ARCHER2 computational training for researchers
13 February 2025
EPCC's Evgenij Belikov writes about the two-day course for researchers he recently delivered at Cardiff University using materials developed by The Carpentries.
Most research undertaken now involves several computational steps which are often critical for advancing scientific discovery. As part of delivering the UK national supercomputing service ARCHER2, EPCC is committed to providing training at all levels of expertise to the user community, from introductory courses to Threaded Programming with OpenMP and Message Passing Programming with MPI, and GPU Programming. We also offer in-depth webinars on profiling, optimisation, and particular parallel applications commonly used on ARCHER2.
Recently we delivered a two-day Data Carpentries course at Cardiff University using the materials developed by The Carpentries, a non-profit organisation that teaches foundational coding and data science skills to researchers worldwide. The workshop includes sections on data organisation with spreadsheets, data exploration and cleaning with OpenRefine, introduction to programming and visualisation using the R programming language and the RStudio integrated development environment, and an introduction to SQL (structured query language used for interaction with databases). See the syllabus and materials on the Data Carpentry website.
The participants followed along the live demonstrations and worked through a number of practical exercises to equip them with skills necessary to effectively work with their own data and apply best practices to improve the reproducibility of their research. Another benefit is the ability to automate repetitive tasks and data visualisation, which is particularly useful for early-career researchers who might still be in search of an optimal workflow for their use cases.
Learner feedback
One of the challenges of providing inclusive introductory, no-prerequisites training workshops is the breadth of the audience. Participants may find the pace of learning either rather fast or slow, and there is limited time available to cover all the materials in two days. However the following feedback reflects that outcomes for participants were positive:
"It was excellent for beginner level users of the various tools. The combination of processing from Spreadsheets -> Openrefine -> R -> SQL was excellent for giving a flavour of the different tools that can be used and how they complement each other."
"Good resources, useful references if I ever need a refresher on the materials."
"I really enjoyed this course, it was pitched generally at a good level, and I learned a lot."
Learning pathway
The knowledge and skills gained on the workshop will allow the participants to continue their learning journey as self-directed learners, using the freely available materials to proceed towards more advanced topics at their own pace. One example pathway is to continue from Data Carpentries to HPC Carpentry (see this introduction to high-performance computing), and then to complete the online ARCHER2 Driving Test, which provides access to ARCHER2 to enable researchers to test their workflows on the national supercomputer.
We would like to thank Jose Munoz Criollo from Advanced Research Computing at Cardiff University for hosting and helping to deliver the workshop.
Please reach out if you are interested in hosting an ARCHER2 training workshop at your institution.
