Teaching basic lab skills
for research computing

Feedback from the First MATLAB Bootcamp

Jan 28: Michael Croucher has also written a good post about what worked and what didn't at this bootcamp.

Last week Software Carpentry paired up with MATLAB to run a bootcamp at the University of Manchester. Apart from shell and version control, the remaining core topics (verification and testing, modular programming with algorithm design) were taught using MATLAB.

The bootcamp was offered to the University of Manchester staff and postgraduate students. We started on 14th January teaching 24 participants shell and then version control with Git. The day finished with the introduction to the MATLAB language and desktop environment. The next day covered algorithm design, testing, verification, debugging and improving performance. All these modules were taught using MATLAB.

The main organiser was Mike Croucher, a Software Sustainability Institute's Fellow. MATLAB material was taught by Jos Martin, Ken Deeley and Juan Martinez from MathWorks. The other two instructors were Shoaib Sufi and myself.

It was the first time that MATLAB was used for teaching at Software Carpentry. For a number of participants, the fact that we used MATLAB was an additional encouragement to attend the bootcamp.

In the feedback questionnaire, the participants said what they thought they learnt that would be useful:

  • Using Git and also some new functions for MATLAB
  • How to use cell arrays and tables in MATLAB, as well as a few useful functions. Very useful were the tips on coding practices and testing environments in MATLAB
  • Tricks in MATLAB, such as the section feature.

Even for more experienced attendees, the bootcamp turned out to be useful:

  • This is a little difficult to answer because this course actually helped me realise how much I already knew! But I did learn a couple of things (the difference between explicit/implicit memory allocation when a variable is initialised, and testing algorithms with various testcases). I am certain these will come in handy as I write algorithms 70% of the time!

Dialogue & Discussion

Comments must follow to our Code of Conduct.