Standardized regression predictors for use with StatsModels.jl (centered, z-scored, etc.)
Author beacon-biosignals
8 Stars
Updated Last
10 Months Ago
Started In
March 2021


Stable Dev CI Coverage

This package provides convenient and modular functionality for standardizing regression predictors. Standardizing predictors can increase numerical stability of some estimation procedures when the predictors are on very different scales or when they are non-orthogonal. It can also produce more interpretable regression models in the presence of interaction terms.

The examples in the docs demonstrate the use of StandardizedPredictors.jl with GLM.jl, but they will work with any modeling package that is based on the StatsModels.jl formula.

Used By Packages

No packages found.