This package provides common abstractions and utilities for specifying, fitting, and evaluating statistical models. The goal is to provide an API for package developers implementing different kinds of statistical models (see the GLM package for example), and utilities that are generally useful for both users and developers when dealing with statistical models and tabular data.
- [Formula notation](@ref Modeling-tabular-data) for transforming tabular data into numerical arrays for modeling.
- Mechanisms for [extending the
@formulanotation](@ref Internals-and-extending-the-formula-DSL) in external modeling packages.
- [Contrast coding](@ref Modeling-categorical-data) for categorical data
- Types and [API](@ref Modeling) for fitting and working with statistical models, extending StatsBase.jl's API to tabular data.
Much of this package was formerly part of [DataFrames.jl](https://www.github.com/JuliaStats/DataFrames.jl) and historically only handled tabular data in the form of a `DataFrame`, but currently supports any table that supports the minimal [Tables.jl](https://github.com/JuliaData/Tables.jl/) interface. It's thus a relatively light dependency.