Closed ffreyer closed 3 years ago
There's also some cleaning up on the interface that needs to be done. For example, I added lattice(::DQMC)
and lattice(::Model)
, so now it makes sense to move away from nsites(model)
and use length(lattice(...))
instead.
Changed in #87
For performance it would also be good/interesting to give more control over what the interaction matrix is and how it is initialized. A Diagonal
would surely perform better than a Matrix
here.
Added in #88
You can currently create am attractive Hubbard model with any lattice, but the docstring and the constructors do nott reflect that.