Footings

A Model Building Library

tests gh-pages license codecov version PyPI version

Summary

Footings is a model building Python library. No out-of-the box models are provided. Instead it is a framework library that provides key objects and functions to help users construct custom models.

Purpose

The footings library was developed with the intention of making it easier to develop actuarial models in Python. Actuarial models are a mix of data engineering and math/calculations. In addition, actuarial models are ususally not defined by one calculation, but a series of calculations. So even though the original purpose is actuarial work, if the problem at hand sounds familiar, others might find this library useful.

Principles

The Footings library was designed as framework library using the below principles -

  • Models are a sequence of linked steps

  • Models need to be easy to understand

  • Models need to have validation built in

  • Models need to be easy to audit

  • Models need to be self documenting

  • Models need to be able to scale up

  • Models need to be able to build off other models

  • Model environments should not be monolithic

These all become easier when you can leverage the amazing Python and wider open source ecosystems as well.

For a more insights into principles see principles

Learning

The best way to learn how to use the footings library to build models is to read through the tutorials.