On 14 May 2010, at 18:38, Michał Maik wrote:

Hi,

I am confused as to how the energy is calculated by the different simulations. When I use the worm or SSE algorithm I receive a different answer then when I use full diagonalization. As an example I have attached a simple parameter file below. Using worm and SSE I get the same energy while full diagonalization gives me a different answer. I have not modified this file at all, just ran it with each of the programs. I would really appreciate an explanation as to why SSE and worms do not give me the same answer as full diagonalization even at such a low temperature. Thank you.

Michal

LATTICE_LIBRARY="/usr/local/lib/xml/lattices.xml"; LATTICE="chain lattice"; L=6

MODEL_LIBRARY="/usr/local/lib/xml/models.xml"; MODEL="boson Hubbard";

U = 1; V = 0; mu = 0;

CONSERVED_QUANTUMNUMBERS = "N"

Nmax = 6; N_total=6;

T=0.01

SWEEPS = 1000000 THERMALIZATION = 50000

{ t = 0.08; }

Hi,

The sse and worm code work in the grand canonical ensemble with mu=0 while for the diagonalization you force the particle number to be 6.

Matthias