3. Introduction

Neutron scattering experiments are useful tools for probing atomic positions and molecular dynamics in materials. Computational simulations and modelling are essential tools to analyse and interpret such experiments. These interpretations help to improve existing materials for bespoke applications and design new ones. Atomistic simulations, particularly molecular dynamics (MD) simulations are being used increasingly for these purposes. However, predicting neutron observables from MD trajectories is not straightforward. A number of operations, such as calculations of mean square displacements, densities of states, velocity and position auto- and cross-correlation functions, Fourier transformations and convolutions with instrument parameters are required to calculate neutron observables that can be compared directly with experimental data.

Some of these steps were implemented in the open source MDANSE (Molecular Dynamics Analysis of Neutron Scattering Experiments) [Ref1]. This software is a Python based application for analysing MD simulation data. This has interface with more than ten MD codes including ab-initio MD codes as listed in Trajectory Converters. MDANSE is shipped with a simple python-based graphical users interface (GUI) which allows the users to run analysis jobs and visualise their results. However, the GUI is not necessary to use MDANSE, as it can also be used through command line utilities and within Python scripts. In addition to this GUI, a well developed molecular viewer and 2D/3D plotter improve the users experience in analysing neutron experimental data.

This users guide provides a detailed overview of the capabilities of MDANSE along with theoretical background and installation instructions on three different platforms Windows, Mac OS and Ubuntu. Authors will be happy to receive any suggestions, feedback and bug reports about the MDANSE software and this Users guide.

Note

MDANSE branch 1.5.x is still based on Python 2, which became obsolete in 2020. For this reason the users of the Apple Silicon family of computers may find it impossible to use the 3D visualisation capabilities of MDANSE. If you do own a computer which is running MacOS on an ARM-based CPU, you may have to wait for the new release of MDANSE in 2023.