I finally recently made the changeover from developing R scripts in
emacs/ESS to Rmarkdown in RStudio. I love RMarkdown facilitating
creating reports and having a central document where I can comment on
code and analysis attempts (which I used to do as comments within R
scripts). The problem I find with Rmarkdown in RStudio, probably as a
common beginner's fault, is that such reports become unwieldly very
quickly especially when I am running complex analyses -- this quickly
rules out the possibility of creating a single report with all these
heavy chunks of code in it, even if the Rmarkdown file itself is small.
I have also noticed that sometimes the inline output of these chunks
disappear when I run code chunks with heavy processes/when RStudio
crashes/etc. (and hence also from the automatically knitted html file)
-- so as I am working on a complex set of analyses on the same data, I
end up having to rerun all previously run chunks if I want it in the
knitted file -- when some chunks involve particularly heavy processes,
this takes an inordinate about of time! I would really appreciate any
advice as to how to best deal with this situation and in general any
recommendations for good working habits with Rmarkdown -- or maybe is what I am doing simply not recommended in RStudio?
I see some useful advice in the thread
such as Andrea's suggestions
including putting analysis-heavy part of code outside of RMarkdown to an R script and then save to a file which can be loaded in RMarkdown. Would love to learn more from the experiences of RMarkdown users out there!
Thanks very much in advance!