recipes
recipes copied to clipboard
tidymodels
Pipeable steps for feature engineering and data preprocessing to prepare for modeling
The step_date function currently does not support the day of month field. I would propose adding feature = 'mday' to step_date which would map to ```lubridate::mday```. I would expect this...
## The problem I'm having trouble selecting columns to impute within `step_impute_knn()` using `has_role()`. Thanks! ## Reproducible example ``` r library(recipes) #> Loading required package: dplyr #> #> Attaching package:...
## The problem I'd like a variable included in the recipe step to be ignored when actually fitting a model (whether it is present in the data or not). In...
## The problem When imputing missing values with `step_impute_knn(all_predictors())` the error `Error in gower_work(x = x, y = y, pair_x = pair_x, pair_y = pair_y, n = n, : Column...
The warning could be better handled on our end since it doesn't specify which variables are the problem. ``` r library(recipes) recipe(~., data = mtcars) |> step_spline_b(vs, am) |> prep()...
It is quite simple for an `NaN` to slip in, and it is almost unnoticeable. But it doesn't have to be. `sd_check()` already checks for zero variable, we can extend...
``` r library(recipes) recipe(~ mpg, data = mtcars) |> step_spline_nonnegative(mpg, deg_free = 5, degree = 6) |> prep() |> bake(new_data = NULL) #> Warning: #> # A tibble: 32 ×...
## Feature This idea came to mind after posting [here](https://community.rstudio.com/t/step-percentile-new-data-outside-range-of-training-data/156694/2) on the Posit Community page which was turned into PR [#1075 ](https://github.com/tidymodels/recipes/pull/1075). I believe it would be useful to have...
As the title says. I think it would be helpful to have some helper functions that would extract what variables got selected by a step and which variables were returned....
This comes with an hard to understand error message ``` r library(recipes) recipe(mpg ~ ., mtcars) |> step_nnmf_sparse(disp, drat) |> prep() #> Error in `step_nnmf_sparse()`: #> Caused by error in...
Metadata
Owner
Metadata
Pipeable steps for feature engineering and data preprocessing to prepare for modeling