# pycancensus

> Python interface to Canadian Census data and geographies via the
> CensusMapper API. An explicit port of the R cancensus package with
> verified-equivalent results: same function names, pandas/GeoPandas
> output. Free API key required (https://censusmapper.ca/users/sign_up).

Datasets cover the 1996-2021 Canadian censuses (CA1996...CA21). The core
workflow is: discover variables ("vectors") -> select regions -> call
get_census(), which returns an analysis-ready DataFrame (or GeoDataFrame
with geo_format="geopandas").

## Docs

- [Usage guide for LLMs and agents](https://pycancensus.readthedocs.io/en/latest/llm_usage.html): exact
  current signatures, 0.2.0 changes, R differences, pitfalls
- [Getting started](https://pycancensus.readthedocs.io/en/latest/tutorials/getting_started.html): discovery,
  hierarchies, region selection, data retrieval
- [API reference](https://pycancensus.readthedocs.io/en/latest/api/index.html): all public functions
- [R to Python migration](https://pycancensus.readthedocs.io/en/latest/migration.html): for R cancensus users
- [Working with geometry](https://pycancensus.readthedocs.io/en/latest/tutorials/working_with_geometry.html)
- [Caching and recalled data](https://pycancensus.readthedocs.io/en/latest/tutorials/caching_data.html)

## Source

- [GitHub repository](https://github.com/dshkol/pycancensus)
- [Changelog](https://github.com/dshkol/pycancensus/blob/main/CHANGELOG.md)
- [R cancensus (reference implementation)](https://github.com/mountainMath/cancensus)

## Optional

- [Full docs as one markdown file](https://pycancensus.readthedocs.io/en/latest/llms-full.txt)