Why I don’t use R Markdown’s ref.label
Want to share your content on R-bloggers? click here if you have a blog, or here if you don't.
Motivation
Consider you have defined variable x
,
x = 1
and define another chunk, where you simply add one up
```{r addOne} sum = x + 1 sum ```
resulting in
[1] 2
To reuse this chunk, an empty code block is created referencing the above chunk
```{r, ref.label = 'addOne'} ```
again resulting in
sum = x + 1 sum [1] 2
Behind the scenes, the chunk basically was copy-pasted and then executed again. One problem is that one can easily lose track of the scope of the variables used in that chunk. For example, let’s assume you use the sum
variable further below in your document to store some other result:
sum = 10
If you now again reuse the above chunk
```{r, ref.label = 'addOne'} ``` sum = x + 1 sum [1] 2
sum
has been overwritten by the chunk:
print(sum) # expect sum == 10 [1] 2
Since the ref.label
chunk is empty, this issue might not be easily spotted.
Another inconvenience arrises with RStudio’s notebook functionality to execute individual code chunks. While the original chunk can be executed, none of the empty ref.label
chunks can. Funnily enough, this inconvenience was what made me think about an alternative solution.
Alternative solution
Luckily, the solution is quite simple – put your entire chunk inside a function and then “reference” the function:
add1 <- function(x) { sum = x + 1 sum } add1(x) [1] 2
Now both the sum
variable is perfectly scoped and the “referenced” call can be executed in the RStudio notebook as usual. Plus, of course, this “chunk” could be easily parametrized:
addY <- function(x, y) { sum = x + y sum } addY(x, y = 1) [1] 2
Summary
Downsides of using ref.label
:
- potential issues with (global) variables as chunk does not provide local scoping
ref.label
chunks are empty and therefore cannot be executed in RStudio notebooks
Proposed solution: encapsulate entire chunk inside a function and then execute the function wherever you would reference the chunk.
R-bloggers.com offers daily e-mail updates about R news and tutorials about learning R and many other topics. Click here if you're looking to post or find an R/data-science job.
Want to share your content on R-bloggers? click here if you have a blog, or here if you don't.