CODEWITHSUNDEEP
rknitrr-markdown
antonio
antonio

Reputation: 11120

Knitr: Do not purl a chunk

Consider the following test.Rmd:

```{r setup, purl=FALSE}
opts_chunk$set(purl=FALSE)
opts_template$set(nopurl = list(purl=FALSE))
```     

```{r test1}
print(1)
```

```{r test2, opts.label='nopurl'}
print(2)
```

```{r test3, purl=FALSE}
print(3)
```

purl('test.Rmd') gives test.R, where none of the test* chunks is supposed to be purled, but:

## ----test1---------------------------------------------------------------
print(1)


## ----test2, opts.label='nopurl'------------------------------------------
print(2)

Only test3 is not purled, the rest are purled, despite the global option opts_chunk$set(purl=FALSE) and the label nopurl.
Why?

Upvotes: 5

Views: 2185

Answers (2)

antonio
antonio

Reputation: 11120

Based on Yihui feedback, the way to go is like follows:

```{r setup, purl=FALSE}
knit_hooks$set(purl = hook_purl)
opts_template$set(nopurl = list(purl=FALSE))
opts_template$set(dopurl = list(purl=TRUE))
```

```{r test}
print(1)
```

```{r test2, opts.label='nopurl'}
print(2)
```

```{r test3, opts.label='dopurl'}
print(3)
```

With this approach, you don't need to:

purl('test.Rmd')

You simply: 

knit('test.Rmd')

and get both the usual test.md and the test.R file. The latter is as follows:

## ----test----------------------------------------------------------------
print(1)

## ----test3, opts.label='dopurl'------------------------------------------
print(3)

As you see, given knit_hooks$set(purl = hook_purl), the default behaviour for chunks is purling.

In actual use nopurl and dopurl will collect more options to be set together with (no-)purling.
Anyway, when considering the following note:

N.B. There is no guarantee that the R script generated by 'purl()' can reproduce the computation done in 'knit()'. The 'knit()' process can be fairly complicated (special values for chunk options, custom chunk hooks, computing engines besides R, and the 'envir' argument, etc). If you want to reproduce the computation in a report generated by 'knit()', be sure to use 'knit()', instead of merely executing the R script generated by 'purl()'. This seems to be obvious, but some people just do not get it.

it seems wise to always use knit_hooks$set(purl = hook_purl) and set the purl=FALSE when purl is not needed.

Upvotes: 2

Yihui Xie
Yihui Xie

Reputation: 30114

It is not a bug, but just that purl() does not executed any code chunks, so the first code chunk was not really executed. The purl() function is not reliable in many ways, and see ?knitr::hook_purl instead. That said, I'd not recommend purling a document at all; knitting is more reliable (See the Note section in ?knitr::purl).

Upvotes: 2

Related Questions