era_pl: empirical research in accounting with Python

era_pl provides Polars data and functions for the Python/Polars edition of Empirical Research in Accounting: Tools and Methods. Install the era-py distribution and import era_pl in Python. This is the API used by the book and its companion materials.

The same distribution also supplies the era_py pandas/Ibis API. Both namespaces use the same packaged datasets, but their function coverage and arguments differ.

See the changelog for recent additions and fixes, or the Python/Polars book for worked examples. The function reference describes the APIs, and the dataset catalog documents the included data.

Installation

Install the package with its Polars extra:

pip install "era-py[polars]"

Add the extras required for your workflow:

pip install "era-py[polars,tables]"
pip install "era-py[polars,plots]"

Python 3.10 or later is required. The Polars extra requires Polars 2 or later.

The development version can be installed from GitHub:

pip install "era-py[polars] @ git+https://github.com/iangow/era_py.git"

Packaged data

Load a dataset and inspect its documentation:

from era_pl import data_info, load_data

camp = load_data("camp_attendance")
camp_info = data_info("camp_attendance")

data_info(name) returns a dataset’s description, source, dimensions, schema, and variable definitions without loading its observations. Use data_catalog() to inspect every packaged dataset.

Polars helpers

Importing era_pl also registers the .era namespace on Polars expressions and dataframes. For example, form quantile groups with:

import polars as pl
import era_pl

data = pl.DataFrame({"value": [1, 2, 3, 4, 5, 6]})
data.with_columns(bucket=pl.col("value").era.ntile(3))

The reference groups functions for event dates, data retrieval, models, data cleaning, tables, plots, and SEC indexes.

Pandas and Ibis API

For pandas workflows, install era-py and import era_py. Its Ibis/Postgres event-return and WRDS helpers additionally require pip install "era-py[ibis]". These functions have a separate section in the reference.

The era_pl event helpers currently construct event dates and windows; era_pl.get_event_rets() and era_pl.copy_inline() are unimplemented. Use the documented Ibis API where cumulative event-return calculations are needed.