Mailman 3 May 2016 - Kwant-discuss

ensuring Hermitian hopping
by Sergey 11 May '16

11 May '16

Dear all, I want to model a random variation of hopping integrals. Simply writing: def hopping(sitei, sitej): return -1 + random.gauss(0,0.2) does not work since this does not ensure that the hopping is Hermitian. Is there a smart way to write something like: def hopping(sitei, sitej): if sitei<sitej: return -1 + random.gauss(0,0.2) else: return Conjugate(hopping(sitej,sitei)) ? Thanks and best wishes, Sergey

5 7

Units of the velocities of the propagating modes
by Michael Beconcini 04 May '16

04 May '16

Dear all, maybe mine is a stupid question, but I’m not sure about the units of measure of the velocities of the propagating modes. I mean, if, e.g., the parameters entering my tight binding system are the lattice constant a, the nearest neighbor hopping energy and the Fermi level, to get the right dimensions of a velocity I need to introduce the Planck constant somewhere. Thus, my question is simply which are the units in which kwant gives back the velocities of the propagating modes? Thank you for your attention! Best regards, Michael Beconcini

2 1

characteristics in nanostructures
by kamal ghosh 04 May '16

04 May '16

Dear All, I find help from the tutorials to study conductance, DOS etc for nano-structures. But when I try to investigate with kwant drain characteristics (Vd verus Id) and transconductance (Vg versus Id) in nanostructures like heterostructure devices of quantum well or QPC, I fall into a mess and cannot arrange my ideas in right direction of proper coding. I find no help file in this direction so that I may proceed. I earnestly request the wise developers and/ any member of my Kwant forum to direct me in this regard. All the best. K.K.Ghosh

3 2

compiling kwant with mumps
by Chong Wang 04 May '16

04 May '16

I am using Arch linux and want to compile kwant against mumps. I installed mumps-seq package from aur and used the following build.conf: [mumps] libraries = zmumps mumps_common pord metis esmumps scotch scotcherr mpiseq include_dirs = /usr/include/mumps-seq However, it does not work well: >>> import kwant /usr/lib/python3.5/site-packages/kwant-1.2.2-py3.5-linux-x86_64.egg/kwant/solvers/default.py:18: RuntimeWarning: MUMPS is not available, SciPy built-in solver will be used as a fallback. Performance can be very poor in this case. "Performance can be very poor in this case.", RuntimeWarning) >>> from kwant.linalg import mumps Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/usr/lib/python3.5/site-packages/kwant-1.2.2-py3.5-linux-x86_64.egg/kwant/linalg/mumps.py", line 17, in <module> from . import _mumps ImportError: ./libzmumps_seq.so: cannot open shared object file: No such file or directory Then I started python in '/usr/lib' where libzmumps_seq.so lives and everything worked well. Does kwant always import zmumps library from the current working dir?

2 1

Aharonov Bohm phase with SOC
by Sumit Ghosh 03 May '16

03 May '16

According to the example ab_ring.py , the AB flux is introduced as the phi parameter of the hopping. In presence of a SOC the hoppings are matrices (spin_orbit.py). # hoppings in x-direction sys[kwant.builder.HoppingKind((1, 0), lat, lat)] = \ -t * sigma_0 - 1j * alpha * sigma_y # hoppings in y-directions sys[kwant.builder.HoppingKind((0, 1), lat, lat)] = \ -t * sigma_0 + 1j * alpha * sigma_x How can I modify the hoppings for such case. I can only figure out that sys[hops_across_cut] has to be a matrix but I don't know which part should be modified for that ! ------------------------------------------------ def fluxphase(site1, site2, phi): return exp(1j * phi) def crosses_branchcut(hop): ix0, iy0 = hop[0].tag # builder.HoppingKind with the argument (1, 0) below # returns hoppings ordered as ((i+1, j), (i, j)) return iy0 < 0 and ix0 == 1 # ix1 == 0 then implied # Modify only those hopings in x-direction that cross the branch cut def hops_across_cut(sys): for hop in kwant.builder.HoppingKind((1, 0), lat, lat)(sys): if crosses_branchcut(hop): yield hop sys[hops_across_cut] = fluxphase -------------------------------------------------- Sumit -- Sumit Ghosh Spintronics Theory Group PSE Division KAUST -- ------------------------------ This message and its contents, including attachments are intended solely for the original recipient. If you are not the intended recipient or have received this message in error, please notify me immediately and delete this message from your computer system. Any unauthorized use or distribution is prohibited. Please consider the environment before printing this email.

2 1

Problem with site color
by T.C. Wu 03 May '16

03 May '16

Dear Kwant developers and users, I am now using Kwant with Python 3.5. After upgrading python from 2.7 to 3.5, I find that site_color=family_colors fails to display different color for A,B sites (such as in http://kwant-project.org/doc/1.0/tutorial/tutorial4). (Note that I cannot see the black and white colors for the A,B sites in tutorial4.) Also, for a 3D system, kwant.plot(sys) fails to display the red dots for the leads (but it did when I was using Python 2.7). I wonder if it is problem with Kwant or it's a problem with my computer. Thanks. Best, Johnny

3 4

Zero velocity mode problem when calculating transmission
by Guodong Yu 02 May '16

02 May '16

Dear Kwant users and admins: I have meet a problem when calculating the transport properties of zigzag-edged graphene nanoribbon. At energy of 0 eV, kwant shows me the error: numpy.linalg.linalg.LinAlgError: singular matrix. After applying the staggered potential, I got this error: RuntimeError: Found a mode with zero or close to zero velocity, because there are two almost flat subbands. I want to know whether kwant can not handle the case when lead exists the flat subbands? How can I solve this problem? Thanks a lot Guodong Yu

2 2

Zero velocity mode problem when calculating transmission
by 于国栋 01 May '16

01 May '16

1 0