tests/testthat/_snaps/impute_knn.md

In tidymodels/recipes: Preprocessing and Feature Engineering Steps for Modeling

options

Code
  new_nn <- gower_topn(x = dat_2, y = dat_1, n = 2, eps = 2)$index
Condition
  Warning in `gower_work()`:
  skipping variable with zero or non-finite range.

impute_with errors with nothing selected

Code
  recipe(~., data = mtcars) %>% step_impute_knn(all_predictors(), impute_with = NULL) %>%
    prep()
Condition
  Error in `step_impute_knn()`:
  ! `impute_with` must not be empty.

warn if all values of predictor are missing

Code
  tmp <- recipe(~., data = mtcars) %>% step_impute_knn(mpg, disp, vs) %>% prep()
Condition
  Warning:
  All predictors are missing; cannot impute.
  Warning:
  All predictors are missing; cannot impute.
  Warning:
  All predictors are missing; cannot impute.

error on wrong options argument

Code
  recipe(~., data = mtcars) %>% step_impute_knn(all_predictors(), options = list(
    wrong = "wrong")) %>% prep()
Condition
  Error in `step_impute_knn()`:
  ! Valid values for `options` are "eps" and "nthread".

Code
  recipe(~., data = mtcars) %>% step_impute_knn(all_predictors(), options = c(
    wrong = "wrong")) %>% prep()
Condition
  Error in `step_impute_knn()`:
  ! `options` should be a named list.

bake method errors when needed non-standard role columns are missing

Code
  bake(imputed_trained, new_data = biomass[, c(-4)])
Condition
  Error in `step_impute_knn()`:
  ! The following required column is missing from `new_data`: hydrogen.

empty printing

Code
  rec
Message

  -- Recipe ----------------------------------------------------------------------

  -- Inputs 
  Number of variables by role
  outcome:    1
  predictor: 10

  -- Operations 
  * K-nearest neighbor imputation for: <none>

Code
  rec
Message

  -- Recipe ----------------------------------------------------------------------

  -- Inputs 
  Number of variables by role
  outcome:    1
  predictor: 10

  -- Training information 
  Training data contained 32 data points and no incomplete rows.

  -- Operations 
  * K-nearest neighbor imputation for: <none> | Trained

printing

Code
  print(rec)
Message

  -- Recipe ----------------------------------------------------------------------

  -- Inputs 
  Number of variables by role
  outcome:   1
  predictor: 5

  -- Operations 
  * K-nearest neighbor imputation for: carbon and nitrogen

Code
  prep(rec)
Message

  -- Recipe ----------------------------------------------------------------------

  -- Inputs 
  Number of variables by role
  outcome:   1
  predictor: 5

  -- Training information 
  Training data contained 536 data points and no incomplete rows.

  -- Operations 
  * K-nearest neighbor imputation for: carbon and nitrogen | Trained

bad args

Code
  recipe(~., data = mtcars) %>% step_impute_knn(all_predictors(), neighbors = 0L) %>%
    prep()
Condition
  Error in `step_impute_knn()`:
  Caused by error in `prep()`:
  ! `neighbors` must be a whole number larger than or equal to 1, not the number 0.

tidymodels/recipes documentation built on Jan. 25, 2025, 7:30 a.m.