/recipes

A preprocessing engine to generate design matrices

Primary LanguageHTML

Build Status Coverage status CRAN_Status_Badge Downloads

The recipes package is an alternative method for creating and preprocessing design matrices that can be used for modeling or visualization. From wikipedia:

In statistics, a design matrix (also known as regressor matrix or model matrix) is a matrix of values of explanatory variables of a set of objects, often denoted by X. Each row represents an individual object, with the successive columns corresponding to the variables and their specific values for that object.

While R already has long-standing methods for creating these matrices (e.g. formulas and model.matrix), there are some limitations to what the existing infrastructure can do.

The idea of the recipes package is to define a recipe or blueprint that can be used to sequentially define the encodings and preprocessing of the data (i.e. "feature engineering"). For example, to create a simple recipe containing only an outcome and predictors and have the predictors centered and scaled:

library(recipes)
library(mlbench)
data(Sonar)
sonar_rec <- recipe(Class ~ ., data = Sonar) %>%
  step_center(all_predictors()) %>%
  step_scale(all_predictors())

To install it, use:

install.packages("recipes")

## for development version:
require("devtools")
install_github("tidymodels/recipes")