How to pass the name of the dataframe to write_csv in dplyr

Question

I would like to write a dataframe with the name of the dataframe passed to the write_csv function. For example, I would normally do the following

iris %>% write_csv("iris.csv")

but I would like to not have to write the name of the file. I tried something like the following

iris %>% write_csv(paste0(., ".csv"))

but I think this will require something more like a quosure

The result of that code is `..csv` not `iris.csv` – Alex Apr 06 '18 at 20:02 — Alex, Apr 06 '18 at 20:02

Martin Schmelzer · Accepted Answer · 2018-04-07T09:37:43.310

3

Just define your own custom write.csv3:

write.csv3 <- function(x, ...) {
  write.csv(x, file = paste0(deparse(substitute(x)), '.csv'), ...)
}
write.csv3(iris)

If you want to use pipes then you could set an attribute name first:

write.csv3 <- function(x, ...) {
  write.csv(x, file = paste0(attr(x, "name"), '.csv'), ...)
}

attr(iris, "name") <- "iris" 
iris %>% write.csv3

Or you try to apply this hacky approach.

edited Apr 07 '18 at 09:37

answered Apr 06 '18 at 20:11

Martin Schmelzer

This does work, but I have not been able to pipe to `write.csv3` – Alex Apr 06 '18 at 23:46
Ok I see. Updated my answer. But there is no *easy* way to fix this. – Martin Schmelzer Apr 07 '18 at 09:40

score 0 · Answer 2 · answered Apr 06 '18 at 20:21

0

The basic use case of function parameter in a dplyr pipe is to treat function parameter as a name something like this:

csv_save <- function(xx){
  csv_name = as.name(xx)
  iris %>% write.csv(.,file = !!csv_name)
}

csv_save("files.csv")

(not tested)

answered Apr 06 '18 at 20:21

Aleksandr

2 Answers2