Software Carpentry is a volunteer organization whose goal is to make scientists more productive, and their work more reliable, by teaching them basic computing skills. Founded in 1998, it runs short, intensive workshops that cover program design, version control, testing, and task automation.
Members of our work group are certified instructors of Software Carpentry. Two workshops were organized in 2016:
• a software carpentry workshop covering basic concepts and tools, including program design, version control, data management, and task automation. This workshop was hosted by the UNIC laboratory and funded by the ENP.
• a workshop based on Software Carpentry2 covering more advanced topics (packages for scientific computing with Python, debugging and testing, and packaging). This workshop was hosted by Telecom ParisTech and funded by the CDS.
We will keep running these workshops in 2017.
Training for Data Challenges
We developed a new curriculum to teach researchers how to organize their own challenges that was taught as part of the Paris-Saclay Master of Data Science.
The greatest need that we have identified after years of practice as researchers in data science is finding people capable of formulating data science problems from raw data. This training aims to introduce students in science in the design of scientific projects based empirical data made available to them by researchers from the CDS, in the form of mini scientific challenges.
Some of the RAMPs are preceded by a single-day training session where we explain the tools used in the RAMP in a hands-on course. We have run four such events on:
• hyperparameter optimization,
• functional data,
• deep learning and computer vision,
• macroeconomic agent-based models.