My df looks like this:

N = 1000

a <- rnorm(N)
b <- rnorm(N)
c <- rnorm(N)

df <- data.frame(a, b, c)

For each of these variables, I would like to perform the following function:

ifelse(df$`i` < 10, paste0("0", df$`i`), paste0(df$`i`))

where i is a , b and c.

Is there a way I can do this in a for-loop? Thanks :)

CodePudding user response：

Do you have to use a loop? dplyr's across lets you iterate a function over many columns:

library(dplyr)

df |> mutate(across(everything(),
             ~ifelse(.x < 10,
                     paste0("0", .x),
                     .x)))

CodePudding user response：

apply(df, 2, \(i) ifelse(i < 10, paste0("0", i), i))

Or for this case, more simply:

apply(df, 2, formatC, width = 2, flag = "0")

CodePudding user response：

If you're hoping to keep them in the frame, then the use of apply comes with a little risk: if there are any character columns existing in it, then when apply converts to a matrix before processing, you will have lost the number-ness of your other columns. (I don't know if your intended use is any more complicated than your sample data.)

Ways to mitigate this:

1. Use apply only on specific columns:

   df[,1:3] <- apply(df[,1:3], 2, sprintf, fmt="f")
   

2. Use lapply:

   df[1:3] <- lapply(df[1:3], sprintf, fmt="f")
   

   Or of you're operating on the whole frame,

   df[] <- lapply(df, sprintf, fmt="f")
   

   The df[] <- is necessary because lapply returns a list, losing the data.frame class. By doing df[] <- instead, you're reassigning the contents without changing the class, so it retains its data.frame class.