For the past 18 months or so, I’ve been involved with the Software Carpentry organization. I got started with this as a result of teaching a “computation for astrophysicists” grad course and being a little frustrated with how well it went – I didn’t think I knew enough to do a good job of it. I took the Software Carpentry instructor course, which was run online over several months and included bi-weekly phone meetings and homework in between sessions. (My first-ever YouTube video was part of the homework; see it here.) Even after a number of years of university teaching, I found the course a very useful synthesis of how people learn and how this should inform teaching.

Since becoming an officially certified instructor I’ve taught at a couple of Software Carpentry workshops in Toronto, helped out at one at the most recent AAS meeting, and taught one of the lessons at the recent Python in Astronomy workshop. But then it was time to bring the knowledge home, so I organized a workshop at Western.

To sponsor the workshop I approached Western’s Center for Planetary Science and Exploration and its Brain and Mind Institute. Between some funds from those two institutes and registration fees, plus very low cost for instructors because most of us were local, I was able to pay for some catering plus the Software Carpentry administration fee. (Although I could have done the workshop without their admin support, I didn’t want to spend the time that would have been needed to collect money, refund it, issue receipts, etc., etc.)
The workshop capacity was 40 people; it filled up to about half capacity almost immediately, then completely filled about a week in advance. A few people registered and then dropped out but we were able to fill their spots.

We ran a slightly different curriculum than usual: normally Software Carpentry workshops are 2 days and cover the command-line, version control, programming, and (sometimes) databases in half-day blocks. We decided to do a 3-day workshop and include a couple of different options for programming: Python and MATLAB simultaneously, plus R for everyone on the third day. The teaching was done by Sahar Rahmani (bash, Python), me (Python, git), Paul Gribble (MATLAB), Xu Fei (R), and Greg Wilson (testing, big picture). Learner feedback after individual sessions was generally positive, although pacing was an issue for some people for the Python and bash sessions.

Overall things went quite smoothly. They would have been even smoother if I had read the appropriate checklist in advance, sigh. Workshop attendees were diligent about installing software beforehand, so we didn’t have too many setup problems aside from some wireless access problems for non-Western folks. I had stupidly booked a room for the workshop that didn’t have power outlets at the desks, but luckily Paul suggested checking this the previous week, so I had time to change the room, and the Evenbrite pre-workshop email let everyone know about it. I didn’t charge high enough registration fees to buy coffee every day (catering is expensive!) but there were cookies, and not too many complaints.

I really enjoyed teaching the git session; this was my fourth or fifth time through the material and I felt that I finally had it well-understood and well-paced. So now I need to teach all the other material the same number of times. One way to do this is to run another workshop here: I’m thinking fall sometime. There is some set-up and administration time involved, but it is not too bad. Please get in touch if you’re interested…