Hi Matthias
Thanks, now I am getting the right plots.
Vipin
On Fri, 19 Feb 2010, Matthias Troyer wrote:
Hi Vipin,
On 19 Feb 2010, at 07:36, Vipin Varma wrote:
Hi Matthias
I just choose a zero chemical-potential; is a non-zero value required to extract the correct t/U dependence?
When I chose Nmax = 2, I want the 2x1 cluster to have totally 2 bosons over the 2 sites, with a maximum of 2 bosons per site i.e. 3 Fock-eigenstates: 02, 11, 20. Is my interpretation of Nmax correct?
You indeed have a maximum of 2 bosons per site, but do not fix the total number. Hence you have the following Hilbert space:
00, 01, 10, 11, 20, 02, 21, 12, 22
To fix the number of particles in the ED codes you need to set CONSERVED_QUANTUMNUMBERS="N" and N_total=2, otherwise you will work in a grand canonical ensemble with mu=0. The QMC codes always work in the grand canonical ensemble.
Matthias