Teaching basic lab skills
for research computing

Software Carpentry in Paris !

A couple of weeks ago, we organized the first Software Carpentry Bootcamp in Paris. We were lucky enough to be able to host it at INRIA's research center, Place d'Italie. The focus of this bootcamp was Python, with 3 interventions on python: beginner's introduction, by Christophe Combelles, numerical computing with numpy by Konrad Hinsen and the python scientific ecosystem by Alexandre Gramfort, and an introduction to DCVS with git, by myself.

22 people joined us for 2 days of intense tutorials (24 places, 2 didn't show up, 15 people on the waiting list). Out of those 22 people, 4 travelled from abroad (UK, Belgium and Germany) to participate, and many travelled from distant cities in France (Rennes, Toulouse, Orléans, Nantes). 10 of the attendees filled in our only survey.

Software Carpentry being initially a north american project, and not benefiting from a joint organisation with a conference, like the italian bootcamp, nor the Greg Wilson effect, we were a bit worried about the communication around the event. We communicated on INRIA's and IRILL's website, and we flooded several universities of posters. We also posted on LinuxFR, a french website about free and opensource technologies and events, on AFPy, the French speaking Python community, and our two sponsors, Majerti and Logilab twitted about the event.

We filled in the bootcamp quite late, and many people subscribed on the mailing list only a couple of days before the event.

As the other bootcamps, we targeted mostly researchers. People worked in different fields: applied mathematics, bio informatics, structural biology, finance, ecology, solid state physics, bio physics, linguistics, and at differents levels: researchers, master students, postdocs, phd students, engineering students. The variety of fields made it hard for speakers to evaluate the levels of the talks and the subjects. One common point of interest of the attendees seemed to be to find a replacement of matlab (because of the licence price), and the Python part of the bootcamp was very attractive.

The bootcamp was pretty intensive, with a lot of quite advanced exercise. I was worried that the level was too high, but it appears that most the attendees found the level just right or hard, but understood the concepts (none of them checked the "Too hard, I didn't understand anything" box!).

Here are some suggestions on how to improve the bootcamp:

  • Ask attendees before hand the python modules they would be interested in.
  • Give pre-bootcamp reading list to smoothen the levels of the participants
  • Don't use vim to edit files! It's confusing for windows participant that don't have any knowledge in using the shell.
  • Make the bootcamp longer (5 days).
  • Make different level bootcamps.

"I think the bootcamp was great, I learnd a lot and especially found it useful that the lecturers pointed towards certain modules coping with the variety of interests and backgrounds of the participants. I will definitely delve deeper into Python now. Hopefully there will be more bootcamps in Europe, maybe even with different levels from beginners to more advanced users. Thanks for this great initiative!"

Overall, attendees were very satisfied with the bootcamp, the facilities! Our sponsors greatly helped us making a successful event. They sponsored coffee breaks, croissant, orange juice, and meals for both days! Thanks once again to the speakers, our sponsors and Feth Azreki, who came to help with the exercises.

Dialogue & Discussion

You can review our commenting policy here.