How can I get R dplyr to report the number of rows that meet each case_when condition?

Question

I'm often using dplyr case_when with several logical conditions. Because case_when evaluates sequentially, sometimes an earlier case applies and then is "taken off" the list even further conditions may also apply. Sometimes I make mistakes and think a case is being activated but it isn't. Has anyone created some code that works with the case_when syntax that shows how many or which rows end up in each case of case_when? Here's an example of case_when that would benefit from this.

mtcars |>
    mutate(
        n_gears = case_when(
            gear == 5 ~ "perfect",
            gear == 4 ~ "bad",
            gear >= 3 ~ "perfect"
        )
)

Answer 1

I don't know of a baked-in function that will do what you want, but you can try this hacked up (under-tested) verbose version of case_when():

case_when_verbose <- function(..., .default = NULL, .envir = parent.frame()) {
  orig_args <- setdiff(names(formals(dplyr::case_when)), c("...", ".default"))
  dots <- list(...)
  if (!is.null(names(dots))) dots <- dots[ !names(dots) %in% orig_args ]
  nms <- sapply(dots, function(dot) deparse(as.list(dot)[[2]]))
  res <- lapply(dots, function(dot) eval(as.list(dot)[[2]], envir = .envir))
  if (!is.null(.default)) {
    nms <- c(nms, ".default")
    res <- c(res, list(!Reduce(`|`, res)))
  }
  res_eff <- c(res[1], lapply(seq_along(res)[-1], function(i) !Reduce(`|`, res[1:(i-1)]) & res[[i]]))
  eff <- tibble(condition = nms, matched = sapply(res, sum, na.rm = TRUE),
                effective = sapply(res_eff, sum, na.rm = TRUE))
  cat("## case_when_verbose():\n", paste("##", format(eff), collapse = "\n"), "\n", sep = "")
  cl <- match.call()
  cl[[1]] <- substitute(dplyr::case_when)
  eval(cl, envir = .envir)
}

mtcars |>
  mutate(
    n_gears = case_when_verbose(
      gear == 5 ~ "perfect",
      gear == 4 ~ "bad",
      gear >= 3 ~ "perfect",
      .default = "unk"
    )
  )
# ## case_when_verbose():
# ## # A tibble: 4 × 3
# ##   condition matched effective
# ##   <chr>       <int>     <int>
# ## 1 gear == 5       5         5
# ## 2 gear == 4      12        12
# ## 3 gear >= 3      32        15
# ## 4 .default        0         0
#                      mpg cyl  disp  hp drat    wt  qsec vs am gear carb n_gears
# Mazda RX4           21.0   6 160.0 110 3.90 2.620 16.46  0  1    4    4     bad
# Mazda RX4 Wag       21.0   6 160.0 110 3.90 2.875 17.02  0  1    4    4     bad
# Datsun 710          22.8   4 108.0  93 3.85 2.320 18.61  1  1    4    1     bad
# Hornet 4 Drive      21.4   6 258.0 110 3.08 3.215 19.44  1  0    3    1 perfect
# Hornet Sportabout   18.7   8 360.0 175 3.15 3.440 17.02  0  0    3    2 perfect
# Valiant             18.1   6 225.0 105 2.76 3.460 20.22  1  0    3    1 perfect
# Duster 360          14.3   8 360.0 245 3.21 3.570 15.84  0  0    3    4 perfect
# Merc 240D           24.4   4 146.7  62 3.69 3.190 20.00  1  0    4    2     bad
# Merc 230            22.8   4 140.8  95 3.92 3.150 22.90  1  0    4    2     bad
# Merc 280            19.2   6 167.6 123 3.92 3.440 18.30  1  0    4    4     bad
# Merc 280C           17.8   6 167.6 123 3.92 3.440 18.90  1  0    4    4     bad
# Merc 450SE          16.4   8 275.8 180 3.07 4.070 17.40  0  0    3    3 perfect
# Merc 450SL          17.3   8 275.8 180 3.07 3.730 17.60  0  0    3    3 perfect
# Merc 450SLC         15.2   8 275.8 180 3.07 3.780 18.00  0  0    3    3 perfect
# Cadillac Fleetwood  10.4   8 472.0 205 2.93 5.250 17.98  0  0    3    4 perfect
# Lincoln Continental 10.4   8 460.0 215 3.00 5.424 17.82  0  0    3    4 perfect
# Chrysler Imperial   14.7   8 440.0 230 3.23 5.345 17.42  0  0    3    4 perfect
# Fiat 128            32.4   4  78.7  66 4.08 2.200 19.47  1  1    4    1     bad
# Honda Civic         30.4   4  75.7  52 4.93 1.615 18.52  1  1    4    2     bad
# Toyota Corolla      33.9   4  71.1  65 4.22 1.835 19.90  1  1    4    1     bad
# Toyota Corona       21.5   4 120.1  97 3.70 2.465 20.01  1  0    3    1 perfect
# Dodge Challenger    15.5   8 318.0 150 2.76 3.520 16.87  0  0    3    2 perfect
# AMC Javelin         15.2   8 304.0 150 3.15 3.435 17.30  0  0    3    2 perfect
# Camaro Z28          13.3   8 350.0 245 3.73 3.840 15.41  0  0    3    4 perfect
# Pontiac Firebird    19.2   8 400.0 175 3.08 3.845 17.05  0  0    3    2 perfect
# Fiat X1-9           27.3   4  79.0  66 4.08 1.935 18.90  1  1    4    1     bad
# Porsche 914-2       26.0   4 120.3  91 4.43 2.140 16.70  0  1    5    2 perfect
# Lotus Europa        30.4   4  95.1 113 3.77 1.513 16.90  1  1    5    2 perfect
# Ford Pantera L      15.8   8 351.0 264 4.22 3.170 14.50  0  1    5    4 perfect
# Ferrari Dino        19.7   6 145.0 175 3.62 2.770 15.50  0  1    5    6 perfect
# Maserati Bora       15.0   8 301.0 335 3.54 3.570 14.60  0  1    5    8 perfect
# Volvo 142E          21.4   4 121.0 109 4.11 2.780 18.60  1  1    4    2     bad

The aesthetics can be cleaned up, but the interim action is to print to the console a quick summary of the conditions, the number of rows that matched, and the number of rows that matched and did not match any of the previous conditions.

This summary frame is just a side-effect printed to the console, it is not stored anywhere. Because the arguments are all passed explicitly to dplyr::case_when(...), this function can be used (afaict) anywhere case_when can during development (I recommend against deploying this to production :-). And because the return value is actually from case_when(), the summary frame is not accessible later in the pipe.

Edit: I added explicit .default= and other case_when() arguments to break it out if used, the output will include an extra row indicating how often it was matched/used (and will not include that row if missing). I also actively guard against other args to case_when(.ptype=, .size=) so that they aren't inadvertently captured. Thanks to @SamR for the suggestion.