Add margin row totals in dplyr chain

Question

I would like to add overall summary rows while also calculating summaries by group using dplyr. I have found various questions asking how to do this, e.g. here, here, and he

Answer 1

Here is my suggestion.

1. Find the combination of relevant grouping variables via the powerSet function.
2. split the data frame into a list, grouped by the powerSet of the grouping variables
3. summarise the data frame using an appropriate summary function (e.g. mean)
4. bind_rows the result - summaries are now NA because these columns are dropped in step 3
5. replace NA values of the grouping variables using appropriate names.

Note. If grouping variables are numeric, they will not be dropped in step 3 - I therefore mutate them to character variables.

powerSetList <- function(df, ...) {
  rje::powerSet(x = c(...))[-1] %>% lapply(function(x, tdf = df) group_by(tdf, .dots=x)) %>% c(list(tibble(df)), .)
} 

mtcars %>% 
  mutate_at(vars("cyl", "gear"), as.character) %>%
  powerSetList("cyl", "gear") %>%
  map(~summarise_if(., is.numeric, .funs = mean)) %>%
  bind_rows() %>%
  replace_na(list(gear = "all gears",
                  cyl = "all cyls"))