Yes, you need to set the boundary conditions in the lattice definition. On 18 Jun 2014, at 08:05, Daniel Guterding <guterding@itp.uni-frankfurt.de> wrote:

Dear Matthias,

thank you very much for your answer. I tried to set BOUNDARY="open"; in the input parameter file, but this had not effect, probably as intended.

Meanwhile I have figured out how to do the calculations by choosing the correct lattice from lattices.xml.

Best regards, Daniel

On 05/06/14 00:34, Matthias Troyer wrote:

...

Hi Daniel,

The models are defined in the models.xml XML file, or alternatively your own model file if you write one.

What do you mean by "chain lattice" with "boundary" set to "open"? How did you set it to open?

Matthias

On 22 May 2014, at 22:02, Daniel Guterding <guterding@itp.uni-frankfurt.de> wrote:

...

Dear ALPS users,

I am trying to use the sparsediag application to solve the standard Heisenberg Hamiltonian for various geometries.

So far I could however not figure out which Hamiltonian is actually implemented in the model "spin". Is it J*S1*S2, 2J*S1*S2 or 0.5J*S1*S2 ? Generally it would be good to know where in the code the model Hamiltonians can be found.

My second question is about boundary conditions. First I tried the "chain lattice" and varied the boundary conditions with the keyword "boundary" set to "open" or "closed" respectively, which had no effect on the determined eigenvalues. Is there a difference between using "open chain lattice" and "chain lattice" with "boundary" set to "open" ? I presume the reason for my problems is a simple user error, hence some help would be appreciated.

Best regards, Daniel