window.md
In recipes: Preprocessing and Feature Engineering Steps for Modeling

error checks

Code
  prep(step_window(rec, y1, size = 6))
Condition
  Error in `step_window()`:
  Caused by error in `prep()`:
  ! `size` should be odd, not 6.

Code
  prep(step_window(rec, y1, size = NA))
Condition
  Error in `step_window()`:
  Caused by error in `prep()`:
  ! `size` must be a whole number, not `NA`.

Code
  prep(step_window(rec, y1, size = NULL))
Condition
  Error in `step_window()`:
  Caused by error in `prep()`:
  ! `size` must be a whole number, not `NULL`.

Code
  step_window(rec, y1, statistic = "average")
Condition
  Error in `step_window()`:
  ! `statistic` must be one of "mean", "median", "sd", "var", "sum", "prod", "min", or "max", not "average".

Code
  prep(step_window(rec, y1, size = 1))
Condition
  Error in `step_window()`:
  Caused by error in `prep()`:
  ! `size` must be a whole number larger than or equal to 3, not the number 1.

Code
  prep(step_window(rec, y1, size = 2))
Condition
  Error in `step_window()`:
  Caused by error in `prep()`:
  ! `size` must be a whole number larger than or equal to 3, not the number 2.

Code
  prep(step_window(rec, y1, size = -1))
Condition
  Error in `step_window()`:
  Caused by error in `prep()`:
  ! `size` must be a whole number larger than or equal to 3, not the number -1.

Code
  prep(step_window(rec, y1, size = 3 + .Machine$double.eps))
Message

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

  -- Inputs 
  Number of variables by role
  predictor: 6

  -- Training information 
  Training data contained 81 data points and no incomplete rows.

  -- Operations 
  * Moving 3-point mean on: y1 | Trained

Code
  prep(step_window(rec, y1, size = 3 + 2 * .Machine$double.eps))
Condition
  Error in `step_window()`:
  Caused by error in `prep()`:
  ! `size` must be a whole number, not the number 3.

Code
  prep(step_window(rec, fac), training = sim_dat)
Condition
  Error in `step_window()`:
  Caused by error in `prep()`:
  x All columns selected for the step should be double or integer.
  * 1 factor variable found: `fac`

Code
  prep(prep(step_window(rec, y1, size = 1000L), training = sim_dat))
Condition
  Error in `step_window()`:
  Caused by error in `prep()`:
  ! `size` should be odd, not 1000.

Code
  prep(bad_names, training = sim_dat)
Condition
  Error in `step_window()`:
  Caused by error in `prep()`:
  ! There were 2 terms selected but 1 value for the new features was passed to `names`.

check_name() is used

Code
  prep(rec, training = dat)
Condition
  Error in `step_window()`:
  Caused by error in `bake()`:
  ! Name collision occurred. The following variable names already exist:
  * `new_value`

error on too large window size

Code
  prep(step_window(recipe(~., data = mtcars), mpg, size = 999))
Condition
  Error in `step_window()`:
  Caused by error in `roller()`:
  ! The window is too large.

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

Code
  bake(rec_trained, new_data = sim_dat[, -1])
Condition
  Error in `step_window()`:
  ! The following required column is missing from `new_data`: x1.

empty printing

Code
  rec
Message

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

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

  -- Operations 
  * Moving 3-point mean 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 
  * Moving 3-point mean on: <none> | Trained

keep_original_cols - can prep recipes with it missing

Code
  rec <- prep(rec)
Condition
  Warning:
  `keep_original_cols` was added to `step_window()` after this recipe was created.
  i Regenerate your recipe to avoid this warning.

printing

Code
  print(rec)
Message

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

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

  -- Operations 
  * Moving 3-point mean on: <none>

Code
  prep(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 
  * Moving 3-point mean on: <none> | Trained