, yielding 60,016 unique unified IDs with zero conflicting populated cells;
• uses semantic public filenames: cpp_clean_core for the stable 185-column schema and cpp_clean_expanded for the 315-column schema;
• retains the older cpp_clean_v3 and cpp_clean_v1 filenames as byte-identical compatibility aliases;
• includes the analysis paper, scripts, data paper, errata, changelogs, repair tools, tests, validation report, and SHA-256 manifests.

The archive passed CRC validation, 51 internal SHA-256 checks, source-semantic validation against all 59,391 CPPVAR records, identifier uniqueness checks, and unit tests. The preceding public release was v1.0, published under Git tag/internal build v3.2.

Artifact-integrity maintenance (2026-08-24): the data paper was rebuilt from source to eliminate a viewer-dependent font defect; the analysis PDF was normalized without changing rendered pages or extracted text; the dated script bundle and direct-download metadata files were synchronized; and machine-local paths were removed from archive audit metadata. No data values, estimates, figures, tables, or interpretations changed in this maintenance pass. Superseded assets remain attached under descriptive provenance names.

Canonical current-source snapshot: use the attached CollaborativePerinatalProject-v2.0-source.zip, built from the non-destructive maintenance tag v2.0-maintenance-20260824. This snapshot includes the final artifact-integrity commits; GitHub's automatic source archives remain tied to the original v2.0 tag for provenance.