May 2009 - Comp-phys-alps-users - lists.phys.ethz.ch

Local measurements and correlations with sparsediag
by Cristian Degli Esposti Boschi 14 May '09

14 May '09

Dear all, I would like to study one-site and two-sites correlation functions in certain variants of the 1D Hubbard model. I have tried to compute the local densities (total, up and down) in a small Hubbard chain (10 sites) with open boundary conditions, so that the one-point observables are non-trivial. However, all the local observables return the value 0. The same happens if I try to compute the correlation cdag_up:c_up (or down). The input file is as follows: MODEL="fermion Hubbard"; LATTICE="open chain lattice"; LATTICE_LIBRARY="./lattices.xml"; MODEL_LIBRARY="./models.xml"; CONSERVED_QUANTUMNUMBERS="Nup; Ndown"; MEASURE_LOCAL[nup]=n_up MEASURE_LOCAL[ndown]=n_down MEASURE_LOCAL[ntot]=n MEASURE_CORRELATIONS[c+up_cup]="cdag_up:c_up" MEASURE_CORRELATIONS[c+dn_cdn]="cdag_down:c_down" U=1; {L=10;Nup_total=3;Ndown_total=3} Is there a reason why the correlations are not properly computed and/or a mistake in the input file? Thanks for your time. Cristian -- ___________________________________________________ Cristian Degli Esposti Boschi CNR, CNISM, Unita' di Ricerca di Bologna, c/o Dipartimento di Fisica, Universita' di Bologna viale Berti-Pichat, 6/2, 40127, Bologna, Italia tel. ++39 051 2095114 fax ++39 051 2095113 e-mail: cristian.degliesposti -AT- unibo.it web: http://www.df.unibo.it/fismat/theory ___________________________________________________

2 1

custom generated basis
by khalid hassan 13 May '09

13 May '09

the basis vectors are written by the convention that the down spin are written first according to formula | basis state > = Multiply(i=1 to i=N) ( for_spin_up ) cDag[i,Up_spin] Multiply(i=1 to i=N) ( for_spin_down ) cDag[i,Down_spin] | 0 > Now for two site system half filling 1 up spin and 1 down spin the basis are | basis 1 > = cDag[ 1, up ] cDag[ 2, down ] |0 > | basis 2 > = cDag[ 2, up ] cDag[ 1, down ] |0 > (* Now position cahanged as down spin is first according to convention *) | basis 3 > = cDag[ 1, up ] cDag[ 1, down ] |0 > | basis 4 > = cDag[ 2, up ] cDag[ 2, down ] |0 > Now when we analyticaaly solve for hubbard hamiltonian we get the matrix ((0,0,-t,-t),(0,0,-t,-t),(-t,-t,U,0),(-t,-t,0.U)) but alps give the matrix ((0,1,-1,0),(1,0,0,1),(-1,0,0,-1),(0,1,-1,0)) the matrices are equivalent as long as their eigen values are same. from which we can have eigen values and eigen vectors I want from alps to have basis vectors with this convention. what I see from naked eye if one of the basis vector get a minus sign then the hamiltonian are equivalent. I want the alps to generate basis states according to this convention down spin first and then compute the hamiltonian matrix, otherwise I think program is very powerful. There is any way that we give our basis to the program to compute hamiltonian matrix. best regards Khalid Loane

2 1

correct basis vector
by khalid hassan 13 May '09

13 May '09

Dear alps team I am using the print_numeric program in the model directory to print the basis vectors and hamiltonian matrices for the fermionic hubbard model. Now the problem is that the basis gives the specified combination but not the correct sign. for example for 2 site hubbard model one up spin, one down spin it gives { [ |0 0 > |1 1 > ] [ |0 1 > |1 0 > ] [ |1 0 > |0 1 > ] [ |1 1 > |0 0 > ] } now the 2nd and 3rd row should have opposite sign either 2nd -ve or 3rd negative according to convention. Now based on this set it calculated the matrix [4,4]((0,1,-1,0),(1,0,0,1),(-1,0,0,-1),(0,1,-1,0)) which is wrong as all terms are negative. so I cannot find correct eigen values from this matrix. tell me how can I solve the problem with sign problem of basis states and find correct hamiltonian matrix Best regards khalid loane

2 2

problem with DMRG of ALPS
by Gang Chen 13 May '09

13 May '09

Hi all, I installed alps-1.3.4 with the latest boost 1.38 on my local Laptop. All the functions (e.g classical MC, fulldiag) all work fine except DMRG. When I try to run dmrg parm file in the tutorial, the job got aborted right after it started. Has anyone had similar problem with DMRG? best, -- GANG CHEN Department of Physics University of California Santa Barbara, CA 93106 Office: Broida 6216 Phone: 805-893-5260 Fax: 805-893-3378

2 1

dvanced correlation function
by zhian asadzadeh 10 May '09

10 May '09

Hi Dear, How I can change the code for measuring those two correlation functions? Thanks, Zhian On Thu, Apr 30, 2009 at 1:30 PM, <comp-phys-alps-users-request(a)phys.ethz.ch>wrote: > Send Comp-phys-alps-users mailing list submissions to > comp-phys-alps-users(a)phys.ethz.ch > > To subscribe or unsubscribe via the World Wide Web, visit > https://webmail.phys.ethz.ch/mailman/listinfo/comp-phys-alps-users > or, via email, send a message with subject or body 'help' to > comp-phys-alps-users-request(a)phys.ethz.ch > > You can reach the person managing the list at > comp-phys-alps-users-owner(a)phys.ethz.ch > > When replying, please edit your Subject line so it is more specific > than "Re: Contents of Comp-phys-alps-users digest..." > > > Today's Topics: > > 1. Advanced correlation functions (zhian asadzadeh) > 2. Re: Advanced correlation functions (Ryo IGARASHI) > > > ---------------------------------------------------------------------- > > Message: 1 > Date: Thu, 30 Apr 2009 10:29:31 +0330 > From: zhian asadzadeh <zhian.asadzadeh(a)gmail.com> > Subject: [ALPS-users] Advanced correlation functions > To: comp-phys-alps-users(a)phys.ethz.ch > Message-ID: > <89e5d6760904292359qaa8200byc8414cd745a85ff1(a)mail.gmail.com> > Content-Type: text/plain; charset="iso-8859-1" > > Hi Dear, > > Please let me know can I change the code for measuring just two kinds of > correlations? > (Correlation of first site with other sites and the middle site with other > sites). > > Thanks. > > On Wed, Apr 29, 2009 at 1:30 PM, > <comp-phys-alps-users-request(a)phys.ethz.ch>wrote: > > > Send Comp-phys-alps-users mailing list submissions to > > comp-phys-alps-users(a)phys.ethz.ch > > > > To subscribe or unsubscribe via the World Wide Web, visit > > > https://webmail.phys.ethz.ch/mailman/listinfo/comp-phys-alps-users > > or, via email, send a message with subject or body 'help' to > > comp-phys-alps-users-request(a)phys.ethz.ch > > > > You can reach the person managing the list at > > comp-phys-alps-users-owner(a)phys.ethz.ch > > > > When replying, please edit your Subject line so it is more specific > > than "Re: Contents of Comp-phys-alps-users digest..." > > > > > > Today's Topics: > > > > 1. Advanced correlation functions (zhian asadzadeh) > > 2. Re: Advanced correlation functions (Justin David Peel) > > > > > > ---------------------------------------------------------------------- > > > > Message: 1 > > Date: Tue, 28 Apr 2009 16:26:47 +0430 > > From: zhian asadzadeh <zhian.asadzadeh(a)gmail.com> > > Subject: [ALPS-users] Advanced correlation functions > > To: comp-phys-alps-users(a)phys.ethz.ch > > Message-ID: > > <89e5d6760904280456y43d0cd4cs665a08729a2cac1f(a)mail.gmail.com> > > Content-Type: text/plain; charset="iso-8859-1" > > > > Hi Dear, > > > > Thank you for your guidance, but the mentioned definition for Sy doesn't > > work. > > And for the first question, how I can change the code for measuring those > > two kind of correlations? I mean the correlation of first site with > others > > and middle site with others too. > > > > Thanks, > > Zhian. > > > > > > > > > > On Tue, Apr 28, 2009 at 2:30 PM, > > <comp-phys-alps-users-request(a)phys.ethz.ch>wrote: > > > > > Send Comp-phys-alps-users mailing list submissions to > > > comp-phys-alps-users(a)phys.ethz.ch > > > > > > To subscribe or unsubscribe via the World Wide Web, visit > > > > > https://webmail.phys.ethz.ch/mailman/listinfo/comp-phys-alps-users > > > or, via email, send a message with subject or body 'help' to > > > comp-phys-alps-users-request(a)phys.ethz.ch > > > > > > You can reach the person managing the list at > > > comp-phys-alps-users-owner(a)phys.ethz.ch > > > > > > When replying, please edit your Subject line so it is more specific > > > than "Re: Contents of Comp-phys-alps-users digest..." > > > > > > > > > Today's Topics: > > > > > > 1. Re: lapack support in ALPS (Synge Todo) > > > 2. Re: Advanced correlation functions (Justin David Peel) > > > 3. Re: Comp-phys-alps-users Digest, Vol 37, Issue 18 > > > (Justin David Peel) > > > > > > > > > ---------------------------------------------------------------------- > > > > > > Message: 1 > > > Date: Mon, 27 Apr 2009 19:51:14 +0900 > > > From: Synge Todo <wistaria(a)comp-phys.org> > > > Subject: Re: [ALPS-users] lapack support in ALPS > > > To: comp-phys-alps-users(a)phys.ethz.ch > > > Message-ID: <04B17733-4F07-4867-8B22-4343337160BB(a)comp-phys.org> > > > Content-Type: text/plain; charset=US-ASCII; format=flowed; delsp=yes > > > > > > Dear Fu-Jiun Jiang, > > > > > > On 2009/04/23, at 23:39, Fu-Jiun Jiang ITP wrote: > > > > When install the ALPS with lapack being supported, what command > > > > should be used? In the instruction, it is said > > > > --with-lapack-dir=/path/to/my/lapack/ --with-lapack="-l my_lapack". > > > > Is it true that first one should point to where the lapack library > is. > > > > How about the second one? Thanks. > > > > > > In many cases, lapack installation will be found automatically without > > > any options. If this is not the case, please your > > > system administrator about the linker options to link your program > > > against lapack. For example, if you are linking > > > liblapack.a and libblas.a in /path/to/my/lapack, then the configure > > > options should be > > > > > > --with-lapack-dir=/path/to/my/lapack --with-lapack="-llapack -lblas" > > > > > > Synge > > > > > > > > > > > > ------------------------------ > > > > > > Message: 2 > > > Date: Mon, 27 Apr 2009 11:35:24 -0600 > > > From: Justin David Peel <justin.peel(a)utah.edu> > > > Subject: Re: [ALPS-users] Advanced correlation functions > > > To: "comp-phys-alps-users(a)phys.ethz.ch" > > > <comp-phys-alps-users(a)phys.ethz.ch> > > > Message-ID: > > > < > > 3E5F591958142A40B693A5FC6672D4942624680106(a)C8V1.xds.umail.utah.edu > > > > > > > Content-Type: text/plain; charset="us-ascii" > > > > > > I'm not sure about only measuring those two correlations (you might > have > > to > > > change the code for that), but, as to your last question, you should be > > able > > > to use Sx:Sx because the Sx operator is defined in the models.xml file > in > > > terms of the raising and lowering operators. If you want to measure Sy, > > you > > > would have to add the following lines to the models.xml file: > > > > > > <SITEOPERATOR name="Sy" site="x"> > > > 1/2*(Splus(x)-Sminus(x)) > > > </SITEOPERATOR> > > > > > > because it hasn't been defined as far as I can tell. > > > > > > Good luck, > > > Justin > > > ________________________________________ > > > From: comp-phys-alps-users-bounces(a)phys.ethz.ch [ > > > comp-phys-alps-users-bounces(a)phys.ethz.ch] On Behalf Of zhian > asadzadeh > > [ > > > zhian.asadzadeh(a)gmail.com] > > > Sent: Sunday, April 26, 2009 1:54 AM > > > To: Comp-phys-alps-users(a)phys.ethz.ch > > > Subject: [ALPS-users] Advanced correlation functions > > > > > > Hi All, > > > > > > I am using DMRG for my problem and I am going to measure diagonal and > > > offdiagonal correlation functions. > > > I am interest to obtain the correlation of first site with other sites > > and > > > the middle site with other sites too. > > > The output of DMRG consist of all the combinations which are possible > > but > > > I just want the two correlations mentioned above because the output > will > > be > > > very large ( I am studying the ferrimagnetic chain with 100 sites). > > > can I do that? how? > > > And also can I use Sx:Sx for measuring off diagonal correlation > > functions? > > > > > > Thanks, > > > Zhian. > > > > > > > > > > > > ------------------------------ > > > > > > Message: 3 > > > Date: Mon, 27 Apr 2009 11:42:56 -0600 > > > From: Justin David Peel <justin.peel(a)utah.edu> > > > Subject: Re: [ALPS-users] Comp-phys-alps-users Digest, Vol 37, Issue > > > 18 > > > To: "comp-phys-alps-users(a)phys.ethz.ch" > > > <comp-phys-alps-users(a)phys.ethz.ch> > > > Message-ID: > > > < > > 3E5F591958142A40B693A5FC6672D4942624680107(a)C8V1.xds.umail.utah.edu > > > > > > > Content-Type: text/plain; charset="us-ascii" > > > > > > Someone correct me if I'm wrong, but I believe this isn't very directly > > > feasible. The ground state vector from the exact ground state will be > in > > a > > > different basis and have a different length than the ground state > vector > > > from DMRG (it's a truncated basis after all). > > > ________________________________________ > > > From: comp-phys-alps-users-bounces(a)phys.ethz.ch [ > > > comp-phys-alps-users-bounces(a)phys.ethz.ch] On Behalf Of Gang Chen [ > > > chggst(a)physics.ucsb.edu] > > > Sent: Saturday, April 25, 2009 9:55 AM > > > To: comp-phys-alps-users(a)phys.ethz.ch > > > Subject: Re: [ALPS-users] Comp-phys-alps-users Digest, Vol 37, Issue 18 > > > > > > Thanks, I want to compute the overlap between the exact ground state > > > and the state I generated from other method. > > > > > > best, > > > Gang > > > > > > On Sat, Apr 25, 2009 at 3:00 AM, < > > > comp-phys-alps-users-request(a)phys.ethz.ch<mailto: > > > comp-phys-alps-users-request(a)phys.ethz.ch>> wrote: > > > Send Comp-phys-alps-users mailing list submissions to > > > comp-phys-alps-users(a)phys.ethz.ch<mailto: > > > comp-phys-alps-users(a)phys.ethz.ch> > > > > > > To subscribe or unsubscribe via the World Wide Web, visit > > > > https://webmail.phys.ethz.ch/mailman/listinfo/comp-phys-alps-users > > > or, via email, send a message with subject or body 'help' to > > > comp-phys-alps-users-request(a)phys.ethz.ch<mailto: > > > comp-phys-alps-users-request(a)phys.ethz.ch> > > > > > > You can reach the person managing the list at > > > comp-phys-alps-users-owner(a)phys.ethz.ch<mailto: > > > comp-phys-alps-users-owner(a)phys.ethz.ch> > > > > > > When replying, please edit your Subject line so it is more specific > > > than "Re: Contents of Comp-phys-alps-users digest..." > > > > > > > > > Today's Topics: > > > > > > 1. extract the ground state from DMRG or Exact diagonalization > > > (Gang Chen) > > > 2. Re: extract the ground state from DMRG or Exact > > > diagonalization (Justin David Peel) > > > > > > > > > ---------------------------------------------------------------------- > > > > > > Message: 1 > > > Date: Fri, 24 Apr 2009 14:57:02 -0700 > > > From: Gang Chen <chggst(a)physics.ucsb.edu<mailto: > chggst(a)physics.ucsb.edu > > >> > > > Subject: [ALPS-users] extract the ground state from DMRG or Exact > > > diagonalization > > > To: comp-phys-alps-users(a)phys.ethz.ch<mailto: > > > comp-phys-alps-users(a)phys.ethz.ch> > > > Message-ID: > > > <3924ab970904241457k17ab51b1tbd002d09ae7287e(a)mail.gmail.com > > <mailto: > > > 3924ab970904241457k17ab51b1tbd002d09ae7287e(a)mail.gmail.com>> > > > Content-Type: text/plain; charset="iso-8859-1" > > > > > > Hi All, > > > > > > If I use DMRG or Exact diagonalization to find the ground state of my > > > Hamiltonian. How can > > > I extract the ground state vector (the numerical vector) from my > results? > > > > > > best, > > > > > > -- > > > GANG CHEN > > > Department of Physics > > > University of California > > > Santa Barbara, CA 93106 > > > Office: Broida 6216 > > > Phone: 805-893-5260 > > > Fax: 805-893-3378 > > >

1 0

DMRG, different type of sites
by Remigiusz 07 May '09

07 May '09

Hello all, This is my first post on this mailing list so I would like say hello to all community members. Great thanks to developers for they work on such nice thing like ALPS. I'm currently using alps version 1.34 with boost version 1.34.1. I have following problem. I defined simple graph in lattices.xml file: <GRAPH name="circle1d"> <VERTEX id="1" type="0"/> <VERTEX id="2" type="1"/> <VERTEX id="3" type="2"/> <VERTEX id="4" type="3"/> <VERTEX id="5" type="4"/> <VERTEX id="6" type="5"/> <VERTEX id="7" type="6"/> <VERTEX id="8" type="7"/> <VERTEX id="9" type="8"/> <VERTEX id="10" type="9"/> <EDGE source="1" target="2" type="0"/> <EDGE source="2" target="3" type="0"/> <EDGE source="3" target="4" type="0"/> <EDGE source="4" target="5" type="0"/> <EDGE source="5" target="6" type="0"/> <EDGE source="6" target="7" type="0"/> <EDGE source="7" target="8" type="0"/> <EDGE source="8" target="9" type="0"/> <EDGE source="9" target="10" type="0"/> <EDGE source="10" target="1" type="0"/> </GRAPH> next I create parameter file in form: LATTICE="circle1d"; SWEEPS=4; MAXSTATES=100; MODEL="spin"; NUMBER_EIGENVALUES=2; J=1; MEASURE_AVERAGE[Magnetization]=Sz; MEASURE_AVERAGE[Exchange]=exchange; MEASURE_LOCAL[Local magnetization]=Sz; MEASURE_CORRELATIONS[Diagonal spin correlations]=Sz; MEASURE_CORRELATIONS[Offdiagonal spin correlations]="Splus:Sminus"; { h0=cos(0.0); h1=cos(2.0*pi/10); h2=cos(4.0*pi/10); h3=cos(6.0*pi/10); h4=cos(8.0*pi/10); h5=cos(10.0*pi/10); h6=cos(12.0*pi/10); h7=cos(14.0*pi/10); h8=cos(16.0*pi/10); h9=cos(18.0*pi/10); Gamma0=sin(0.0); Gamma1=sin(2.0*pi/10); Gamma2=sin(4.0*pi/10); Gamma3=sin(6.0*pi/10); Gamma4=sin(8.0*pi/10); Gamma5=sin(10.0*pi/10); Gamma6=sin(12.0*pi/10); Gamma7=sin(14.0*pi/10); Gamma8=sin(16.0*pi/10); Gamma9=sin(18.0*pi/10); } When I started this task I get "segmentation fault (core dumped)" error. I notice that when using graph with less then 5 node types I don't have any problem with this. What I do wrong? Is there any restrictions for number of node types ? Great Thanks, Remigiusz Ziemkiewicz

1 0