All of our events may be recorded and shared via the University of Sheffield Kaltura platform so those who cannot attend may still benefit. We will consider your attendance implict consent to this.
Abstract: Researchers often have to make trade-offs between expressiveness and performance when implementing their models. For sufficiently specific problems, these trade-offs can be removed by interfaces that reduce the gap between the way they would describe their problem and optimal ways for a computer to solve it.Ordinary Differential Equations (ODEs) turn up in many areas of infectious disease modelling and represent a trade-off of this form, especially in high-level languages such as R. We have developed a new R package “odin” (https://mrc-ide.github.io/odin) to reduce this trade-off by creating a domain specific language (DSL) hosted in R that compiles a subset of R to C in order to efficiently express and solve ODEs.
I will present applications of “odin” both in a research context for implementing epidemiological models with 10’s of thousands of equations and in teaching contexts where we are using the introspection built into the DSL to automatically generate interactive web-based interfaces. I will also discuss how our RSE group works to support a large group of epidemiological modellers, how we use agile processes to manage our development, and to briefly highlight other tools from our group that help with general problems.
Bio: Rich has been a research software engineer in the “RESIDE” group (https://reside-ic.github.io/) at the Department of Infectious Disease Epidemiology and MRC Centre for Global Infectious Disease Analysis for the last 4 years. His focuses are infrastructure and tools that generalise problems common to research groups across the department. He is interested in reproducible research and in helping researchers get more science done per line of code that they write. Before moving to work full-time as an RSE, his research career involved modelling coexistence in tropical forests, diversification over macro-evolutionary timescales and the potential for gene flow from genetically-modified crops.
For queries relating to collaborating with the RSE team on projects: email@example.com
Join our mailing list so as to be notified when we advertise talks and workshops by subscribing to this Google Group.