This chapter describes work carried out to optimise trial/guiding wavefunctions for use in VMC and DMC calculations. The motivation for optimising trial wavefunctions is introduced in the light of recent results for the ground state energy of solid germanium in the diamond structure using VMC and DMC[33, 50]. These results illustrate the difference in the quality of the standard trial/guiding wavefunction and the `groundstate' wavefunction that the DMC algorithm propagates towards. The variance minimisation technique used to perform the optimisation and the various choices of functional form to optimise are then described in a similar order to that in which the work was originally performed. Preliminary results are given at the end of each stage of the optimisation as they had a major influence on the future direction of the project.

- Motivation
- Optimisation Method
- Optimising the function
- Optimising the
*u*Function - A New
*u*function - Applying the New
*u*Function to Solids - Optimising Wavefunctions for Atoms
- Variance Minimisation on Parallel Computers
- Variance Minimisation with Non-Local Pseudopotentials
- Limits of Variance minimisation

Tue Nov 19 17:11:34 GMT 1996