Open pablosanjose opened 3 months ago
The reason is that h
is non-hermitian. Closing, as this case is not intended to work with band connection
There is still an issue. If we make this Hermitian, the Dirac points fail to be properly connected
function build()
lat = LP.triangular()
σ1 = SA[0 0; 1 0]
model = onsite(σ1+σ1') + hopping((r, dr) -> dr[2] == 0 ? zero(σ1) : dr[1] > 0 ? σ1 : σ1')
h = lat |> hamiltonian(model; orbitals = 2)
return h
end
Note that the conventional graphene case (which would read model = onsite(σ1+σ1') + hopping((r, dr) -> dr[2] == 0 ? zero(σ1) : dr[2] > 0 ? σ1 : σ1')
) is properly connected
An equivalent one-liner: HP.graphene() |> supercell(1, -1) |> bands(range(0, 2π, 19), range(0, 2π, 19)) |> qplot
This code produces a completely wrong band connection pattern