Hi everyone, I want to simulate the Hard-Core Bose-Hubbard model using an initial Mott-like state of alternating filled and unfilled sites, and study its evolution toward a superfluid state. As I am new here, I would appreciate any ideas or guidance.
Hi Souritra,
Sorry for the slow reply on this!
The easiest way to make an initial state which is a product state is to use our special MPS constructor that takes an array of strings which are ‘state names’. The allowed names are listed on this documentation page.
For the case of hardcore bosons, you can do this:
using ITensors
using ITensorMPS
# ...
N = 6
sites = siteinds("Boson",N; dim=2)
state = [isodd(j) ? "0" : "1" for j=1:N]
psi = MPS(sites,state)
The default dimension of “Boson” sites is actually 2, so the dim=2 keyword isn’t required above, but I put it to show how can one can also do regular (non-hard-core) bosons up to some maximum occupation.