Dear Prof. Troyer Thanks for the reply. Something bothers me as follows. Let's say if in the script I have NONLOCAL = 0 U = ... t = ... V = 1.0 Will the simulation be carried out with V = 0? I checked the trial runs using above parameters against the known results in the literature regarding the extened Bose-Hubbard model with non-vanishing V and find a good agreement.... Answers will be gratefully appreciated! Sincerely, Fu-Jiun On Mon, 23 Nov 2009, Matthias Troyer wrote:

On Nov 23, 2009, at 4:32 PM, Fu-Jiun Jiang ITP wrote:

...

Dear ALPS users

For the extended Bose-Hubbard

H = -t\sum_{<i,j>} a^{\dagger}_i a_j +\frac{U}{2}\sum_{i}n_i(n_i-1) + V\sum_{<i,j>}n_in_j -\mu\sum_{i}n_i,

I wonder why worm and dirloop_sse in ALPS show two completely different results?

Another technique problem as follows. Can one change parameters Nmax = 2 or NONLOCAL = 0 in the tutorial? I am not sure what's the impact of these two parameters. Any help will be greatly appreciated. THANKS!

Sincerely, Fu-Jiun

NONLOCAL = 0 means there are no nearest neighbor interactions besides the hopping. That means the V term will be ignored

Matthias

Dear Prof. Troyer Thank you very much for your quick response. Sincerely, Fu-Jiun On Thu, 26 Nov 2009, Matthias Troyer wrote:

On Nov 26, 2009, at 3:14 PM, Fu-Jiun Jiang ITP wrote:

...

Dear Prof. Troyer

Thanks for the reply. Something bothers me as follows. Let's say if in the script I have

NONLOCAL = 0 U = ... t = ... V = 1.0

Will the simulation be carried out with V = 0? I checked the trial runs using above parameters against the known results in the literature regarding the extened Bose-Hubbard model with non-vanishing V and find a good agreement.... Answers will be gratefully appreciated!

Sincerely, Fu-Jiun

Using the dirloop_sse code there should be no difference. NONLOCAL is used only in the worm code.

Matthias