This is an experiment to compute harmonic differentials in a translation surface; there are lots of convergence problems at the moment and we might end up doing things very differently.
TODO
[x] We need parts of #211 to compare differentials on differently triangulated surfaces.
[x] On the octagon: Compare quality of differential with the explicitly computed differential.
[x] We need to compute the same differential in another triangulation so we to be able to map cohomology classes through deformations
[x] To compare the differentials, we just need to evaluate at points which means we need to be able to map points through deformations.
[ ] Given a differential, compute the location of its zeros and their order of vanishing. (Up to some numerical error to decide whether two zeros are the same or not.)
[ ] For an L, construct the 1-form associated to any real class in the symplectic orthogonal of span(Re(ω), Im(ω)). Check that it has a double zero.
[ ] Given a 1-form η, compute -1/2i int η wedge bar{η}, see _area_upper_bound() in the old optimization approach.
This is an experiment to compute harmonic differentials in a translation surface; there are lots of convergence problems at the moment and we might end up doing things very differently.
TODO
_area_upper_bound()in the old optimization approach.Checklist
doc/news/.doc/geometryordoc/graphical.