CODEWITHSUNDEEP
rr-s3
MayaGans
MayaGans

Reputation: 1845

Function for columns to inherit custom class of data frame in R

I have a list of dataframes 

dd <- list()
dd$dat <- list(
  one = data.frame(a = c(1), b = c(2)),
  two = data.frame(c = c(3), d = c(4)),
  three = data.frame(e = c(5), f = c(6))
)

and wrote a function to append a custom class to each dataframe:

# append classes 
append_classes <- function(x, nm) {
  class(x) <- 
    case_when(
      nm == "one" ~ c(class(x), "foo"),
      nm == "two" ~ c(class(x), "bar"),
      TRUE ~ c(class(x), "custom")
    )
  return(x)
}

dd$dat <- imap(dd$dat, append_classes)

class(dd$dat[[1]])

It works!

[1] "data.frame" "foo"

But now I want to use class inheritance to have the columns in each data frame inherit the foo, bar, and custom classes respectively - how would I do this?

Desired Output

class(dd$dat$one$a)

[1] "numeric" "foo"

class(dd$dat$two$d)

[1] "numeric" "bar"

I'm very new to using S3 so any help is appreciated!!

Upvotes: 1

Views: 168

Answers (1)

akrun
akrun

Reputation: 887028

We can use the imap recursively or use map inside 

library(purrr)
dd$dat <-  imap(dd$dat, ~ {nm1 <- .y
       map_dfr(append_classes(.x, nm1), ~ append_classes(.x, nm1))
       })

class(dd$dat$one$a)
#[1] "numeric" "foo"    
class(dd$dat$two$d)
#[1] "numeric" "bar"

Or this can be done with base R using Map/lapply 

dd$dat <- Map(function(x, y) {
     tmp <- append_classes(x, y)
    tmp[] <- lapply(tmp, append_classes, nm = y)
    tmp} , dd$dat, names(dd$dat))
class(dd$dat$one$a)
#[1] "numeric" "foo"

Upvotes: 1

Related Questions