Exact spherical harmonics coefficients for weighting function h by Malecki
With the help of Mathematica I could get the analytical terms for eta, so we won't need numerical integration in the setup of AXDTOperator anymore. Tomorrow I'll measure the speed improvement, but I hope my code should be substantially faster... EDIT: No need to measure, it's noticeably faster!
The coefficients finally agree with the numerical integration (modulo normalization)!
Also I refactored getGeometryAt to throw...
Edited by Cederik Höfs