On May 17, 2006, at 10:30 AM, Andrea Taroni wrote:

Dear all,

I am in the very early stages of experimenting around with ALPS, with the specific goal of using the software as a benchmark on some classical calculations. I hope you don't mind, but I have a number of basic questions I would like to ask, I'm finding the documentation rather hard to follow at the moment!

1. If I were to add a term to, say, the basic Heisenberg Hamiltonian

what files should I be looking to do this in?

Look at the file on.h which implements the couplings for an XY model, Heisenberg model or O(N) model. Actually the current 1.3a1 pre- release already contains single-ion terms, and anisotropis so please just tell me what you need - it might already be there.

2. In order to recompile the code once this is done, is it just a case

of running gcc on the .C files in the subdirectory of interest (in this case (~/alps-applications-1.2.2/mc/spins) or do I need to recompile the whole of ALPS? If so how is this done?

You would just have to type make install in the subdirectory of interest

3. Is there a protocol in place in order to inform the main developers

that my new (fictitious) Heisenberg code should be in the next version of ALPS?

Send an informal e-mail to the developers list.

Best regards

Matthias Troyer