R: Is there a good replacement for plyr::rbind.fill in dplyr?

Question

for tidyverse users, dplyr is the new way to work with data.

For users trying to avoid older package plyr, what is the equivalent function to rbind.fill in dplyr?

Prelude

Answer 1

Yes. dplyr::bind_rows

Credit goes to commenter.

Prelude

Answer 2

An alternative:

rbindf <- function(...) {
  
  l <- list(...)
  if(length(l) == 1) l <- l[[1]]
  nn <- length(l)

  x <- l[[1]]
  if(length(l)>1){
      for(i in 2:nn) {
        y <- l[[i]]
        if(nrow(x) > 0 & nrow(y) > 0) {
          if(!all(yinx <- names(y) %in% names(x))) {
            x[, names(y)[!yinx]] <- NA
          } 
          if(!all(xiny <- names(x) %in% names(y))) {
             y[, names(x)[!xiny]] <- NA
          } 
        }
        x <- rbind(x, y)
      }
  }
  return(x)
}

From https://github.com/sashahafner/biogas (originally from https://github.com/jonkatz2/monitoR).

Curiosity

Answer 3

Same Thing with R Base:

df2 <- data.frame(a = c(1:5), b = c(6:10))
df3 <- data.frame(a = c(11:15), b = c(16:20), c = LETTERS[1:5])
df2[setdiff(names(df3), names(df2))] <- NA
df3[setdiff(names(df2), names(df3))] <- NA

rbind(df2, df3)

Matronna