This is a git repository for the HEC Gridding algorithm Latest information: Phase screen and FFT are consistent with normal w-projection kernels. We suspect that the bicubic is not selecting the correct neighbours/values during interpolation, and appears to be unbalanced. I suspect this is because we have been using full support * oversample in logic, when we should probably be using half support * oversample.
adamrcampbell/HECGridder
Datasets and precomputed grids for comparison can be found here: https://drive.google.com/drive/folders/1I_LP_rRmhlMUUfRwHxNBTsK1OCNpIjUC?usp=sharing
C