The mapping to a spin model is not needed and will actually give wrong results. WHy don't you just use a Bose Hubbard model with a cutoff?
Also a parameter file would indeed help, as well as knowing the problems you have.
On 22 Mar 2011, at 14:37, Fernando Cucchietti wrote:
Can you send the parameter file?
I'd say you have to put a cutoff in the number of particles per site, use a mapping to a spin model, and make Sz a conserved number.
Fer
On 22/03/2011, at 20:15, "Michał Maik" michal.maik@uj.edu.pl wrote:
Hi
I was wondering if DMRG works with the boson Hubbard model? I tried making a parameter file for a 1D open chain lattice with filling factor of one and the DMRG routine wouldn't work. Are there certain parameter restrictions that must be put into place for it to work? Thanks
Regards
Michal Maik