2015 Post-Workshop Instructor Debriefing, Round 22

This post originally appeared on the Software Carpentry website.

There were relatively few attendees at the 22nd round of debriefing for instructors on November 10, but we thought we'd include a few quick notes about two specific topics: using Software and Data Carpentry workshops to reach out to potential users of high-performance computing (HPC) resources, and exploring new lesson materials related to databases.

High Performance Computing

University of Manitoba hosted a Software Carpentry workshop at the end of October in conjunction with Compute Canada. In addition to the normal half-day lessons in Unix shell, Git, and intro to Python, this workshop featured an introduction to using WestGrid/Compute Canada resources on the afternoon of the second day. The instructor for this workshop, Hossein Pourreza, said that this last lesson module was used as a capstone for helping reinforce and integrate material from the previous lessons, and appeared well-received by the students. Given that there are few training opportunities for using the cluster, this appears to be a great way to introduce new potential users to the basic tools they'll need to get started with larger-scale analysis.

This recent workshop is a great example of catering the lesson materials to meet the needs of a particular audience. There are some resources currently available for outlining essential skills for remote computing. For example, Data Carpentry has a genomics lesson involving cloud computing under development.

Databases

Continuing the theme of piloting new material, a recent workshop at the Pacific Northwest National Laboratory included a short lesson on MongoDB, added at the end of the official Software Carpentry SQL material. The motivation for including MongoDB (an example of a NOSQL database) was the instructor's own use of it in his work, and it's growing influence in scientific computing.

It was hard to draw conclusions about the value of introducing NOSQL, mostly because there isn't enough time in a 2-day workshop to do git, shell, Python *and* databases, especially with the lessons as written. If instructors want to include databases (SQL or otherwise) in their workshops, they should be aware that the Python lesson will probably have to be significantly shortened to fit into half a day. This could be a good strategy for workshops with a specific audience, where all the participants are not novice programmers. Alternatively, if a workshop is being taught with local instructors, databases could be a follow-on day or half day after the first two days.

Question for the community: do other instructors use NOSQL databases in their daily work? In what circumstances is it a useful tool or skill?

Thanks

We are grateful to the instructors who attended debriefing sessions this round:

• Hossein Pourreza
• Donny Winston