On 27 Mar 2009, at 06:08, Fu-Jiun Jiang ITP wrote:

Hello all,

I got 2 questions for the Kondo lattice in the ALPS. The Hamiltonian in the models.xml for Kondo lattice reads

-t*fermion_hop(i,j) + J*exchange(i,j)

1. Does the code only work at half-filling?

2. The second term in the Hamiltonian does not look like the original

one, namely, the original Hamiltonian of the Kondo model contains a term \sum_{i} s_i \cdot S_i, here the s (S) are the spin of the conducting (local) electrons. I wonder what's going on.

If you look at the definition of the exchange() term then it is just what you write

Matthias