Help: Inputting Random Potential on Graphene
Hello, My name is Raymond Blackwell and I am attempting to use Kwant to simulate some experimental data. Our data involves dopant atoms intercalated under graphene, so I would like to simulate the LDOS but with a potential that randomly varies at several points within the system. I am trying to modify the code for the kernel_polynominal_method, but I cannot seem to get a random potential. I have used a similar method to randomly select a few points and modify the hopping parameter, but I am struggling with the seemingly more simple case where the potential varies. The relevant portion of my code is below. def make_syst(r=30, t=-1, a=1): syst = kwant.Builder() lat = kwant.lattice.honeycomb(a, norbs=1) def circle(pos): x, y = pos return x ** 2 + y ** 2 < r ** 2 sites2 = list(syst.sites()) Random_sites= random.choices(sites2, k=10) def onsite(site, salt): return 0.5 * kwant.digest.gauss(repr(Random_sites), salt) syst[lat.shape(circle, (0, 0))] = onsite(bytes(3),bytes(3)) Any help is greatly appreciated! Best, -Raymond
Dear Raymond, I think that the problem comes from the last line of your code.
syst[lat.shape(circle, (0, 0))] = onsite(bytes(3),bytes(3))
calls the function onsite with the 2 arguments you provided and then assign the result to all the sites. What you want to do is
syst[lat.shape(circle, (0, 0))] = onsite
In that case, when iterating over the sites, kwant will call the onsite function for each site. See the documentation for providing the salt parameter, e.g. https://kwant-project.org/doc/1/tutorial/spin_potential_shape in the quantum_well.py example. Best regards, Xavier
Le 12 mai 2022 à 00:26, Blackwell, Raymond <rblackwel@bnl.gov> a écrit :
Hello,
My name is Raymond Blackwell and I am attempting to use Kwant to simulate some experimental data. Our data involves dopant atoms intercalated under graphene, so I would like to simulate the LDOS but with a potential that randomly varies at several points within the system. I am trying to modify the code for the kernel_polynominal_method, but I cannot seem to get a random potential. I have used a similar method to randomly select a few points and modify the hopping parameter, but I am struggling with the seemingly more simple case where the potential varies.
The relevant portion of my code is below.
def make_syst(r=30, t=-1, a=1): syst = kwant.Builder() lat = kwant.lattice.honeycomb(a, norbs=1)
def circle(pos): x, y = pos return x ** 2 + y ** 2 < r ** 2 sites2 = list(syst.sites()) Random_sites= random.choices(sites2, k=10)
def onsite(site, salt): return 0.5 * kwant.digest.gauss(repr(Random_sites), salt)
syst[lat.shape(circle, (0, 0))] = onsite(bytes(3),bytes(3))
Any help is greatly appreciated!
Best, -Raymond
Hi Xavier, I appreciate your help! I updated my code to the following but now I receive an error that traces back to the Random_sites line and says that the list index is out of range. If I attempt to print sites2 then the output is an empty list. This is a bit confusing to me, is it because the system is not yet finalized? Once again thank you so much for your help! def make_syst(r=30, t=-1, a=1): syst = kwant.Builder() lat = kwant.lattice.honeycomb(a, norbs=1) sites2 = list(syst.sites()) Random_sites = random.choices(sites2, k = 10) def onsite(site): return 0.5 * kwant.digest.uniform(repr(Random_sites)) + 0.5 def circle(pos): x, y = pos return x ** 2 + y ** 2 < r ** 2 syst[lat.shape(circle, Random_sites)] = onsite syst[lat.neighbors()] = t syst.eradicate_dangling() return syst From: Xavier Waintal <xavier.waintal@cea.fr> Sent: Thursday, May 12, 2022 2:25 AM To: Blackwell, Raymond <rblackwel@bnl.gov> Cc: kwant-discuss@python.org Subject: Re: [Kwant] Help: Inputting Random Potential on Graphene Dear Raymond, I think that the problem comes from the last line of your code. syst[lat.shape(circle, (0, 0))] = onsite(bytes(3),bytes(3)) calls the function onsite with the 2 arguments you provided and then assign the result to all the sites. What you want to do is syst[lat.shape(circle, (0, 0))] = onsite In that case, when iterating over the sites, kwant will call the onsite function for each site. See the documentation for providing the salt parameter, e.g. https://kwant-project.org/doc/1/tutorial/spin_potential_shape<https://urldefense.com/v3/__https:/kwant-project.org/doc/1/tutorial/spin_potential_shape__;!!P4SdNyxKAPE!BY_Tx3Czr9bNjUVgs-Z_mSOn4Hj52FpSmhTjzfGAyP2tvHrviEg84oyfyg4fzcl-1n0Em6DZCCdCyArUVg0mceoRPw$> in the quantum_well.py example. Best regards, Xavier Le 12 mai 2022 à 00:26, Blackwell, Raymond <rblackwel@bnl.gov<mailto:rblackwel@bnl.gov>> a écrit : Hello, My name is Raymond Blackwell and I am attempting to use Kwant to simulate some experimental data. Our data involves dopant atoms intercalated under graphene, so I would like to simulate the LDOS but with a potential that randomly varies at several points within the system. I am trying to modify the code for the kernel_polynominal_method, but I cannot seem to get a random potential. I have used a similar method to randomly select a few points and modify the hopping parameter, but I am struggling with the seemingly more simple case where the potential varies. The relevant portion of my code is below. def make_syst(r=30, t=-1, a=1): syst = kwant.Builder() lat = kwant.lattice.honeycomb(a, norbs=1) def circle(pos): x, y = pos return x ** 2 + y ** 2 < r ** 2 sites2 = list(syst.sites()) Random_sites= random.choices(sites2, k=10) def onsite(site, salt): return 0.5 * kwant.digest.gauss(repr(Random_sites), salt) syst[lat.shape(circle, (0, 0))] = onsite(bytes(3),bytes(3)) Any help is greatly appreciated! Best, -Raymond
Hi Raymond, Indeed, sites2 is an empty list because when you define this variable, you have not yet added any site to your builder (those are added in the line syst[lat.shape…]= onsite. Even more problematic is the fact that the function onsite(site) does not actually depend on the variable site, therefore it will return the same value on each site. Also, there is one mistake in using lat.shape. So overall, you probably want to - remove the lines that defines sites2 and Random_sites and - change onsite into > def onsite(site): > return 0.5 * kwant.digest.uniform(repr(site)) + 0.5 - and last modify > syst[lat.shape(circle, (0,0))] = onsite (the second argument of lat.shape is one set of coordinates that belongs to the shape) and I think you should be fine. Best regards, Xavier > Le 12 mai 2022 à 16:23, Blackwell, Raymond <rblackwel@bnl.gov> a écrit : > > Hi Xavier, > > I appreciate your help! I updated my code to the following but now I receive an error that traces back to the Random_sites line and says that the list index is out of range. If I attempt to print sites2 then the output is an empty list. This is a bit confusing to me, is it because the system is not yet finalized? Once again thank you so much for your help! > > > def make_syst(r=30, t=-1, a=1): > > syst = kwant.Builder() > lat = kwant.lattice.honeycomb(a, norbs=1) > sites2 = list(syst.sites()) > Random_sites = random.choices(sites2, k = 10) > > def onsite(site): > return 0.5 * kwant.digest.uniform(repr(Random_sites)) + 0.5 > > def circle(pos): > x, y = pos > return x ** 2 + y ** 2 < r ** 2 > > > syst[lat.shape(circle, Random_sites)] = onsite > syst[lat.neighbors()] = t > syst.eradicate_dangling() > > return syst > > From: Xavier Waintal <xavier.waintal@cea.fr> > Sent: Thursday, May 12, 2022 2:25 AM > To: Blackwell, Raymond <rblackwel@bnl.gov> > Cc: kwant-discuss@python.org > Subject: Re: [Kwant] Help: Inputting Random Potential on Graphene > > Dear Raymond, > > I think that the problem comes from the last line of your code. > > syst[lat.shape(circle, (0, 0))] = onsite(bytes(3),bytes(3)) > > calls the function onsite with the 2 arguments you provided and then assign the result to all the sites. > What you want to do is > > syst[lat.shape(circle, (0, 0))] = onsite > > In that case, when iterating over the sites, kwant will call the onsite function for each site. > See the documentation for providing the salt parameter, e.g. https://kwant-project.org/doc/1/tutorial/spin_potential_shape <https://urldefense.com/v3/__https:/kwant-project.org/doc/1/tutorial/spin_potential_shape__;!!P4SdNyxKAPE!BY_Tx3Czr9bNjUVgs-Z_mSOn4Hj52FpSmhTjzfGAyP2tvHrviEg84oyfyg4fzcl-1n0Em6DZCCdCyArUVg0mceoRPw$> in the quantum_well.py example. > > Best regards, > > Xavier > > > > > Le 12 mai 2022 à 00:26, Blackwell, Raymond <rblackwel@bnl.gov <mailto:rblackwel@bnl.gov>> a écrit : > > Hello, > > My name is Raymond Blackwell and I am attempting to use Kwant to simulate some experimental data. Our data involves dopant atoms intercalated under graphene, so I would like to simulate the LDOS but with a potential that randomly varies at several points within the system. I am trying to modify the code for the kernel_polynominal_method, but I cannot seem to get a random potential. I have used a similar method to randomly select a few points and modify the hopping parameter, but I am struggling with the seemingly more simple case where the potential varies. > > The relevant portion of my code is below. > > def make_syst(r=30, t=-1, a=1): > syst = kwant.Builder() > lat = kwant.lattice.honeycomb(a, norbs=1) > > def circle(pos): > x, y = pos > return x ** 2 + y ** 2 < r ** 2 > sites2 = list(syst.sites()) > Random_sites= random.choices(sites2, k=10) > > def onsite(site, salt): > return 0.5 * kwant.digest.gauss(repr(Random_sites), salt) > > syst[lat.shape(circle, (0, 0))] = onsite(bytes(3),bytes(3)) > > Any help is greatly appreciated! > > Best, > -Raymond
participants (2)
-
Blackwell, Raymond -
Xavier Waintal