A decade ago a new sampling scheme, nested sampling, was introduced in the field of applied probability and inference, to sample probability densities in high-dimensional spaces where the regions contributing most of the probability mass are exponentially localised. In recent years I have been working on adapting the nested sampling approach for exploring atomic configurational phase. Nested sampling automatically generates all relevant atomic configurations, and makes the direct calculation of the partition function – and thus all thermodynamic observables – possible, and it does not require advance knowledge of the studied system.
For the nearer future my aims are to extend the nested sampling method to be
suitable for routine use in a wide variety of research fields, from material science to
biochemistry, and to make it available for a wider community. Currently I am working on calculating the phase diagram of a variety of metals with nested sampling.