Skip to contents

Automatically prep a data.frame/tibble for use in the cubist algorithm.

Usage

hai_cubist_data_prepper(.data, .recipe_formula)

Arguments

.data

The data that you are passing to the function. Can be any type of data that is accepted by the data parameter of the recipes::reciep() function.

.recipe_formula

The formula that is going to be passed. For example if you are using the diamonds data then the formula would most likely be something like price ~ .

Value

A recipe object

Details

This function will automatically prep your data.frame/tibble for use in the cubist algorithm. The cubist algorithm is for regression only.

This function will output a recipe specification.

Author

Steven P. Sanderson II, MPH

Examples

library(ggplot2)

hai_cubist_data_prepper(.data = diamonds, .recipe_formula = price ~ .)
#> 
#> ── Recipe ──────────────────────────────────────────────────────────────────────
#> 
#> ── Inputs 
#> Number of variables by role
#> outcome:   1
#> predictor: 9
#> 
#> ── Operations 
#>  Zero variance filter on: recipes::all_predictors()
rec_obj <- hai_cubist_data_prepper(diamonds, price ~ .)
get_juiced_data(rec_obj)
#> # A tibble: 53,940 × 10
#>    carat cut       color clarity depth table     x     y     z price
#>    <dbl> <ord>     <ord> <ord>   <dbl> <dbl> <dbl> <dbl> <dbl> <int>
#>  1  0.23 Ideal     E     SI2      61.5    55  3.95  3.98  2.43   326
#>  2  0.21 Premium   E     SI1      59.8    61  3.89  3.84  2.31   326
#>  3  0.23 Good      E     VS1      56.9    65  4.05  4.07  2.31   327
#>  4  0.29 Premium   I     VS2      62.4    58  4.2   4.23  2.63   334
#>  5  0.31 Good      J     SI2      63.3    58  4.34  4.35  2.75   335
#>  6  0.24 Very Good J     VVS2     62.8    57  3.94  3.96  2.48   336
#>  7  0.24 Very Good I     VVS1     62.3    57  3.95  3.98  2.47   336
#>  8  0.26 Very Good H     SI1      61.9    55  4.07  4.11  2.53   337
#>  9  0.22 Fair      E     VS2      65.1    61  3.87  3.78  2.49   337
#> 10  0.23 Very Good H     VS1      59.4    61  4     4.05  2.39   338
#> # ℹ 53,930 more rows