# bad values Code sacr_rec %>% step_profile(everything(), profile = vars(sqft)) %>% prep(data = Sacramento) Condition Error in `step_profile()`: Caused by error in `prep()`: ! The profiled variable cannot be in the list of variables to be fixed. --- Code sacr_rec %>% step_profile(sqft, beds, price, profile = vars(zip, beds)) %>% prep(data = Sacramento) Condition Error in `step_profile()`: Caused by error in `prep()`: ! Only one variable should be profiled --- Code sacr_rec %>% step_profile(city, profile = vars(sqft), pct = -1) %>% prep(data = Sacramento) Condition Error in `step_profile()`: ! `pct should be on [0, 1]` --- Code sacr_rec %>% step_profile(city, profile = vars(sqft), grid = 1:3) %>% prep( data = Sacramento) Condition Error in `step_profile()`: ! `grid` should have two named elements. See ?step_profile --- Code sacr_rec %>% step_profile(city, profile = vars(sqft), grid = list(pctl = 1, len = 2)) %>% prep(data = Sacramento) Condition Error in `step_profile()`: ! `grid$pctl should be logical.` --- Code fixed(rep(c(TRUE, FALSE), each = 5)) Condition Error in `error_cnd()`: ! Conditions must have named data fields # empty printing Code rec Message -- Recipe ---------------------------------------------------------------------- -- Inputs Number of variables by role outcome: 1 predictor: 10 -- Operations * Profiling data set for: mpg --- 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 * Profiling data set for: mpg | Trained # printing Code print(rec) Message -- Recipe ---------------------------------------------------------------------- -- Inputs Number of variables by role predictor: 10 -- Operations * Profiling data set for: sqft --- Code prep(rec) Message -- Recipe ---------------------------------------------------------------------- -- Inputs Number of variables by role predictor: 10 -- Training information Training data contained 20 data points and no incomplete rows. -- Operations * Profiling data set for: sqft | Trained