What is JHPCE? JHPCE or the Joint High-Performance Computing Exchange is a computation resource available to members at the Johns Hopkins Medicial Institute (JHMI). Several research groups across the School of Public Health and School of Medicine tap into the power of JHPCE to run computationally expensive analyses - our transplant epidemiology group included. The JHPCE introduction to new users describes some use cases. This blog will be a brief introduction to getting your system ready to log onto the JHPCE cluster to use Stata-MP.
Would you be surprised to find out that the figures above were made in Stata? I discover the work of Daniel Bischof about a year ago and it has changed the way I think about Stata figures. I have been a long time user of ggplot2, the graphing R package by R Guru Hadley Wickham. Some of my publication figures using ggplot2 are dozens of lines long. I enjoy fiddling with every aspect of the aesthetic to make a perfectly clear visualization.
I wanted to give a quick shoutout to the Effort Report Podcast hosted by Roger Peng and Elizabeth Matsui. Both hosts are faculty members at Johns Hopkins (School of Public Health & School of Medicine) and they discuss issues that any budding academic should consider. I’ve listened to every podcast (some multiple times) and I’ve even joined the conversation on Twitter.
I recently discovered that Mintty is a great way to make your Cygwin terminal prettier. Mintty made my conversation from Fedora to Windows 10 and Cygwin more palatable. Like all Cygwin pacakges/add-ons you’ll have to install it using the Cygwin installer. You can find some quick instructions at HowToGeek. While there are a number of themes out there, I really like this Base-16 inspired set. I’ve also done some other tweaks to make my Cygwin terminal behave like my Mac Terminal (such as transparency when not in focus).
I prefer to create new log files every time I (fully) run a Stata script. When I’m writing a script or testing code I might stop logging, but in general it’s wise to keep logs of important runs. I store my logs in their own directory so that they don’t clog up my code directory (see organization). I title them with the scripts name followed by the date and time.