create_starterframe: Set up a nested data frame with transformations
In rsangole/tidyrabbit: Tidy up Your Carret Work.

Description Usage Arguments Value Examples

In many machinelearning projects you want to start with original values, but also transformations of those values. This function combines all transformations into 1 nested dataframe that contains the transformed data frames

1 2	create_starterframe(transformations = list(), y = NULL, .suffix = NULL, .prefix = NULL)

`transformations`	a named list of transformations
`y`	the values you want to predict, can be empty
`.suffix`	a vector of the same length as the transformation list that contains the suffix you want to apply to your variables
`.prefix`	a vector of the same length as the transformation list that contains the prefix you want to apply to your variables

a nested data frame with trainingsets, name and y values (if supplied)

\notrun{
library(mlbench)
data("BostonHousing")
prepTrain <- preProcess(x = BostonHousing[,preds.original], method = c('BoxCox'))
data_transformation(list(
      original = BostonHousing[,1:13],
      boxcoxed = predict(prepTrain,newdata = BostonHousing[,preds.original]),
      squared  = (BostonHousing[,c(1:3,5:13)])^2
       ),
        .suffix = c(NA, "boxed","squared"))
}