tests/testthat/_snaps/corr.md

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

many missing values

Code
  filtering_trained <- prep(filtering, training = dat2, verbose = FALSE)
Condition
  Warning:
  The correlation matrix has missing values. 1 column was excluded from the filter.

occasional missing values

Code
  filtering_trained <- prep(filtering, training = dat3, verbose = FALSE)
Condition
  Warning:
  The correlation matrix has sporadic missing values. Some columns were excluded from the filter.

case weights

Code
  filtering_trained
Message

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

  -- Inputs 
  Number of variables by role
  predictor:    8
  case_weights: 1

  -- Training information 
  Training data contained 100 data points and no incomplete rows.

  -- Operations 
  * Correlation filter on: V3_dup, V1, V2 | Trained, weighted

Code
  filtering_trained
Message

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

  -- Inputs 
  Number of variables by role
  predictor:    8
  case_weights: 1

  -- Training information 
  Training data contained 100 data points and no incomplete rows.

  -- Operations 
  * Correlation filter on: V6, V1, V3 | Trained, ignored weights

corr_filter() warns on many NA values

Code
  tmp <- recipe(~., data = mtcars) %>% step_corr(all_predictors()) %>% prep()
Condition
  Warning:
  Too many correlations are `NA`; skipping correlation filter.

empty printing

Code
  rec
Message

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

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

  -- Operations 
  * Correlation filter on: <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 
  * Correlation filter on: <none> | Trained

printing

Code
  print(rec)
Message

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

  -- Inputs 
  Number of variables by role
  predictor: 7

  -- Operations 
  * Correlation filter on: all_predictors()

Code
  prep(rec)
Message

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

  -- Inputs 
  Number of variables by role
  predictor: 7

  -- Training information 
  Training data contained 100 data points and no incomplete rows.

  -- Operations 
  * Correlation filter on: V6 and V1 | Trained

bad args

Code
  recipe(mpg ~ ., mtcars) %>% step_corr(all_predictors(), threshold = 2) %>% prep()
Condition
  Error in `step_corr()`:
  Caused by error in `prep()`:
  ! `threshold` must be a number between 0 and 1, not the number 2.

Code
  recipe(mpg ~ ., mtcars) %>% step_corr(all_predictors(), use = "this") %>% prep()
Condition
  Error in `step_corr()`:
  Caused by error in `prep()`:
  ! `use` must be one of "all.obs", "complete.obs", "pairwise.complete.obs", "everything", or "na.or.complete", not "this".

Code
  recipe(mpg ~ ., mtcars) %>% step_corr(all_predictors(), method = "my dissertation") %>%
    prep()
Condition
  Error in `step_corr()`:
  Caused by error in `prep()`:
  ! `method` must be one of "pearson", "kendall", or "spearman", not "my dissertation".

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