Dear Synge,
This is exactly what I was trying to do (quite easily with the python tools) when I got your message but your answer is really welcome as a confirmation and for the precisions you add.
I have now another question : as I use a transverse field, I will be interested in studying the magnetization along the x-axis. How could I reach such an observable with the looper code ?
Many thanks,
Jean-David
Le 28 juil. 2010 à 04:10, Synge Todo a écrit :
Dear Jean-David,
By specifying
MEASURE[Correlations] = 1
the correlation functions between all the site paris will be calculated. Then, you can Fourier transform them by yourself afterward.
Or you can specify
USE_SITE_INDICES_AS_TYPES = 1 USE_BOND_INDICES_AS_TYPES = 1
by which you can use different coupling constants and magnetic fields for every site/bond with using a smaller unitcell.
c.f) http://wistaria.comp-phys.org/alps-looper/
Synge
On 2010/06/26, at 14:51, Jean-David Picon wrote:
Hi,
I am using ALPS qmc codes (currently looper because of a non 0 transverse field) with a honneycomb unitcell containing with 24 sites lying themselves on a regular smaller honeycomb lattice.
My question : is there an "automatic" way to make ALPS calculate the structure factors corresponding to the smaller underlying lattice without having to define by hand the N^2 corresponding structure factors.
Let me be more precise : if i define a square unitcell (to simplify) of linear size "a" and a system of (linear) length "Na" and ask for the structure factors, I will get all the S(q) with q=(2pi nx/L,2pi ny/L) and n ranging from 1 to N but if I define a system of only ONE unitcell of length Na, ALPS apparently only calculates the structure factors corresponding to q=(0,0) because the size of the system is 1.
I need to work with a 24 sites unitcell, but I also need to get information about the 24 structure factors.
Many thanks,
Jean-David