How to get the object name as a string inside a function in r

I want to change column names of a data frame with a function.

To overwrite my data frame with the new column names, I used assign(), which first argument has to be the name of the same data frame as a string. To get the name as a string, I used deparse(substitute(x)), which worked outside the function. But inside the function, it returns the content of my data frame as a string instead of the name itself...


df <- data.frame(
  emp_id = c (1:5), 
  emp_name = c("Rick","Dan","Michelle","Ryan","Gary"),
  stringsAsFactors = FALSE
)

deparse(substitute(df))

rename_fun <- function(x) {
  colnames(x)[1] <- "___0"
  colnames(x)[2] <- "___1"

  y <- deparse(substitute(x))
  
    assign(y, x, envir = .GlobalEnv)      
}

rename_fun(df)

I also tried

as.character(substitute(x))

but the same problem...

Solution

An alternative is to use as.character(match.call()$x), which can be used at any point in the function:

rename_fun <- function(x) {
  colnames(x)[1] <- "___0"
  colnames(x)[2] <- "___1"
  assign(as.character(match.call()$x), x, envir = .GlobalEnv)      
}

Giving

rename_fun(df)

df
#>   ___0     ___1
#> 1    1     Rick
#> 2    2      Dan
#> 3    3 Michelle
#> 4    4     Ryan
#> 5    5     Gary

Note though that functions which write objects into the global environment as a side effect are not recommended, even if they are overwriting an existing object. A function should return the altered data frame, which the user may then choose to use to overwrite the object.

A better way to write the function would be:

rename_fun <- function(x) {
  
  colnames(x)[1] <- "___0"
  colnames(x)[2] <- "___1"
  x
}

Which would be called like this:

df <- rename_fun(df)

And give the same result, while leaving the option of having a copy of the original data frame if the caller wants.

^{Created on 2022-11-18 with reprex v2.0.2}