August 2009 - Comp-phys-alps-users

Problems with worms??
by Jakub Zakrzewski 27 Oct '10

27 Oct '10

Hello, Sometimes QMC worm code (used on Bose-Hubbard) terminates with the following error: here come first earlier succesful simulations then ... Created run 85 remote on Host ID: 84 Created run 86 remote on Host ID: 85 Created run 87 remote on Host ID: 86 Created run 88 remote on Host ID: 87 All processes have been assigned Checking if Simulation 1 is finished: not yet, next check in 120 seconds ( 0% done). q = -0 state1 = 0 state2 = 0 bond_type = 0 zero matrix element in remove_jump application called MPI_Abort(MPI_COMM_WORLD, -2) - process 67 Could somebody advice what I can do? Best regards Kuba

2 2

biquadratic terms in QMC
by Kazuo Hida 31 Aug '09

31 Aug '09

Dear all Thank you for developping ALPS. I am just trying to use ALPS QMC library. I have a simple question. Does any of ALPS QMC code for quantum spin models support the biquadratic terms (\vec{S}_i\cdot\vec{S}_j)^2 for S \geq 1? With best regards Kazuo Hida Division of Material Science Graduate School of Science and Engineering Saitama University Saitama, 338-8570 JAPAN Tel&FAX:+48-858-3613 e-mail: hida(a)phy.saitama-u.ac.jp Homepage: http://www.phy.saitama-u.ac.jp/~hida/index_e.html

4 4

Re: [ALPS-users] inverse-sqare 1d ising chain (Synge Todo)
by Tadeusz Wasiutynski 24 Aug '09

24 Aug '09

Dear, Is there a method for the simulation of the system with competing interactions e.g dipolar r^{-3} and quadrupolar r^{-5} distance dependence? From the paper Fukui and Todo it is not clear for me. Best regards, Tadeusz On Sat, 22 Aug 2009 12:00:03 +0200, comp-phys-alps-users-request wrote > > Message: 1 > Date: Fri, 21 Aug 2009 19:07:39 +0900 > From: Synge Todo <wistaria(a)comp-phys.org> > Subject: Re: [ALPS-users] inverse-sqare 1d ising chain > To: Andrea Taroni <andrea.taroni(a)fysik.uu.se>, > comp-phys-alps-users(a)phys.ethz.ch > Message-ID: <EDEE50B2-E084-4602-B562-AA97D75E62B3(a)comp-phys.org> > Content-Type: text/plain; charset=US-ASCII; format=flowed; delsp=yes > > Dear Andrea, > > Currently, ALPS lattice and model libraries do not support long- > range interactions, though you can simulate such models by > specifying a graph with N(N-1)/ 2 bonds with different bond types > for different bond lengths (and it will take O(N^2) CPU time). > > Or, I can send you our QMC code used in Fukui and Todo (2009) if you > want. > > Best, > Synge > > On 2009/08/21, at 17:45, Andrea Taroni wrote: > > > Dear All, > > > > Is there a straightforward way to simulate the classical inverse- > > square > > Ising model using ALPS? By straightforward, I mean using the kind of > > recipe that is described in the tutorials section of the APLS > > applications website. > > > > I am interested in comparing any eventual results with those presented > > by Fukui and Todo J. Comp. Phys. 228, 2629 (2009), using a novel > > order-N > > cluster MC method. > > > > Thanks in advance! > > > > Andrea > > End of Comp-phys-alps-users Digest, Vol 41, Issue 3 > *************************************************** Prof. dr hab. Tadeusz Wasiutynski H. Niewodniczanski Institute of Nuclear Physics PAN tel. +48(12)6628251 fax: +48(12)6628251 e-mail: tadeusz.wasiutynski(a)ifj.edu.pl

2 1

inverse-sqare 1d ising chain
by Andrea Taroni 21 Aug '09

21 Aug '09

Dear All, Is there a straightforward way to simulate the classical inverse-square Ising model using ALPS? By straightforward, I mean using the kind of recipe that is described in the tutorials section of the APLS applications website. I am interested in comparing any eventual results with those presented by Fukui and Todo J. Comp. Phys. 228, 2629 (2009), using a novel order-N cluster MC method. Thanks in advance! Andrea

2 1

only some disorder seeds working in sparsediag?
by Bob Gooding 20 Aug '09

20 Aug '09

There may be a problem with sparsediag (my previous coding error notwithstanding). Here is the parameter file used for both sparsediag and fulldiag LATTICE_LIBRARY = "/usr/local/ALPS/lib/xml/lattices.xml"; LATTICE = "inhomogeneous simple cubic lattice"; MODEL_LIBRARY= "/usr/local/ALPS/lib/xml/models.xml"; MODEL = "alternative fermion Hubbard"; CONSERVED_QUANTUMNUMBERS="N,Sz"; MEASURE_LOCAL[Local charge density]=n; t = 1.0; L = 2; U = 9.0; N_total = 4; Sz_total = 0; TRANSLATION_SYMMETRY = false; {DISORDERSEED = 11138; mu = -6.0*(random()-0.5);} Here is the relevant portion of the task1.out file generated by sparsediag - the densities are all zero, but the ground state energy is correct (see below). <EIGENVALUES number="1"> -11.43877071841712 </EIGENVALUES> <EIGENSTATES number="1"> <EIGENSTATE number="0"> <SCALAR_AVERAGE name="Energy"><MEAN>-11.43877071841712</MEAN></SCALAR_AVERAGE> <VECTOR_AVERAGE name="Local charge density"> <SCALAR_AVERAGE indexvalue="( 0,0,0 )"><MEAN>0</MEAN></SCALAR_AVERAGE> <SCALAR_AVERAGE indexvalue="( 0,0,1 )"><MEAN>0</MEAN></SCALAR_AVERAGE> <SCALAR_AVERAGE indexvalue="( 0,1,0 )"><MEAN>0</MEAN></SCALAR_AVERAGE> <SCALAR_AVERAGE indexvalue="( 0,1,1 )"><MEAN>0</MEAN></SCALAR_AVERAGE> <SCALAR_AVERAGE indexvalue="( 1,0,0 )"><MEAN>0</MEAN></SCALAR_AVERAGE> <SCALAR_AVERAGE indexvalue="( 1,0,1 )"><MEAN>0</MEAN></SCALAR_AVERAGE> <SCALAR_AVERAGE indexvalue="( 1,1,0 )"><MEAN>0</MEAN></SCALAR_AVERAGE> <SCALAR_AVERAGE indexvalue="( 1,1,1 )"><MEAN>0</MEAN></SCALAR_AVERAGE> </VECTOR_AVERAGE> Here is a portion of the output from fulldiag - same energy as sparsediag, but now the local charge densities are non zero. <EIGENVALUES number="784"> -11.43877071841714 </EIGENVALUES> ... <EIGENSTATE number="0"> <SCALAR_AVERAGE name="Energy"><MEAN>-11.43877071841714</MEAN></SCALAR_AVERAGE> <VECTOR_AVERAGE name="Local charge density"> <SCALAR_AVERAGE indexvalue="( 0,0,0 )"><MEAN>0.3727847497057824</MEAN></SCALAR_AVERAGE> <SCALAR_AVERAGE indexvalue="( 0,0,1 )"><MEAN>0.8699453754269064</MEAN></SCALAR_AVERAGE> <SCALAR_AVERAGE indexvalue="( 0,1,0 )"><MEAN>0.4528964351309576</MEAN></SCALAR_AVERAGE> <SCALAR_AVERAGE indexvalue="( 0,1,1 )"><MEAN>0.7330138222807983</MEAN></SCALAR_AVERAGE> <SCALAR_AVERAGE indexvalue="( 1,0,0 )"><MEAN>0.6288755119873222</MEAN></SCALAR_AVERAGE> <SCALAR_AVERAGE indexvalue="( 1,0,1 )"><MEAN>0.1885341339328754</MEAN></SCALAR_AVERAGE> <SCALAR_AVERAGE indexvalue="( 1,1,0 )"><MEAN>0.3029849265253151</MEAN></SCALAR_AVERAGE> <SCALAR_AVERAGE indexvalue="( 1,1,1 )"><MEAN>0.4509650450100266</MEAN></SCALAR_AVERAGE> </VECTOR_AVERAGE>

1 0

only some disorder seeds working in sparsediag?
by Bob Gooding 19 Aug '09

19 Aug '09

If I use mu = -W*(rand()-1/2) then this is equivalent in to the Anderson Hubbard model with random on-site energies - a box distribution of strength W. (That is, the grand Hamiltonian H - mu*N becomes the Anderson Hubbard Hamiltonian (H only).) We are completing our work in the fixed particle number canonical ensemble, and my understanding of the conserved quantum numbers option is that if I specify N_total=4 (say) then the calculation is completed in the canonical ensemble. So, perhaps my error (???) is in the use of CONSERVED_QUANTUMNUMBERS="N_total"; N_total=4; > Could it be just that the values of mu are such that the ground state has no particles? > Matthias

2 1

only some disorder seeds working in sparsediag?
by Bob Gooding 19 Aug '09

19 Aug '09

We are using some of our own lattices and models, but the problem we are encountering can be identified using existing lattices ("inhomogeneous square lattice" and "inhomogeneous simple cubic lattice") and different fermion Hubbard models. As one concrete example, with the following (simplified) parameter file, only for some of the runs does one get local density data. Two example outputs are shown. Note that the same thing occurs (of course) if we use two separate parameter files, each one listing a particular disorder seed. Seeds 11113, 11114 work, 11115 does not, etc. Also, the problem exists for N_total=6 and 4 ... So, do constraints exist on valid choices of DISORDERSEED? (Also, since we would like to know the complexion of disorder associated with a given ground state wave function, we would in fact like to know the random number generator being used.) Thanks. ---- LATTICE_LIBRARY = "/usr/local/ALPS/lib/xml/lattices.xml"; LATTICE = "inhomogeneous simple cubic lattice"; MODEL_LIBRARY= "/usr/local/ALPS/lib/xml/models.xml"; MODEL = "alternative fermion Hubbard"; CONSERVED_QUANTUMNUMBERS="Sz_total"; CONSERVED_QUANTUMNUMBERS="N_total"; MEASURE_LOCAL[Local charge density]=n; t = 1.0; U = 2.0; L = 2; Sz_total = 0; N_total = 8; TRANSLATION_SYMMETRY = false; { DISORDERSEED = 11111; mu=-6.0*(random()-0.5); } { DISORDERSEED = 11112; mu=-6.0*(random()-0.5); } some of task1 output: <EIGENVALUES number="1"> -13.57576898386332 </EIGENVALUES> <EIGENSTATES number="1"> <EIGENSTATE number="0"> <SCALAR_AVERAGE name="Energy"><MEAN>-13.57576898386332</MEAN></SCALAR_AVERAGE> <VECTOR_AVERAGE name="Local charge density"> <SCALAR_AVERAGE indexvalue="( 0,0,0 )"><MEAN>1.02945417967653</MEAN></SCALAR_AVERAGE> <SCALAR_AVERAGE indexvalue="( 0,0,1 )"><MEAN>0.9407378693270886</MEAN></SCALAR_AVERAGE> <SCALAR_AVERAGE indexvalue="( 0,1,0 )"><MEAN>0.8692266888738961</MEAN></SCALAR_AVERAGE> <SCALAR_AVERAGE indexvalue="( 0,1,1 )"><MEAN>0.2243604886901401</MEAN></SCALAR_AVERAGE> <SCALAR_AVERAGE indexvalue="( 1,0,0 )"><MEAN>0.9836579726343326</MEAN></SCALAR_AVERAGE> <SCALAR_AVERAGE indexvalue="( 1,0,1 )"><MEAN>0.9533153322138351</MEAN></SCALAR_AVERAGE> <SCALAR_AVERAGE indexvalue="( 1,1,0 )"><MEAN>1.256609332250588</MEAN></SCALAR_AVERAGE> <SCALAR_AVERAGE indexvalue="( 1,1,1 )"><MEAN>1.742638136334122</MEAN></SCALAR_AVERAGE> </VECTOR_AVERAGE> </EIGENSTATE> </EIGENSTATES> some of task2 output: <EIGENVALUES number="1"> -12.3429442715644 </EIGENVALUES> <EIGENSTATES number="1"> <EIGENSTATE number="0"> <SCALAR_AVERAGE name="Energy"><MEAN>-12.3429442715644</MEAN></SCALAR_AVERAGE> <VECTOR_AVERAGE name="Local charge density"> <SCALAR_AVERAGE indexvalue="( 0,0,0 )"><MEAN>0</MEAN></SCALAR_AVERAGE> <SCALAR_AVERAGE indexvalue="( 0,0,1 )"><MEAN>0</MEAN></SCALAR_AVERAGE> <SCALAR_AVERAGE indexvalue="( 0,1,0 )"><MEAN>0</MEAN></SCALAR_AVERAGE> <SCALAR_AVERAGE indexvalue="( 0,1,1 )"><MEAN>0</MEAN></SCALAR_AVERAGE> <SCALAR_AVERAGE indexvalue="( 1,0,0 )"><MEAN>0</MEAN></SCALAR_AVERAGE> <SCALAR_AVERAGE indexvalue="( 1,0,1 )"><MEAN>0</MEAN></SCALAR_AVERAGE> <SCALAR_AVERAGE indexvalue="( 1,1,0 )"><MEAN>0</MEAN></SCALAR_AVERAGE> <SCALAR_AVERAGE indexvalue="( 1,1,1 )"><MEAN>0</MEAN></SCALAR_AVERAGE> </VECTOR_AVERAGE> </EIGENSTATE> </EIGENSTATES>

2 1