Driving simulation and data analysis of magnetic nanostructures through Jupyter Notebook


Upvotes: DownVotes:
Age: a year     Page Views: 81
Votes / View: 0    Wilson Score: 0.00

We present ongoing work from a project that makes a particular computersimulation (implemented in C++ and Tk-Tcl) accessible through a Pythoninterface, and through the Jupyter Notebook. The talk describes themotivation and current status of the project.In more detail, the computer simulation in question is the ObjectOriented Micromagnetic Modelling Framework(`OOMMF `__) which is likely the mostwidely used micromagnetic simulation package. It can be driven through agraphical (Tk) user interface or through a configuration file thatdefines a simulation run.In this talk, we first show a Python interface to OOMMF that allows thedriving of OOMMF simulations from a Python program or interpreterprompt. This way we embed a widely used scientific code from 1990s in ageneral purpose programming language[`1 `__] and enable the full use ofthe ecosystem of scientific libraries available for Python. For example,design optimisation, specialised post-processing, and the creation offigures can all be carried out using a single script; making the workmore easily reproducible.Second, we integrate the Python interface to OOMMF into a Jupyternotebook, so that all existing benefits of using Jupyter are inheritedfor the use in computational micromagnetics, which is the reason wenamed our code Jupyter- OOMMF (`JOOMMF `__). A`JupyterHub installation `__ of the toolreduces barriers in uptake, and all the code is `ongithub `__.We discuss the benefits of driving computer simulation and data analysisthrough Jupyter Notebooks.This project is a part of the Jupyter-OOMMF (JOOMMF) activity in the`OpenDreamKit `__ project and we acknowledgefinancial support from Horizon 2020 European Research Infrastructuresproject (676541). The work is also supported by the EPSRC CDT in NextGeneration Computational Modelling EP-L015382-1, and the EPSRC grantsEP-M022668-1 and EP-N032128-1.For additional context: micromagnetic modelling is a key research methodin academia and industry to support development of high-capacitymagnetic storage devices that are cheap, fast, and reliable, and toenable research into future alternative storage and processingtechnologies such as spintronics. The OOMMF modelling package has beenused in `over 2500publications `__ since1999.[1] Beg, M., Pepper, R. A., and Fangohr, H. User interfaces forcomputational science: A domain specific language for OOMMF embedded inPython. AIP Advances 7, 056025 (2017), https:--doi.org-10.1063-1.4977225