ANSTO internship No. 4
- ANSTO Project supervisor : Andrew Nelson
- Project duration: 3 months
- Nominal project commencement date: Flexible
refnx is an open source Python package for the analysis of neutron and X-ray reflectometry (NR, XRR) data, https://refnx.readthedocs.io/en/latest/. The package is used by researchers worldwide who have used those techniques to study thin interfacial films, https://refnx.readthedocs.io/en/latest/testimonials.html. Recently the project has also developed functionality to analyse ellipsometry data, making it possible to co-refine Neutron, X-ray and light reflection data using a joint model. refnx has a modular object oriented architecture, allowing extensibility for new features.
When analysing NR data refnx is only able to analyse unpolarised measurements. In contrast Polarised NR measurements are used for structural characterisation of thin magnetic films; for example the size and direction of a film’s magnetic moment can be determined in different surface strata. This project will add functionality to refnx to be able to analyse polarised NR data.
Ideally the intern should be familiar with polarised neutron reflectometry, or have a physics/maths background capable of understanding how a magnetic thin film structure can be used to generate a polarised neutron reflectometry curve.
Work will proceed by porting code for the underlying physics calculation from another analysis package, before adding refnx components that can use that calculation.
refnx is written in Python, so a basic knowledge of programming and Python is required. However, supervision and guidance will given in that area. During the project the intern will:
- develop their Python skills
- gain a deeper understanding of scientific programming
- learn and practice test driven development
- learn how to use git/Github for distributed version control
- develop a web-based interface for calculating diffraction angles via a UB-matrix
Supervisor bio: Andrew Nelson is a scientist using Neutron and X-ray reflectometry to characterise thin films at interfaces. His research interests are Organic Semiconductors, stimulus responsive polymer brushes, and NR/XRR technique development. He is particularly interested in how computer software can be used to faciliate more complex experiments. This lead him to create Motofit and refnx for analysing NR/XRR data. Powerful features, such as Bayesian statistics for model selection/posterior determination, has greatly aided scientists around the world in coming to grips with their data. Andrew is recognised as a lead developer of scientific software, and is a core member of the SciPy, refnx, and refellips projects.
- Project location: ANSTO Lucas Heights campus
- Requirement for ANSTO site security clearance (requires 3 months advanced processing): Yes
- Option for to work remotely? Yes, but will need to comply with site attendance expectations. It’s difficult to give supervision for this project remotely.
- Will an interview step be required? Yes
- Site attendance expectations: negotiable, but at least 1 – 2 days a week
- Requirement for a non-disclosure agreement (must be handled by the ANSTO affiliate itself): No