Uncertainty quantification in diffusion#

kinisi is an open-source Python package focussed on accurately quantifying the uncertainty in diffusion processes in atomic and molecular systems.

The approach used by kinisi ensured an accurate and statistically efficient estimation of the diffusion coefficient and ordinate offset. More information about how kinisi determines the diffusion coefficient can be found in the detailed methodology.

kinisi can handle simulation trajectories from many common molecular dynamics packages, including VASP and those that can be read by MDAnalysis. Examples of some of these analyses are shown in the tutorials, also given there is an example of using kinisi to investigate the Arrhenius relationship of diffusion as a function of temperature.

An example of the kinisi analysis for the diffusion of lithium in a superionic material.

An example of the output from a kinisi analysis; showing the determined mean-squared displacements (solid black line), the estimated Einstein diffusion relationship (blue regions representing descreasing credible intervals), and the estimate of the start of the diffusive regime using the maximum of the non-Gaussian parameter (green vertical line).#


Andrew R. McCluskey | Benjamin J. Morgan


Index | Module Index | Search Page