I need to calculate the moving, weighted average by group in R.

Below is a sample of my code:

df %>% 
arrange(Type, Year)
group_by(Type) %>% 
mutate(
"weighted.rolling.mean" = rollapplyr(
    df %>% select(Value, Area),
    width = 50,
    function(z){
      return(
        weighted_mean = weighted.mean(z[,"Value"],z[,"Area"], partial = TRUE)
      )
    },
    by.column = FALSE,
    align = "right",
    fill=NA
  )
)

Without the group_by argument, the code works fine.

If I group_by inside the rollapplyr function, I get the error below:

Error in `mutate()`:
ℹ In argument: `weighted.rolling.mean = rollapplyr(...)`.
ℹ In group 1: `Type = "Etageadskillelse"`.
Caused by error in `x * w`:
! non-numeric argument to binary operator
Backtrace:
  1. ... %>% ...
  9. zoo::rollapplyr(...)
 11. zoo:::rollapply.default(..., align = align)
 14. zoo:::rollapply.zoo(zoo(data), ...)
 15. base::mapply(...)
 16. zoo (local) `<fn>`(dots[[1L]][[50L]], dots[[2L]][[50L]], data = `<chr[,4]>`)
 17. FUN(data[posns, ], ...)
 19. stats:::weighted.mean.default(z[, "Value"], z[, "Area"], partial = TRUE)

Many thanks in advance!