Garret Christensen–BITSS Project Scientist
If you’re coming to the Annual Meeting this Thursday, you may be curious what exactly we’re doing during the optional training workshop Thursday morning. I’ll give a brief overview of software tools that can help with reproducible workflow, and then we’ll spend most of the time demonstrating dynamic documents (which are a means of producing reproducible, automated papers that have your code and your final output all in one place so you don’t make copy and paste mistakes transferring results from R/Stata to your final paper in Word/LaTeX/whatever) and maybe a tiny bit of version control (a very powerful way to keep track of versions of your code so you don’t end up doing this). The hope is that you may be able to play along a little bit with the live demo, but we’ll also briefly expose you to some more advanced tools that we won’t have enough time to fully teach.
Which exact programs we’ll talk about depends on who shows up, but the two major options are:
There are several ways to do dynamic documents in Stata, but I’m most familiar with markdoc. If you’ve ever installed it before, run
adoupdate to make sure you’ve got the latest version. If you’ve never installed it before, run:
ssc install markdoc
ssc install weaver
ssc install statax
You may also be prompted to install pandoc and wkhtmltopdf, which need to run behind the scenes to translate documents from one format to another. In my experience, these installs can be done from within Stata. If you’re going to use the PDF/LaTeX capabilities, you’re also going to need to have LaTeX installed. (It’s a big file, and must be installed outside Stata.)
In my opinion the dynamic document capabilities in R are vastly superior to those in Stata. Download both R and R Studio. If you already have them installed, it’s a very good idea to update, especially if your version of R Studio is < version 1.0. (In my experience you’ll often find things breaking if you don’t update R frequently.) Again, you’ll also need to have LaTeX installed if you want to use any of the PDF capabilities.
For those installing on Windows, things usually work out better if you right-click on the install files that you download and
run as administrator. Things should still work if you don’t have admin access, as you can usually install in your user directory.
It would be great if you can attempt to install things before the workshop, but don’t worry if you have issues. We’ll have helpers walking around to, well, help! Lastly, I mentioned version control. We may only have time to talk about this briefly, and I’ll demonstrate it using the built-in version control GUI in R Studio, but you can install Git here.