Error in parse_block(g[-1], g[1], params.src): duplicate label 'unnamed-chunk-1

Question

There is something strange with this error. Sometimes the same rmd file with r chunks runs without problems and knitr assigns sequential number after the 'unnamed-chunk-' but sometimes knitr reports error.

I can't find any way to identify under what circumstances the process runs smoothly and when it generates error. All r chunks within rmd file are without the name/label.

Below there is the log of command in case everything is going smoothly and the not labels chunks are given sequential number

outfile = c(outfile, knit("File1.rmd"))    
processing file: File1.Rmd

  |.....                                                            |   8%
  ordinary text without R code

  |...........                                                      |  17%
label: unnamed-chunk-1 (with options) 
List of 1
 $ echo: logi FALSE

  |................                                                 |  25%
   inline R code fragments

  |......................                                           |  33%
label: unnamed-chunk-2 (with options) 
List of 4
 $ echo      : logi FALSE
 $ connection: symbol db.conn
 $ output.var: chr "db.data"
 $ engine    : chr "sql"

  |...........................                                      |  42%
  ordinary text without R code

  |................................                                 |  50%
label: unnamed-chunk-3 (with options) 
List of 1
 $ echo: logi FALSE

  |......................................                           |  58%
  ordinary text without R code

  |...........................................                      |  67%
label: unnamed-chunk-4 (with options) 
List of 1
 $ echo: logi FALSE

  |.................................................                |  75%
  ordinary text without R code

  |......................................................           |  83%
label: unnamed-chunk-5 (with options) 
List of 1
 $ echo: logi FALSE

  |............................................................     |  92%
  ordinary text without R code

  |.................................................................| 100%
label: unnamed-chunk-6 (with options) 
List of 1
 $ echo: logi FALSE


output file: Test_DSN-ARAMIS-01.md

> outfile = c(outfile, knit_child("File2.Rmd"))


processing file: File2.Rmd
  |.............                                                                                                                                          |   8%
  ordinary text without R code

  |.........................                                                                                                                              |  17%
label: unnamed-chunk-1 (with options) 
List of 1
 $ echo: logi FALSE

  |......................................                                                                                                                 |  25%
   inline R code fragments

  |..................................................                                                                                                     |  33%
label: unnamed-chunk-2 (with options) 
List of 4
 $ echo      : logi FALSE
 $ connection: symbol db.conn
 $ output.var: chr "db.data"
 $ engine    : chr "sql"

  |...............................................................                                                                                        |  42%
  ordinary text without R code

  |............................................................................                                                                           |  50%
label: unnamed-chunk-3 (with options) 
List of 1
 $ echo: logi FALSE

  |........................................................................................                                                               |  58%
  ordinary text without R code

  |.....................................................................................................                                                  |  67%
label: unnamed-chunk-4 (with options) 
List of 1
 $ echo: logi FALSE

  |.................................................................................................................                                      |  75%
  ordinary text without R code

  |..............................................................................................................................                         |  83%
label: unnamed-chunk-5 (with options) 
List of 1
 $ echo: logi FALSE

  |..........................................................................................................................................             |  92%
  ordinary text without R code

  |.......................................................................................................................................................| 100%
label: unnamed-chunk-6 (with options) 
List of 1
 $ echo: logi FALSE

and the second execution of the same files gives the following output when processing second file

> outfile = c(outfile, knit_child("File2.Rmd"))


processing file: File2.Rmd
Error in parse_block(g[-1], g[1], params.src) : 
  duplicate label 'unnamed-chunk-1'

Solution 1

There are 2 situations:

1. You could have a duplicate chunk name
2. You could have copied and pasted the same code over. If so do the following:

options(knitr.duplicate.label = "allow")

https://bookdown.org/yihui/rmarkdown-cookbook/duplicate-label.html

Solution 2

It looks like there is an explanation and solution to this exact problem give here https://bookdown.org/yihui/rmarkdown-cookbook/duplicate-label.html

options(knitr.duplicate.label = "allow")

I've tried it on my own system, when faced with the same problem, and it was resolved.

Comments

[Prevost]

Can you add unique names to each chunk?

[Jan Pips]

Yes I can and that solves the error but not solves the issue. Adding unique labels to a chunk is workable for a dozen of files with say less than 10 chunks within each file. But doing that for 60 files is a pain in the ass unless there is the way to generate randomly the chunk labels which I am not aware of.

[Prevost]

Good point. I work in.rnw files. For printing auto-genreated xtables I create a list of print(xtable(x)) objects each with their own auto-genreated caption and run lapply(x, cat) or for (i in 1:length(x)){}. For multiple auto-generated figures, I also put those into a list and then use invisible(lapply(x,print))` to print all the figures. The caption is the same for all the figures. I differentiate between the figures by applying a descriptive legend title. You would need a chunk for each type of plot though.