Dear Hosein, You are creating a system with essentially all possible hoppings. These will take a considerable computational cost regardless of what you do. Try limiting the number of hoppings to only a few nearest neighbors per site. Best, Anton On Wed, 14 Aug 2019 at 09:40, Khani Hosein <hoseinkhaniphy@gmail.com> wrote:

Dear developers, I constructed Lat_A and Lat_B for a system. A1,A2=Lat_A.sublattices B1,B2=Lat_B.sublattices I want to set hoppings between Lat_A and Lat_B, and the hoppings are dependent on the site position.Every atom of Lat_A is coupled to all the atoms in Lat_B.

Now I have tried it like this : def coupling(site1,site2): xp1,yp1=site1.pos xp2,yp2=siye2.pos return t*sqrt((xp1-xp2)**2+(yp1-yp2)**2)/d

for i in range(100): for j in range(100): sys[kwant.builder.HoppingKind(( i, j), A1, B1)] = coupling sys[kwant.builder.HoppingKind(( i, j), A1, B2)] = coupling sys[kwant.builder.HoppingKind(( i, j), A2, B1)] = coupling sys[kwant.builder.HoppingKind(( i, j), A2, B2)] = coupling

It can not work because it take too long to do the calculations. I am considering a bilayer system but rotations are considered, for example a twisted bilayer graphene. Could you please help me to solve this problem? Regards, Hosein Khani