Closed stevengj closed 2 years ago
I will look further into this but my trials for now seem to indicate that it might be more difficult to adapt chetrd
than changing a few signs.
I think I have found how to adapt the algorithm chetrd
, the complex hessenberg reduction should be working soon
The complex hessenberg works fine but I must improve the memory allocations.
Might be interesting as an exercise to generalize
sktrd!
to work on complex-Hermitian matrices … hopefully it's as simple as putting a few complex-conjugations in the right places.