Cabo Verde Y-chromosome statistical supplement
Version: 1.0.0
Prepared date: 2026-09-30
Project: Badiu Nation Research Project
Article: https://www.badiunation.page/essays/ydna-island-weighting
Input dataset version: 2026-09-13
Participants: 1,257 in nine paternal-island groups

First release (v1.0.0), versioned with the project website.
It contains aggregate data only, not participant records or Y-STR haplotypes.
The script outputs are the numerical source for Tables S1-S7 and the added
manuscript uncertainty and sensitivity results.

This supplement v1.0.0 is included with the Zenodo article version 3.0 package,
following version 2 (https://doi.org/10.5281/zenodo.22774640).
The v3 deposit is prepared but not yet published; its DOI will be added after
publication. It replaces the older PDF in Zenodo version 2 and adds
the aggregate data, calculation script, and tables as supporting files.

Files
  aggregate-data.json      Frozen counts, category definitions, scenario weights.
  census-inputs.json       INE population extraction, query, and source metadata.
  LICENSE-MIT.txt          License for the calculation script.
  aggregate-counts.csv     Same counts; one island per row, one category per column.
  calculate-supplement.mjs Standalone JavaScript calculation script.
  results.json             Full-precision calculated results (proportion units).
  tables.md                Formatted supplementary tables and interpretation.
  manifest.json            Version metadata and SHA-256 hashes of these files.

Reproduce
Tested with Node.js v24.15.0 on Windows.
Download all files into one directory. With Node.js installed, run:
  node calculate-supplement.mjs
  node calculate-supplement.mjs --json
  node calculate-supplement.mjs --csv
These commands print tables, full-precision results, and counts respectively.
They do not modify files. No additional packages or network access are needed.
The repository copy is scripts/ydna-supplement.mjs. In the repository,
npm run check:research checks the archived output and manuscript for agreement.

Definitions and assumptions
Island means reported birthplace of the paternal grandfather, not residence.
CSV counts are integer participant assignments, including observed zeros.
A1a + E1a + E1b1a + E2b is an analytical grouping, not an ancestry proportion.
The main standardised summary uses 2021 census island resident shares (63.17%).
These cover all ages and both sexes, not adult men or paternal-grandfather
origins; this is not a validated national paternal-lineage frequency estimate.
Table 1 is a secondary illustrative scenario of undocumented origin.
Island intervals use Wilson scores. Weighted normal intervals use conditional
fixed-stratum variance, independent Bernoulli observations within groups, and
known fixed weights. Paired differences use weight differences on the same data.
Recruitment bias, residual relatedness, uncertain weights, and resident versus
paternal-origin population differences are not covered by these intervals.
The reported variance ratio is not an established survey design effect.
No full 9-by-10 exact or Monte Carlo test has been performed.

Sources
Study counts and Table 1 weights: aggregate dataset version 2026-09-13,
as reproduced in the article. The demographic provenance of Table 1 weights
remains unresolved; they must not be described as validated population weights.
Census announcement: Instituto Nacional de Estatistica, Cabo Verde (2022).
INE apresenta os resultados definitivos do V Recenseamento Geral da Populacao
e Habitacao (RGPH-2021), 4 April 2022.
https://ine.cv/censo2020/?p=5489
The page confirms 505,044 people counted, including 491,233 residents.
It contains no island totals. The previously suggested presentation-PDF
address redirected to INE's Not Found page when checked on 2026-09-30.

Exact integer island inputs: INE ArcGIS layer ILHAS (ID 1), within service
INECV_GEODATA/RE2023_INDICADORES_GERAIS_EMPRESAS_ILHA.
This service concerns 2023 enterprise indicators; the ILHAS layer also carries
the population field Pop_Res_2021, alias "População Residente (2021)".
We used this population field, not any enterprise-count field.
https://webgis10.ine.cv/arcgisinecv/rest/services/INECV_GEODATA/RE2023_INDICADORES_GERAIS_EMPRESAS_ILHA/MapServer/1
Values first checked 2026-09-29 and retrieved again 2026-09-30:
Santiago 273988; Sao Vicente 75845; Santo Antao 36950; Fogo 33754;
Sal 33615; Boa Vista 12798; Sao Nicolau 12306; Maio 6330; Brava 5647.
Total: 491,233 residents. Santa Luzia's value is null and it is outside the
nine inhabited-island study groups; its null was not converted to zero.
census-inputs.json preserves the normalized extraction and full query URL.
The script reads that file; results.json and Table S3 reproduce the inputs.

Version and preservation
The manifest hashes the exact UTF-8, LF-line-ending bytes in this directory.
Keep the files together and preserve their bytes when checking these hashes.
Any change to a published snapshot should receive a new version directory.
This is a versioned website supplement, not a new Zenodo deposit.
The existing article PDF and https://doi.org/10.5281/zenodo.22745204 predate
this supplement. No new DOI or Zenodo deposition is claimed for version 1.0.0.
Licenses
Study aggregate data, generated tables (tables.md and results.json), and
project-authored documentation: Creative Commons Attribution 4.0 International.
https://creativecommons.org/licenses/by/4.0/
Attribute Badiu Nation Research Project, supplement v1.0.0, and indicate changes.
Calculation script (calculate-supplement.mjs): MIT License; see LICENSE-MIT.txt.
Third-party INE material is attributed separately. These licenses do not claim
ownership of, or relicense, INE source material or private participant records.