recipes
recipes copied to clipboard
tidymodels
Pipeable steps for feature engineering and data preprocessing to prepare for modeling
`step_dummy()` gives a warning about NA values, but it isn't specified that it is coming from `step_dummy()` or what could be done to fix this problem. Adding `step_unknown()` seems to...
In the step_other() function, it would be nice to have the option of naming the "other" value as the name combination of the collapsed levels. For example, if a variable...
For tidymodels/workflows#112, we'll create an S3 method for steps and the recipe that indicates whether novel factor levels are handled in the step (and the recipe). Most steps will default...
In cases where `step_other` does not change factor levels of a variable, novel values will create NA (as documented). Please add an example that e.g. prepends `step_novel(all_nominal_predictors(),new_level="other")` to `step_other` to...
A few times I've come across a situation where I want have very few levels of a numeric variable, and I want these to be factors. But I've noticed every...
This is an option to `prep()` but it is really intrinsic to the recipe so it should probably an argument there. We can soft-deprecate the `prep()` argument.
Currently, step_other() is equivalent to fct_lump_prop() or fct_lump_min() depending on the threshold argument value. It would be nice to have an optional n argument (& a ties.method) that would make...
> _I'm just familiarizing myself with `recipes`, so apologies if there's a simple solution that I haven't been able to dig up._ # Suggestions - Just like there are selectors...
There are some places where the operations [might be costly and are embarrassingly parallel](https://github.com/topepo/recipes/blob/master/R/bag_imp.R). A parallel option might be a good idea here but might bite the user if the...
The "splines" package used in the `step_ns` and `step_bs` does not allow for cyclic splines, while mgcv::s does. Would it be possible to add a step wrapping calls to s...
Metadata
Owner
Metadata
Pipeable steps for feature engineering and data preprocessing to prepare for modeling