Difference in PDFs when manually generated with Knit-Button vs rmarkdown::render() - e.g. section numbers not always included - r-markdown

I try to generate multiple reports (1 per group) with rmarkdown::render(), but it produces no section numbers.
I use the following file-structure for this in case you want to reproduce the sample (I tried to simplify each file and use only necessary code to reproduce the error):
*) groups.R: defines groups and subgroups in a nested list (all groups stored in a list and each group is a list itself)
groups <- list(
list(c("subgroup1","subgroup2"),"maingroup1"),
list(c("subgroup3","subgroup4"),"maingroup2")
)
*) main.Rmd: template for the reports
---
output:
pdf_document:
number_sections: true
classoption: a4paper
geometry: left=2cm,right=1cm,top=1.5cm,bottom=1cm,includeheadfoot
fontfamily: helvet
fontsize: 11pt
lang: en
header-includes:
- \usepackage{lastpage}
- \usepackage{fancyhdr}
- \pagestyle{fancy}
- \fancyhf{}
- \fancyhead[R]{\fontsize{9}{11} \selectfont \leftmark}
- \fancyhead[L]{\fontsize{9}{11} \selectfont Special report xxx}
- \fancyfoot[R]{\fontsize{9}{0} \selectfont Page \thepage\ of
\pageref{LastPage}}
---
```{r setup, include = FALSE}
knitr::opts_chunk$set(echo =
FALSE,comment=NA,warning=FALSE,message=FALSE)
```
\thispagestyle{empty}
\tableofcontents
\newpage
\setcounter{page}{1}
# Introduction
Some text...
```{r results="asis"}
source("graphics.R")
```
*) graphics.R: generates graphics for each subgroup (sections/section numbers are produced with cat() for each subgroup)
load("actgroup.RData")
source("template_graphics.R")
for (g in 1:length(act.group[[1]][[1]])) {
subgroup.name <- act.group[[1]][[1]][g]
cat("\\clearpage")
cat("\n# ",subgroup.name, "\n")
template_graphics(cars)
cat("\n\n")
cat("\\clearpage")
template_graphics(iris)
cat("\n\n")
cat("\\clearpage")
template_graphics(airquality)
cat("\n\n")
cat("\\clearpage")
cat("\n")
}
*) template_graphics.R: template for plotting
template_graphics <- function(data) {
plot(data)
}
*) loop.R: used for generating all reports as PDF - 1 per group
setwd("YOUR DIRECTORY HERE")
library(rmarkdown)
source("groups.R")
for(i in 1:length(groups)) {
act.group = list(groups[[i]])
save(act.group,file="actgroup.RData")
rmarkdown::render("main.Rmd",
output_format=pdf_document(),
output_file=paste0("Special Report ",act.group[[1]][[2]],".pdf"),
output_dir="~/Reports")
}
The problem is, that the final documents do not show the section numbers. When I knit manually in main.Rmd (pressing knit-Button), the section numbers are printed.
Version rmarkdown::render
Version knit-Button
I thought that pressing the knit-Button also starts the rendering-process with rmarkdown::render()? So it's surprising that the reports are not identical?
In advance I installed tinytex::install_tinytex(). The used latex-packages in main.Rmd were automatically installed during the first time rendering.
I am not sure what the problem is. I use R 4.1.0 and RStudio 2022.02.2.
Thanks for your help!!

The behaviour of pdf_document() as the output-format in rmarkdown::render caused the missing section-numbers.
In my YAML-header in main.Rmd I chose to keep the section numbers with number_sections: true. If this should also be rendered when using rmarkdown::render, it has to be an argument in the function:
pdf_document(number_sections=TRUE)
The code of loop.R produces now pdfs with section numbers:
library(rmarkdown)
source("groups.R")
for(i in 1:length(groups)) {
act.group = list(groups[[i]])
save(act.group,file="actgroup.RData")
rmarkdown::render("main.Rmd",
output_format=pdf_document(number_sections=TRUE),
output_file=paste0("Special Report ",act.group[[1]][[2]],".pdf"),
output_dir="~/Reports")
}
More information on pdf_document() can be found here:
https://pkgs.rstudio.com/rmarkdown/reference/pdf_document.html
Alternatively, fill in just a text-reference as output-format: output_format="pdf_document". When set like this, the options of the YAML-Header are not overwritten and the numbers are also included.

Related

Change default behavior of callout blocks in Quarto

In Quarto, I'd like to change the default behavior of a single callout block type so that it will
always automatically have the same caption (e.g. "Additional Resources")
always be folded (collapse="true")
Let's say I want this for the tip callout block type while the others (note, warning, caution, and important) should not be affected.
In other words, I want the behavior/output of this:
:::{.callout-tip collapse="true"}
## Additional Resources
- Resource 1
- Resource 2
:::
by only having to write this:
:::{.callout-tip}
- Resource 1
- Resource 2
:::
Update:
I have actually converted the following lua filter into a quarto filter extension collapse-callout, which allows specifying default options for specific callout blocks more easily. See the github readme for detailed instructions on installation and usage.
As #stefan mentioned, you can use pandoc Lua filter to do this more neatly.
quarto_doc.qmd
---
title: "Callout Tip"
format: html
filters:
- custom-callout.lua
---
## Resources
:::{.custom-callout-tip}
- Resource 1
- Resource 2
:::
## More Resources
:::{.custom-callout-tip}
- Resource 3
- Resource 4
:::
custom-callout.lua
local h2 = pandoc.Header(2, "Additional Resources")
function Div(el)
if quarto.doc.isFormat("html") then
if el.classes:includes('custom-callout-tip') then
local content = el.content
table.insert(content, 1, h2)
return pandoc.Div(
content,
{class="callout-tip", collapse='true'}
)
end
end
end
Just make sure that quarto_doc.qmd and custom-callout.lua files are in the same directory (i.e. folder).
After a look at the docs and based on my experience with customizing Rmarkdown I would guess that this requires to create a custom template and/or the use of pandoc Lua filters.
A more lightweight approach I used in the past would be to use a small custom function to add the code for your custom callout block to your Rmd or Qmd. One drawback is that this requires a code chunk. However, to make your life a bit easier you could e.g. create a RStudio snippet to add a code chunk template to your document.
---
title: "Custom Callout"
format: html
---
```{r}
my_call_out <- function(...) {
cat(":::{.callout-tip collapse='true'}\n")
cat("## Additional Resources\n")
cat(paste0("- ", ..., collapse = "\n\n"))
cat("\n:::\n")
}
```
```{r results="asis"}
my_call_out(paste("Resource", 1:2))
```
Blah blah
```{r results="asis"}
my_call_out("Resource 3", "Resource 4")
```
Blah blah

Citations not generating in RMarkdown

new to Stackoverflow, and relatively new to writing documents in RMarkdown. I have most other things down, however, I cannot generate citations and the bibliography in RMarkdown. For the example code below, I have used the \cite{REFNAME} as well as [#REFNAME] syntax. The [#REFNAME] will not even let the document be generated, the \cite{} generates the document but the citation is [?] and the reference is not after the #References section.
I have looked through this site and I cannot get suggested fixes to work. I am hoping that the error I am showing after the code sections will tell someone what the error really is.
---
title: "Title for my document"
date: "`r format(Sys.Date(), '%d %B, %Y')`"
output:
pdf_document:
fig_caption: yes
keep_tex: true
number_sections: true
geometry: margin=0.5in
header-includes: \usepackage{float}
urlcolor: blue
bibliography: bibliography_file.bib
---
```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = FALSE)
knitr::knit_hooks$set(mysize = function(before, options, envir) {
if (before)
return(options$size)
})
```
text written \cite{neil_increasing_2008} more text written
# References
My .bib file entry looks like this
#article{neil_increasing_2008,
title = {Increasing incidence of legionellosis in the {United} {States},
1990-2005: changing epidemiologic trends.},
volume = {47},
journal = {Clinical Infectious Diseases},
author = {Neil, K. and Berkelman, R.},
year = {2008},
pages = {591--599}
}
The error is:
Warning message: LaTeX Warning(s): Citation `neil_increasing_2008'
on page 1 undefined on input lin There were undefined references.
If I use the [#neil_increasing_2008] as opposed to the \cite{} then the following error is generate, and it does not generate the document either
pandoc-citeproc: Cannot decode byte '\x87':
Data.Text.Internal.Encoding.Fusion.streamUtf8: Invalid UTF-8 stream
pandoc.exe: Error running filter pandoc-citeproc Filter returned error
status 1 Error: pandoc document conversion failed with error 83
Execution halted

RMarkdown not rendering Stan code chunk as code

When I knit the following RMarkdown document:
title: "Reprex"
author: "Jeremy Colman"
date: "17/07/2018"
output:
html_document: default
pdf_document: default
word_document: default
---
```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)
```
```{stan, output.var='priors', eval = FALSE, tidy = FALSE}
parameters {
real<lower = 0> qtilde1;
}
model {
qtilde1 ~ gamma(38.9, 0.67);
}
```
The Stan code chunk is rendered as ordinary text, including the three reverse single quotes and word stan from the chunk header. I cannot show that in this post because stackoverflow tells me, correctly but unhelpfully, "Your post appears to contain code that is not properly formatted as code". That sums up my problem!
Code chunks in R are rendered correctly.
Your
```{stan output.var='priors', eval = FALSE, tidy = FALSE}
needs to be flush-left, but you have a leading space before the ```

How to parameterize an rmarkdown in blogdown?

Thanks to parameterized reports support by Rmarkdown, it is easy to make an .Rmd file rendered with parameters as follows (this example is from here):
---
title: "`r params$dynamictitle`"
params:
dynamictitle: My report
reportdate: !r Sys.Date()
date: "`r params$reportdate`"
---
This would be fine to knit to a .html file, but cannot be rendered in blogdown, with the following error:
Quitting from lines 2-5 ()
<simpleError in eval(parse_only(code), envir = envir): object 'params' not found>
I want to know whether there is a workaround way to make an .Rmd file parameterized?

break a slide of reveal.js (xaringan) when it's long

I'm actually using xaringan, but it uses reveal.js, so it should be the same.
I have a slide which prints bibliography using RefManageR, and I'd like to use as many slides as needed:
---
```
{r results = "asis", echo = FALSE}
PrintBibliography(bib, .opts = list(check.entries = FALSE, sorting = "ynt"))
```
---
I guess I'm looking for some type of allowframebreaks, but I couldn't manage to find one.
Xaringan uses remark.js, not reveal.