CODEWITHSUNDEEP
rfunctionreplace
MinzKi
MinzKi

Reputation: 11

How to exchange parts of a variable name in a function in R?

I try to run a chunk of code multiple times (t1-t13). Since all the variable names start with the name of the concerning timepoint, I tried to write a function in which I wanted to simply replace the “t1” in the variable name, e.g., t1_ls_3. To do so, I used the paste function to convert the following chunk of code into a function, where only the timepoint should be replaced. A short example:

I need a way to change t1 in this line: data$t1_ls_3 <- 11

I tried this out and it worked:

data[paste(“t1”, "ls_3", sep = "_")] <- 11
data <- data[paste(“t1”, "ls_3", sep = "_")]

Then I tried to write it as a function:

func_replace <- function(x = “time”) {
  data[paste(“time”, "ls_3", sep = "_")] <- 11
  data <- data[paste(“time”, "ls_3", sep = "_")]
 
}
func_replace (x = “t1“)

However, all R does when I run the function is to put a new variable into the data frame, time_ls_3 and does not change the actual variable t1_ls_3. Do you have any idea what the problem is or how I can exchange parts of a variable name in a function? I know that I could work around the problem by using a long-format data set. But I would like to know how to exchange parts of a variable in a function in general.

Upvotes: 1

Views: 333

Answers (1)

Ian Campbell
Ian Campbell

Reputation: 24838

Your fundamental misunderstanding seems to be about character strings and symbols.

You define a character string in R with " or '. Thus the following:

"time"

is a character string with the word time in it.

In contrast, symbols are names that can refer to objects in R. You can create a symbol (pointing to an object) holding a time called time like this:

time <- "03:00"

When you type time at the console, you get back the character string:

time
#"03:00"

Therefore, when you try to use paste with "time", you're going to get the same result no matter what:

paste("time", "ls_3", sep = "_")
#[1] "time_ls_3"

Instead, you seem to want to use x which is the symbol your function is assigning "t1" to.

Perhaps your data looks something like this:

data <- setNames(data.frame(t(rep(0,5))),sapply(1:5,\(x)paste0("t",x,"_ls_3")))
data
#  t1_ls_3 t2_ls_3 t3_ls_3 t4_ls_3 t5_ls_3
#1       0       0       0       0       0

The last problem is that assignments inside functions only work in the function environment, so you need to use the <<- operator.

func_replace <- function(x = NULL) {
  data[paste(x, "ls_3", sep = "_")] <<- 11
  }

func_replace(x = “t1“)
data
#  t1_ls_3 t2_ls_3 t3_ls_3 t4_ls_3 t5_ls_3
#1      11       0       0       0       0

Upvotes: 2

Related Questions