I'm trying to compile an article using BiocWorkflowTools and I'm pretty close to tearing all my hair out. I've been growing my hair for a few years now, so I would appreciate any help in delaying that process.

The suggestion on the "workflow workflow", for us wretched creatures who don't use Rstudio, is to make an article like so:

rmd_file <- rmarkdown::draft(
  "Article.Rmd",
  template = "f1000_article",
  package = "BiocWorkflowTools",
  edit = FALSE
)

If I then start an R session in the created Article directory and run rmarkdown::render("Article.Rmd") I get a latex error because f1000_styles.sty isn't included:

! LaTeX Error: Environment tabledata undefined.

Okay, so I can add this to the yaml frontmatter:

---
[...]
header-includes:
   - \usepackage{f1000_styles}
---

Now I wonder what happens if I render again?

! LaTeX Error: Option clash for package geometry.

Fantastic. Seems to be a check here in BiocWorkflowTools, because rmarkdown does something funny with geometry. No worries, add it to the yaml frontmatter.

---
[...]
geometry: "hmargin=1.8cm,vmargin=2.2cm"
---

What now? Oh boy. Apparently there's some nonsense with titlesec and pandoc too (here).

! Argument of \paragraph has an extra }.
<inserted text> 
                \par 
l.1290 \ttl@extract\paragraph

Fine. Stress levels rising even recreating this debugging process, but it's fine, latex alone is annoying and there are multiple other programs/languages interacting here so it's inevitable. Throw another ~~stick on the fire~~ entry in the yaml frontmatter:

---
[...]
subparagraph: yes
---

Now?

! LaTeX Error: \begin{minipage} on input line 110 ended by \end{center}.

This is where the trail of tears ends, I'm afraid. I have no debugging stamina remaining. I have absolutely no idea what the problem is here and I can't be arsed to debug the latex at the minute (I will probably try again tomorrow). One hypothesis is that it's the author field being scrambled. For example, if I look at the tex document I see

\title{\emph{F1000Research} Software Tool Article Template}
\author{true \and true}

An interesting interpretation, to be sure, but not the one that I initially intended, nor one that I particularly expected.

If I run rmarkdown::render([...], clean = FALSE) and inspect Article.knit.md, I see the same author field as in the template:

---
[...]
author: 
  - name: Author Name1
    affiliation: Address of author1
  - name: Author Name2
    affiliation: Address of author2
---

So it seems to me that pandoc is messing with me.

If I change to author: [A, B] I see \author{A \and B} which is obviously a slight improvement over random coercion to booleans, but obviously lacks affiliations and other niceties that journals sometimes expect. Even author: [A, B] doesn't save me from the proverbial river of ~~sh~~ excrement that is ! LaTeX Error: \begin{minipage} on input line 111 ended by \end{center}..

If you've made it this far I commend you, weary traveller. Rest assured my tone is one of general frustration, not annoyance with the authors. I hope that my ~~irritating~~ attempts at levity are seen as such, as I am simply trying to laugh as ~~Rome~~ my manuscript burns. I'm happy to make a PR to patch whatever problems I find.

Per the formalities, I'll try to list my software versions. I'm using the latest (1.12.0) BiocWorkflowTools

pandoc -v gives:

pandoc 1.19.2.4
Compiled with pandoc-types 1.17.0.5, texmath 0.9.4.4, skylighting 0.3.3.1

While pandoc-citeproc -V gives pandoc-citeproc 0.10.5.1.

My sessionInfo():

R version 3.6.2 (2019-12-12)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: Ubuntu 18.04.4 LTS

Matrix products: default
BLAS:   /usr/lib/x86_64-linux-gnu/blas/libblas.so.3.7.1
LAPACK: /usr/lib/x86_64-linux-gnu/lapack/liblapack.so.3.7.1

locale:
 [1] LC_CTYPE=en_GB.UTF-8       LC_NUMERIC=C              
 [3] LC_TIME=en_GB.UTF-8        LC_COLLATE=en_GB.UTF-8    
 [5] LC_MONETARY=en_GB.UTF-8    LC_MESSAGES=en_GB.UTF-8   
 [7] LC_PAPER=en_GB.UTF-8       LC_NAME=C                 
 [9] LC_ADDRESS=C               LC_TELEPHONE=C            
[11] LC_MEASUREMENT=en_GB.UTF-8 LC_IDENTIFICATION=C       

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base     

loaded via a namespace (and not attached):
 [1] Rcpp_1.0.3               bookdown_0.17            digest_0.6.24           
 [4] R6_2.4.1                 magrittr_1.5             git2r_0.26.1            
 [7] evaluate_0.14            BiocWorkflowTools_1.12.0 httr_1.4.1              
[10] stringi_1.4.5            rlang_0.4.4              fs_1.3.1                
[13] rstudioapi_0.10          rmarkdown_2.1            BiocStyle_2.14.4        
[16] tools_3.6.2              stringr_1.4.0            glue_1.3.1              
[19] tinytex_0.19             xfun_0.12                yaml_2.2.1              
[22] parallel_3.6.2           compiler_3.6.2           BiocManager_1.30.10     
[25] htmltools_0.4.0          knitr_1.27               usethis_1.5.1           

I also note this knitr error for what it's worth, though I obviously don't care about this one:

!!! Error: Input file `Article_files/figure-latex/frog-picture-1.pdf' not found!

Does it work any better for you if you grab the version from Github?

BiocManager::install('grimbough/BiocWorkflowTools')

I think some changes to the rmarkdown package has prevented it from using the template thats provided by BiocWorkflowTools, hence why you keep having to shove more and more things in the YAML header - those should all be defined in the template.

I can now do the following without error:

> rmd_file <- rmarkdown::draft(
+   "Article.Rmd",
+   template = "f1000_article",
+   package = "BiocWorkflowTools",
+   edit = FALSE
+ )
> setwd("Article/")
> rmarkdown::render('Article.Rmd', quiet = TRUE)
Warning message:
LaTeX Warning: Float too large for page by 20.96321pt on input line 308. 

There's clearly still some issue with resizing of images, so I'll look at that tomorrow, but it seem to at least compile for me now.