I want to use the collapse package to summarise the following data:

library(collapse)
df <- data.frame(id = c(1, 1, 1,
                        2, 2, 2),
                 var = c(NA, "a", "b",
                         "e", "f", "g"),
                 var1 = c("q", "w", "r",
                          "l", "k", NA),
                 other = c(4, 5, 6, 7, 8, 9))
df
#   id  var var1 other
# 1  1 <NA>    q     4
# 2  1    a    w     5
# 3  1    b    r     6
# 4  2    e    l     7
# 5  2    f    k     8
# 6  2    g <NA>     9

into the following where I simply get the first row of var, the last row of var1 (excluding NAs) and the mean of other, by id:

#   id  var var1 other
# 1  1 <NA>    r     5
# 2  2    e    k     8

The default behaviour of na.rm for ffirst and flast is to ignore NAs:

logical. TRUE skips missing values and returns the first / last
non-missing value i.e. if the first (1) / last (n) value is NA, take
the second (2) / second-to-last (n-1) value etc..

However, I can only have na.rm = TRUE and na.rm = FALSE set for ffirst and flast at the same time (not separately for each):

# na.rm = TRUE (default)
collap(df, ~id,
       custom = list(ffirst = "var", flast = "var1", fmean = "other"))
#   id var var1 other
# 1  1   a    r     5
# 2  2   e    k     8

# na.rm = FALSE
collap(df, ~id,
       custom = list(ffirst = "var", flast = "var1", fmean = "other"),
       na.rm = FALSE)
#   id  var var1 other
# 1  1 <NA>    r     5
# 2  2    e <NA>     8

Is there a way to include na.rm = FALSE for flast only within the collap call?

thanks