- Laplace Regulariser
- Integrate w.r.t. r or x
- Apply Laplace w.r.t. r, x, or R?
- How to incorporate environment
- next logical step: simple sums of pair potentials for the environment representation
- then bond-angles?
- "Correct" weights for forces and virials
- Choice of space transform
- speed vs accuracy, polynmomials vs exponentials
- possibly optimise the meta-parameters e.g. the a in exp( - a k r )
- high-pressure data and match with 2B for core (forgot the name)
- different choices of the angle variable, angle, cos, or just ip (current)
- BL vs BA
- Is there some way to use orthogonal invariant polynomials?