On 29 April 2011 13:41, Matthias Troyer <troyer@phys.ethz.ch> wrote:

On Apr 29, 2011, at 11:29 AM, Oles Zaburannyi wrote:

Hello,

I have question about multi-site terms support in hamiltonian definition.

I found some mention about that in ALPS workshop (2006)

In recent wiki announced that

`Extensions to more complex terms, such as 3- and 4-site terms are in preparation and will be included here as soon as the ALPS libraries support such terms.`

There is possibility to include now such terms?

Not yet - we need to find someone to code it. This will not be hard but someone needs to do it.

Ok, thanks. I have downloaded sources, but estimate my chances as rather small.

Indeed, it will have to be someone who could talk to me for half a day or a day and then we can fix the design. I might get around to doing it myself this summer.

Will arise huge problems in DMRG methods after including such milti-site terms?

No, why should there be huge problems?

I don't know, just asking.

I take a look at lattice and hamiltonian definition. Why we can't define lattice unit with two quantum numbers s1z and s2z

and define s_i s_{i+1} s_{i+2} as follows

for odd $i$ by acting on quantum numbers in single unit cell (siteterm)

and for even $i$ by acting on quantum numbers for two unit cells (bondterm)

Sure, that will work for your particular model

Matthias