R recipes step_profile 創建數據集的分析版本

step_profile(
  recipe,
  ...,
  profile = NULL,
  pct = 0.5,
  index = 1,
  grid = list(pctl = TRUE, len = 100),
  columns = NULL,
  role = NA,
  trained = FALSE,
  skip = FALSE,
  id = rand_id("profile")
)

data(Sacramento, package = "modeldata")

# Setup a grid across beds but keep the other values fixed
recipe(~ city + price + beds, data = Sacramento) %>%
  step_profile(-beds, profile = vars(beds)) %>%
  prep(training = Sacramento) %>%
  bake(new_data = NULL)
#> # A tibble: 6 × 3
#>   city      price  beds
#>   <fct>     <int> <int>
#> 1 ANTELOPE 220000     1
#> 2 ANTELOPE 220000     2
#> 3 ANTELOPE 220000     3
#> 4 ANTELOPE 220000     4
#> 5 ANTELOPE 220000     5
#> 6 ANTELOPE 220000     8

##########

# An *additive* model; not for use when there are interactions or
# other functional relationships between predictors

lin_mod <- lm(mpg ~ poly(disp, 2) + cyl + hp, data = mtcars)

# Show the difference in the two grid creation methods

disp_pctl <- recipe(~ disp + cyl + hp, data = mtcars) %>%
  step_profile(-disp, profile = vars(disp)) %>%
  prep(training = mtcars)

disp_grid <- recipe(~ disp + cyl + hp, data = mtcars) %>%
  step_profile(
    -disp,
    profile = vars(disp),
    grid = list(pctl = FALSE, len = 100)
  ) %>%
  prep(training = mtcars)

grid_data <- bake(disp_grid, new_data = NULL)
grid_data <- grid_data %>%
  mutate(
    pred = predict(lin_mod, grid_data),
    method = "grid"
  )

pctl_data <- bake(disp_pctl, new_data = NULL)
pctl_data <- pctl_data %>%
  mutate(
    pred = predict(lin_mod, pctl_data),
    method = "percentile"
  )

plot_data <- bind_rows(grid_data, pctl_data)

library(ggplot2)

ggplot(plot_data, aes(x = disp, y = pred)) +
  geom_point(alpha = .5, cex = 1) +
  facet_wrap(~method)