tests/testthat/_snaps/logit.md

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

out of bounds logit trans

Code
  prep(rec, training = ex_dat, verbose = FALSE)
Condition
  Error in `step_logit()`:
  Caused by error in `binomial()$linkfun()`:
  ! Value -0.77772 out of range (0, 1)

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

Code
  bake(rec_trained, new_data = ex_dat[, 2:3])
Condition
  Error in `step_logit()`:
  ! 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 
  * Logit transformation 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 
  * Logit transformation on: <none> | Trained

printing

Code
  print(rec)
Message

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

  -- Inputs 
  Number of variables by role
  predictor: 3

  -- Operations 
  * Logit transformation on: x1

Code
  prep(rec)
Message

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

  -- Inputs 
  Number of variables by role
  predictor: 3

  -- Training information 
  Training data contained 20 data points and no incomplete rows.

  -- Operations 
  * Logit transformation on: x1 | Trained

bad args

Code
  recipe(~., data = ex_dat) %>% step_logit(x1, offset = "sure") %>% prep()
Condition
  Error in `step_logit()`:
  Caused by error in `prep()`:
  ! `offset` must be a number, not the string "sure".

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