This package implements a few polynomial basis types, convenient methods for evaluation, derivatives up to second order and (hopefully fast) batched evaluation. The bases currently implemented include:
- Various orthogonal polynomials via 3-point recursion
- Trigonometric polynomials
- Complex and real spherical and solid harmonics
- Some quantum chemistry basis sets
- Utilities to recombine them into (tensor) product basis sets
- Utilities to implement cluster expansion models
We also aim to provide full Lux.jl integration to build layered models. A possible application of this might be to implement various flavours of equivariant deep neural networks.