Open-source software & research for civic data.

Python for NYC civic data — 311, subway accessibility, shared geography — and reproducible-research tooling for causal / econometric work that generalizes beyond the city.

Toolkits

NYC civic data

nyc-geo-toolkit
nyc-geo-toolkitstable
pip install nyc-geo-toolkit

Shared geography foundation — boundaries, normalization, and spatial helpers

DocsPyPIv0.4.1
nyc311
nyc311stable
pip install "nyc311[all]"

311 complaint analysis — SDK, CLI, and a factor-factory adapter

DocsPyPIv1.0.3
subway-access
subway-accessstable
pip install "subway-access[factor-factory,tearsheets]"

NYC subway accessibility — stations, outages, and tract-level need

DocsPyPIv0.5.1

AI / ML / Data Science

citeformer
citeformerstable
pip install citeformer

Verifiably-cited LLM text via constrained decoding

DocsPyPIv0.3.0
rrxiv-pythonbeta
pip install "rrxiv @ git+https://github.com/random-walks/rrxiv-python.git@main"

Reference implementation of the rrxiv preprint protocol

Docsv0.1.0
factor-factorystable
pip install "factor-factory[all]"

Panel + engine framework for causal / econometric analysis

DocsPyPIv1.0.3
jellycellstable
pip install "jellycell[server]"

Plain-text notebooks with content-hashed caching

DocsPyPIv1.4.0

Featured

Rat Containerization and Complaint Volume: Did NYC's Mandatory Bin Rollout Causally Reduce Rodent Sightings?

/nyc311

Did NYC's trash-bin containerization reduce rats? Staggered DiD across the 2023 pilot and 2024 citywide rollout finds roughly 12 fewer 311 rat complaints per neighborhood per month, corroborated by DOHMH inspections.

Staggered DiDTWFECallaway-Sant'AnnaSun-AbrahamBorusyak-Jaravel-SpiessSynthetic ControlParallel TrendsMDE / PowerBH CorrectionMoran's IRDDfactor-factoryjellycellnyc311

Research

All