sdc-census10to20 0.1.0

Redistribute 2010-2019 census data onto 2020 census boundaries

sdc-census10to20

Redistribute 2010-2019 census data onto 2020 census boundaries.

This is the Python port of the R package sdc.census10to20, used by the Social Data Commons pipelines to standardize tract- and block-group-level data across the 2020 decennial census boundary changes.

Install

pip install sdc-census10to20

Quick start

import pandas as pd
from sdc_census10to20 import standardize_all

df = pd.DataFrame({
    "geoid": ["51059450100", "51059450200"],
    "year": [2018, 2018],
    "measure": ["population", "population"],
    "value": [3000, 4500],
    "moe": [pd.NA, pd.NA],
})

standardized = standardize_all(df)

See the documentation for the full API reference and a worked example with Virginia tract-level data.

Releasing

This package uses PyPI Trusted Publishing, so no API tokens are stored in GitHub.

One-time setup (done by a maintainer):

1. Register the project name on TestPyPI and PyPI with a "pending publisher" pointing at this repo and workflow file (.github/workflows/publish.yml).
2. Configure two GitHub environments on the repo: testpypi and pypi.

To cut a release:

1. Bump version in pyproject.toml, update CHANGELOG.md.
2. Commit and tag: git tag census10to20-v0.1.0 && git push --tags.
3. The publish.yml workflow builds the wheel, uploads to TestPyPI first, then PyPI on success.

To publish docs:

The docs.yml workflow runs on every push to main that touches packages/sdc-census10to20/. No tag required.

License

MIT