Hi Sarosh Ali,

please send your questions to the ALPS user list instead of personay to me. You'll get much faster answers that way.

On 24 Dec 2007, at 13:36, sarosh ali wrote:

I am a student of IIT Kharagpur. I am attempting to numerically solve a exciton-lattice system. I tried to use ALPS, using the spin- less fermion basis with various terms in the hamiltonian defined. I tried to do full-diagonalization for 3 sites chain. However, either our computers hung or I got errors like "cannot get quantumnumber". Probably, we have not set up ALPS properly or the system configuration here is too low.

Maybe your input files have problems. Can you post the files you use and we can try to help.

So, in order to solve the problem, I intend to write a C code myself for arbitrary number of sites on a chain. I do have a matrix diagonalization code. Could you please provide me with an algorithm for how to order the basis elements and how to label them. My basis is a lattice basis for two bands with electrons in the conduction band and holes in the valence band is as follows: | n1,n2,n3....> X | m1,m2,m3.....> , where ni indicates the number of electrons and mi the number of holes at the lattice site i. I donot want to trouble you with my problem. However, if you could provide me with a generic way to generate basis, it would be a great help.

Sorry, I just don't have the time to explain to you how to write a complete exact diagonalization code from scratch - since this is essentially what you ask here.

P.S: I have another question not related to this. How to use Alps for quasi-particles?

What do you mean by quasi-particles? What is your model?

Matthias Troyer