A nice notebook and presentation - thank you.
In trying to reproduce the results I find that I have to set periodicboundaryconditions to False (instead of True). In that case I get the notebook results. If I set the condition to True (as it is in the notebook) I get a slightly different Hamiltonian along the diagonal, and though the rest of the notebook results are generally similar, the numerical values are rather different.
Also the Eigensystem solver warns that the Arnoldi method is not converging after 1000 iterations. The results however are still as presented in the notebook.
Mike Burns