Learn R Programming

⚠️There's a newer version (1.2.1) of this package.Take me there.

recipes (version 1.2.0)

Preprocessing and Feature Engineering Steps for Modeling

Description

A recipe prepares your data for modeling. We provide an extensible framework for pipeable sequences of feature engineering steps provides preprocessing tools to be applied to data. Statistical parameters for the steps can be estimated from an initial data set and then applied to other data sets. The resulting processed output can then be used as inputs for statistical or machine learning models.

Copy Link

Version

Install

install.packages('recipes')

Monthly Downloads

218,325

Version

1.2.0

License

MIT + file LICENSE

Issues

Pull Requests

Stars

Forks

Repository

https://github.com/tidymodels/recipes

Homepage

https://recipes.tidymodels.org/

Maintainer

Max Kuhn

Last Published

March 17th, 2025

Functions in recipes (1.2.0)

Detect if a particular step or check is used in a recipe

Quantitatively check on variables

Check range consistency

Discretize Numeric Variables

.recipes_preserve_sparsity

Does step destroy sparsity of columns

Helper Functions for Profile Data Sets

Helpers for printing step functions

.recipes_estimate_sparsity

Estimate sparity of a recipe

.recipes_toggle_sparse_args

Toggle all auto sparse arguments

Check to see if a recipe is trained/prepared

.get_data_types

Get types for use in recipes

Create a formula from a prepared recipe

get_keep_original_cols

Get the keep_original_cols value of a recipe step

Estimate a preprocessing recipe

developer_functions

Developer functions for creating recipes steps

Extract transformed training set

recipes_eval_select

Evaluate a selection with tidyselect semantics specific to recipes

Wrapper function for preparing recipes within resampling

recipes: A package for computing and preprocessing design matrices.

Prototype of recipe object

recipes_ptype_validate

Validate prototype of recipe object

Create a recipe for preprocessing data

Make a random identification field for steps

recipes_pkg_check

Update packages

recipes_extension_check

Checks that steps have all S3 methods

Box-Cox transformation for non-negative data

step_YeoJohnson

Yeo-Johnson transformation

Methods for selecting variables in step functions

Manually alter roles

Internal Functions

recipes-role-indicator

Role indicators

recipes_remove_cols

Removes columns if options apply

step sets the class of the step and check is for checks.

Using sparse data with recipes

Objects exported from other packages

required_pkgs.step_classdist_shrunken

S3 methods for tracking which additional packages are needed for steps.

Centering numeric data

step_bin2factor

Create a factors from A dummy variable

Distances to class centroids

Sort rows using dplyr

Create counts of patterns using regular expressions

Cut a numeric variable into a factor

remove_original_cols

Removes original columns if options apply

Impute via bagged trees

step_classdist_shrunken

Compute shrunken centroid distances for classification models

High correlation filter

B-spline basis functions

step_factor2string

Convert factors to strings

Distance between two locations

step_discretize

Discretize Numeric Variables

step_filter_missing

Missing value column filter

Create traditional dummy variables

step_dummy_extract

Extract patterns from nominal data

step_dummy_multi_choice

Handle levels in multiple predictors together

Date feature generator

Filter rows using dplyr

step_impute_mean

Impute numeric data using the mean

ICA signal extraction

step_impute_bag

Impute via bagged trees

step_impute_knn

Impute via k-nearest neighbors

Add sin and cos terms for harmonic analysis

step_impute_linear

Impute numeric variables via a linear model

step_impute_lower

Impute numeric data below the threshold of measurement

step_hyperbolic

Hyperbolic transformations

step_impute_median

Impute numeric data using the median

Holiday feature generator

Convert values to predefined integers

step_impute_mode

Impute nominal data using the most common value

Add intercept (or constant) column

Inverse transformation

step_impute_roll

Impute numeric data using a rolling window statistic

Inverse logit transformation

Impute via k-nearest neighbors

Isomap embedding

step_indicate_na

Create missing data column indicators

Create interaction variables

Kernel PCA signal extraction

Linear combination filter

step_medianimpute

Impute numeric data using the median

Radial basis function kernel PCA signal extraction

Logit transformation

Logarithmic transformation

Create a lagged predictor

Polynomial kernel PCA signal extraction

step_meanimpute

Impute numeric data using the mean

step_lowerimpute

Impute numeric data below the threshold of measurement

Natural spline basis functions

Center and scale numeric data

step_num2factor

Convert numbers to factors

Add new variables using dplyr

Non-negative matrix factorization signal extraction

step_nnmf_sparse

Non-negative matrix factorization signal extraction with lasso penalization

step_modeimpute

Impute nominal data using the most common value

Mutate multiple columns using dplyr

Simple value assignments for novel factor levels

Remove observations with missing values

Orthogonal polynomial basis functions

PCA signal extraction

step_ordinalscore

Convert ordinal factors to numeric scores

Partial least squares feature extraction

step_percentile

Percentile transformation

Create a profiling version of a data set

Near-zero variance filter

step_poly_bernstein

Generalized bernstein polynomial basis

Scaling numeric data to a specific range

Collapse infrequent categorical levels

Ratio variable creation

Detect a regular expression

Rename variables by name using dplyr

General variable filter

Scaling numeric data

step_rollimpute

Impute numeric data using a rolling window statistic

Apply (smoothed) rectified linear transformation

Sample rows using dplyr

Rename multiple columns using dplyr

Relevel factors to a desired level

Square root transformation

step_spline_nonnegative

Non-negative splines

Select variables using dplyr

Filter rows by position using dplyr

step_spatialsign

Spatial sign preprocessing

step_spline_convex

step_spline_natural

Natural splines

step_spline_monotone

Monotone splines

Shuffle variables

Select terms in a step function.

tidy.step_BoxCox

Tidy the result of a recipe

step_string2factor

Convert strings to factors

Summarize a recipe

Zero variance filter

Convert ordered factors to unordered factors

Moving window functions

Time feature generator

Update a recipe step

Assign missing categories to "unknown"

update_role_requirements

Update role specific requirements

Add a New Operation to the Current Recipe

Check if all columns are present

Check for required column at bake-time

check_new_values

Check for new values

Check for missing values

Apply a trained preprocessing recipe

case-weight-helpers

Helpers for steps with case weights

Check variable class

Using case weights with recipes

check that newly created variable names don't overlap