Compute correlations with a target variable — compute

compute_correlation() allows to compute correlation of several numerical predictors with respect to a target variable. Optionally, the computation can be done by groups.

Usage

compute_correlation(data, target_var, by = NULL, ...)

Arguments

data: A tibble.
target_var: The variable against which the correlations needs to be computed. For instance, dx_gap.
by: A character vector.
...: Optional arguments passed to correlate()

Value

A tibble.

Examples

if (FALSE) {
tbl <-
  build_tbl(
    "tb",
    year = NULL,
    estimated = "who_estimates.e_inc_num",
    notified = "who_notifications.c_newinc",
    vars = extract_vars("tb")
  ) |>
  dplyr::mutate(is_hbc = forcats::as_factor(is_hbc))

tbl |>
  dplyr::filter(year == 2019) |>
  compute_correlation(dx_gap, by = "year")

tbl |>
  compute_correlation(dx_gap, by = c("year", "is_hbc"))

car_tbl <- tibble::as_tibble(mtcars, rownames = "car_name")
my_cars <- dplyr::select(car_tbl, -dplyr::all_of(c("vs", "am", "carb")))
compute_correlation(my_cars, mpg, by = c("gear", "cyl"))
compute_correlation(my_cars, mpg, by = NULL) # car_name is dropped by default
}