recipes: Preprocessing Tools to Create Design Matrices

An extensible framework to create and preprocess design matrices. Recipes consist of one or more data manipulation and analysis "steps". Statistical parameters for the steps can be estimated from an initial data set and then applied to other data sets. The resulting design matrices can then be used as inputs into statistical or machine learning models.

Version: 0.1.3
Depends: R (≥ 3.1), dplyr, broom
Imports: tibble, stats, ipred, dimRed (≥ 0.1.0), lubridate, timeDate, ddalpha, purrr (≥ 0.2.3), rlang (≥ 0.1.1), gower, RcppRoll, tidyselect (≥ 0.1.1), magrittr, Matrix, tidyr, pls
Suggests: testthat, rpart, kernlab, fastICA, RANN, igraph, knitr, rsample (≥ 0.0.2), ggplot2, rmarkdown, covr
Published: 2018-06-16
Author: Max Kuhn [aut, cre], Hadley Wickham [aut], RStudio [cph]
Maintainer: Max Kuhn <max at>
License: GPL-2
NeedsCompilation: no
Materials: NEWS
CRAN checks: recipes results


Reference manual: recipes.pdf
Vignettes: Custom Steps
Dummy Variables and Interactions
Ordering of Steps
Selecting Variables
Basic Recipes
On Skipping Steps
Package source: recipes_0.1.3.tar.gz
Windows binaries: r-devel:, r-release:, r-oldrel:
OS X binaries: r-release: recipes_0.1.3.tgz, r-oldrel: recipes_0.1.3.tgz
Old sources: recipes archive

Reverse dependencies:

Reverse depends: embed
Reverse imports: caret, formulize, healthcareai, olsrr, rsample, tidymodels
Reverse suggests: C50, modelgrid


Please use the canonical form to link to this page.