Using as.numeric, with functions and pipes in R

Question

I have a function that looks like this

 calc_df <- function(A_df, B_df){
     C_df <- filter(A_df, Type == "Animal") %>% 
         left_join(B_df)  %>%
         as.numeric(C$Count)

Where I cannot get the last lime to work, the first 3 work properly, but I would like the last line to take the column "Count" from the new df calculated in the function and make it numeric. (Right now it is a character vector)

** I have to do this at the end of the function because before the filter command, the Count column contains letters and cannot be made as.numeric

Answer 1

Looks like you're using dplyr, and that you want to change or add a column. This is what the dplyr::mutate function does.

Replace

as.numeric(C$Count)

with

mutate(Count = as.numeric(Count))

to replace the old, non-numeric Count column with the coerced-to-numeric replacement.

As to why your code didn't work, there are a few problems:

• dplyr is made for working with data frames, and the main dplyr functions (select, filter, mutate, summarize, group_by, *_join, ...) expect data frames as the first argument, and then return data frames. By piping the result of a left join into as.numeric, you are really calling as.numeric(unnamed_data_frame_from_your_join, C$Count), which clearly doesn't make much sense.

• You are trying to reference a data frame called C inside a definition for a data frame called C_df, which I think you mean to be the same thing. There's two issues here: (1) the mismatch between the names C and C_df, and (2) you can't reference C_df inside it's own definition.