Abstract

UBiKuity is a software library that I am currently developing to model the behaviour of charged particles in magnetic and electric fields, with the intention of eventually converting it into a full magnetospheric plasma model. It uses an under-explored mathematical formulation of charged particle motion to generate powerful insights into the phase space of possible particle behaviour. It combines a variety of numerical methods to find solutions to differential equations, with most of the numerical code written in Fortran. As it is intended to be shared both within my research group, and eventually with the wider community, I am simultaneously writing a Python wrapper to improve its usability.

Like many other pieces of software in the scientific community, UBiKuity has been written by someone with little to no training in programming software engineering. Because of this, I have tried to treat this first foray into research software engineering as something of a learning experience. This talk will discuss some of the challenges I have faced in attempting to write shareable, performant code to solve a complex research problem.