January 2009 - Comp-phys-alps-users

one solution to compilation issues with GCC
by Jeff Hammond 22 Jan '09

22 Jan '09

I was not able to compile ALPS 1.3.3 with GCC 4.3.1 but had no problems once I installed GCC 4.2.4 and compiled with that instead. Hopefully this information will be of use to others who use a Linux distro which ships with GCC 4.3 and subsequently struggle to compile ALPS. The only downside of this solution is that compiling GCC 4.2.4 takes a few hours with the default set of features. Best, Jeff -- Jeff Hammond The University of Chicago http://home.uchicago.edu/~jhammond/

3 6

uniform susceptibility
by Tadeusz Wasiutynski 22 Jan '09

22 Jan '09

hello, I try to understand all examples in alps-applications. How can I figure out what exactly is in the output as "uniform susceptibility" (say in Ising) and why it is not simply <M2>-<M>2? In which part of the code it is hidden? Tadeusz

2 1

depleted lattice
by Brian Keith 21 Jan '09

21 Jan '09

I have two questions on the depleted lattice. I am using the depleted lattice as defined in the lattices.xml that ALPS provides. I would like to run a simulation with the DEPLETION variable set to something extremely small like .0004. Do I need to make the lattice size huge, on the order of 1000's of spins, to ensure that the lattice will have a reasonable chance of getting a hole in it? I will run the simulation on a spin ladder if that makes a difference. I would also like to simulate a lattice where the DEPLETION variable is modified by the number of holes close to it, i.e. I would like to encourage the simulation to form small clusters of these spinless holes. Is this easily done? -- Brian Keith Dept. of Physics Clark University

2 1

Classical Monte Carlo - ground states
by Prsa Krunoslav 21 Jan '09

21 Jan '09

Dear All, I am running a Classical Monte Carlo simulation (spinsim, ALPS v. 1.3) on a certain frustrated magnetic lattice. I am trying to simulate one material's behaviour in low temperature (T/J~0.01). I have made quick and dirty modifications of the spinsim.h in order to extract the actual snapshots of the simulated system. Via snapshots, I am extracting the system's structure factor and calculating the neutron Bragg peak intensities vs. field. My material is complicated and its field properties (and this, after looking at the results seems to be the case with the structure factors, not so with magnetization) may depend on the ground state and the details of the simulation. I have a following question: If I specify the input file like the following: LATTICE="my lattice" LATTICE_LIBRARY="../../lattices.xml" T=0.01 J=-1 THERMALIZATION=100000 SWEEPS=5000000 UPDATE="local" MODEL="Ising" L=100 W=100 {h=0;} {h=1;} {h=2;} what actually happens? Does ths start three simulations that are completely separate (starting from i.e. random values of the ground state before), or it is the same simulation meaning that here the simulation with h=0 finishes, the field is applied to *that* ground state and then the simulation continues for h=1? I seem to need the latter case. I hope my question is clear, if it is not, I can rephrase and add more details. Best and thanks for the great work, Kruno --------------------------------------- Krunoslav Prsa, Ph. D. Student Laboratory for Neutron Scattering Paul Scherrer Institute and ETH-Zürich CH-5232 Villigen PSI, Switzerland tel: +41 56 310 20 91 mob: +41 76 386 17 99 ----------------------------------------

2 1

spin correaltion
by Fu-Jiun Jiang ITP 21 Jan '09

21 Jan '09

Hello all I am new to ALPS, but I found it very useful! I got a quick question regrding the " spin correlation ". For a usual Heisenberg model, the output file reads like (0,0) -- (0,1) .... and so. What dose the first (second) coordinate stand for? I think my question is as follows, namely, for a square lattice, say 10 by 10 lattices, how the coordinates for these lattice sites are arranged in the output file. Thanks in advance. FJ

2 1

charge correlation for 4*2 sites
by khalid hassan 20 Jan '09

20 Jan '09

Dear all I only want to find the charge correlation function <n:n> for a 4*2 lattice with 2 particles. as this is not the lowest state it did not come in sparsediag result. fulldiag calculates the whole spectrum which takes much time and i do not need. If there can be any setting of parameter that we calculate "n:n" only for this state. e.g Nup=1 Ndown=1 LATTICE_LIBRARY="/temp/ALPS/xml/lattices.xml" MODEL_LIBRARY="/temp/ALPS/xml/models.xml" MODEL="fermion Hubbard" LATTICE="anisotropic square lattice" L=4 W=2 mu=-0.1 CONSERVED_QUANTUMNUMBERS="Nup,Ndown" MEASURE_CORRELATIONS[charge correlation]="n:n" MEASURE_AVERAGE[Total Particles]="n" U=0 {t=1.0} I need a program that calculate only for one particular state without needing to calculate all the spectrum. I am copying my program for information Best Regards Khalid Loane

2 1

Re: [ALPS-users] Expectation value
by Matthias Troyer 14 Jan '09

14 Jan '09

Sy-reflection is a symmetry, and thus there is always the y-reflected state as well which has the same energy. For the Sy-Sy correlations you can just measure S+ - S+, S+ - S- and S- - S- correlations and calculate it from the, without the need for complex arithmetic Matthias On Jan 14, 2009, at 12:56 PM, zhian asadzadeh wrote: > Dear > I don't think so, when we increase the field this is true. and for > tow point correlations <Sy Sy>, I need to define Sy. > Thanks > Zhian > > On Thu, Jan 15, 2009 at 12:20 AM, Matthias Troyer > <troyer(a)phys.ethz.ch> wrote: > Then <Sy> should be 0 by symmetry > > > On Jan 14, 2009, at 12:48 PM, zhian asadzadeh wrote: > >> Hi >> I have an isotropy in Z direction, and the field is in X direction >> Thanks >> Zhian >> >> On Thu, Jan 15, 2009 at 12:15 AM, Matthias Troyer <troyer(a)phys.ethz.ch >> > wrote: >> Is the transverse field in the x or y direction? >> >> >> On Jan 14, 2009, at 12:43 PM, zhian asadzadeh wrote: >> >>> Hi >>> >>> My model doesn't conserve U(1) symmetry. My model is spin chain >>> in transverse magnetic field. if we want to define the Sy we must >>> use complex number i, so can we define complex number in Model.XML >>> file? >>> >>> Thanks, >>> Zhian >>> >>> On Wed, Jan 14, 2009 at 9:54 PM, Matthias Troyer <troyer(a)phys.ethz.ch >>> > wrote: >>> Does your model conserve SU(2) or U(1) symmetry? In that case the >>> expectation value is 0. Otherwise you should be able to measure it >>> >>> Matthias >>> >>> >>> On 30 Dec 2008, at 04:33, zhian asadzadeh wrote: >>> >>> Hi Dears >>> >>> Can i measure the expectation value of the y component of Spin >>> operator (Sy)?( in DMRG ) >>> >>> Thanks >>> >>> Zhian >>> >>> >> >> > >

1 0

Re: [ALPS-users] Expectation value
by Matthias Troyer 14 Jan '09

14 Jan '09

Then <Sy> should be 0 by symmetry On Jan 14, 2009, at 12:48 PM, zhian asadzadeh wrote: > Hi > I have an isotropy in Z direction, and the field is in X direction > Thanks > Zhian > > On Thu, Jan 15, 2009 at 12:15 AM, Matthias Troyer > <troyer(a)phys.ethz.ch> wrote: > Is the transverse field in the x or y direction? > > > On Jan 14, 2009, at 12:43 PM, zhian asadzadeh wrote: > >> Hi >> >> My model doesn't conserve U(1) symmetry. My model is spin chain in >> transverse magnetic field. if we want to define the Sy we must use >> complex number i, so can we define complex number in Model.XML file? >> >> Thanks, >> Zhian >> >> On Wed, Jan 14, 2009 at 9:54 PM, Matthias Troyer >> <troyer(a)phys.ethz.ch> wrote: >> Does your model conserve SU(2) or U(1) symmetry? In that case the >> expectation value is 0. Otherwise you should be able to measure it >> >> Matthias >> >> >> On 30 Dec 2008, at 04:33, zhian asadzadeh wrote: >> >> Hi Dears >> >> Can i measure the expectation value of the y component of Spin >> operator (Sy)?( in DMRG ) >> >> Thanks >> >> Zhian >> >> > >

1 0

Re: [ALPS-users] Expectation value
by Matthias Troyer 14 Jan '09

14 Jan '09

Is the transverse field in the x or y direction? On Jan 14, 2009, at 12:43 PM, zhian asadzadeh wrote: > Hi > > My model doesn't conserve U(1) symmetry. My model is spin chain in > transverse magnetic field. if we want to define the Sy we must use > complex number i, so can we define complex number in Model.XML file? > > Thanks, > Zhian > > On Wed, Jan 14, 2009 at 9:54 PM, Matthias Troyer > <troyer(a)phys.ethz.ch> wrote: > Does your model conserve SU(2) or U(1) symmetry? In that case the > expectation value is 0. Otherwise you should be able to measure it > > Matthias > > > On 30 Dec 2008, at 04:33, zhian asadzadeh wrote: > > Hi Dears > > Can i measure the expectation value of the y component of Spin > operator (Sy)?( in DMRG ) > > Thanks > > Zhian > >

1 0

error in instaling alps
by Huan Tran 14 Jan '09

14 Jan '09

Dear all, I am reinstall alps-1.3.3 on my new Ubuntu. During the installation of alps-applications-1.3.3, I got the following message: spinmc_factory.C: In constructor ?AbstractSpinSim<MAT>::AbstractSpinSim(const alps::ProcessList&, const alps::Parameters&, int) [with MAT = DiagMatrix<double, 3>]?: spinmc_factory.C:194: internal compiler error: Segmentation fault Please submit a full bug report, with preprocessed source if appropriate. See <file:///usr/share/doc/gcc-4.3/README.Bugs> for instructions. make[2]: *** [spinmc_factory.o] Error 1 make[2]: Leaving directory `/usr/local/src/alps-applications-1.3.3/mc/spins' make[1]: *** [install-recursive] Error 1 make[1]: Leaving directory `/usr/local/src/alps-applications-1.3.3/mc' make: *** [install-recursive] Error 1 I guess that this comes from boost library. Although I have tried many version, from 1.33.1 to 1.37.0, it does not work. Searching on the web does help, too. Could anybody have a hint for me? Thanks Huan Tran ----------------------------------------------------------------

4 4