5 Final Notes on The OxThesis template and on collaboration
5.1 Unnumbered chapters
To make chapters unnumbered (this is normally only relevant to the Introduction and/or the Conclusion), follow the chapter header with {-}
, e.g. # Introduction {-}
.
When you do this, you must also follow the heading with these two latex commands:
\adjustmtc
\markboth{The Name of Your Unnumbered Chapter}{}
Otherwise the chapter’s mini table of contents and the running header will show the previous chapter.
5.2 Beginning chapters with quotes
The OxThesis LaTeX template lets you inject some wittiness into your thesis by including a block of type savequote
at the beginning of chapters.
To do this, use the syntax ```{block type='savequote'}
.4
Add the reference for the quote with the chunk option quote_author="my author name"
.
You will also want to add the chunk option include=knitr::is_latex_output()
so that quotes are only included in PDF output.
It’s not possible to use markdown syntax inside chunk options, so if you want to e.g. italicise a book name in the reference use a ‘text reference’: Create a named piece of text with ‘(ref:label-name) My text,’ then point to this in the chunk option with quote_author='(ref:label-name)'
.
5.3 Highlighting corrections
For when it comes time to do corrections, you may want to highlight changes made when you submit a post-viva, corrected copy to your examiners so they can quickly verify you’ve completed the task. You can do so like this:
5.3.1 Short, inline corrections
Highlight short, inline corrections by doing [like this]{.correction}
— the text between the square brackets will then be highlighted in blue in the output.
5.3.2 Blocks of added or changed material
Highlight entire blocks of added or changed material by putting them in a block of type correction
, using the syntax ```{block type='correction'}
.5
Like so:
For larger chunks, like this paragraph or indeed entire figures, you can use the correction
block type. This environment highlights paragraph-sized and larger blocks with the same blue colour.
5.3.3 Stopping corrections from being highlighted in the output
For PDF output, go to index.Rmd and (i) set corrections: false
under params
in the YAML header (stops block of corrections from being highlighted), (ii) comment out pandoc_args: ["--lua-filter=scripts_and_filters/correction_filter.lua"]
(stops inline corrections from being highlighted).
For gitbook output, go to style.css and comment out the styling for .correction
.
5.4 Diving in to the OxThesis LaTeX template
For LaTeX minded people, you can read through templates/template.tex to see which additional customisation options are available as well as templates/ociamthesis.cls which supplies the base class. For example, template.tex provides an option for master’s degree submissions, which changes identifying information to candidate number and includes a word count. At the time of writing, you must set this directly in template.tex rather than from the YAML header in index.Rmd.
5.5 Collaborative writing
Best practices for collaboration and change tracking when using R Markdown are still an open question.
In the blog post One year to dissertate by Lucy D’Agostino, which I highly recommend, the author notes that she knits .Rmd files to a word_document
, then uses the googledrive
R package to send this to Google Drive for comments / revisions from co-authors, then incorporates Google Drive suggestions by hand into the .Rmd source files.
This is a bit clunky, and there are ongoing discussions among the R Markdown developers about what the best way is to handle collaborative writing (see issue #1463 on GitHub, where CriticMarkup is among the suggestions).
For now, this is an open question in the community of R Markdown users. I often knit to a format that can easily be imported to Google Docs for comments, then go over suggested revisions and manually incorporate them back in to the .Rmd source files. For articles, I sometimes upload a near-final draft to Overleaf, then collaboratively make final edits to the file there. I suspect some great solution will be developed in the not-to-distant future, probably by the RStudio team.