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

DMRG for Antiferromagnetic spin one half chain
by zhian asadzadeh 15 Dec '08

15 Dec '08

Hi Dears I have got following results:" for open boundary condition" SWEEP=8, MAXSTATES=400, L=200, E0= -0.44220777784 SWEEP=50, MAXSTATES=500, L=200, E0= -0.4422077778446504 SWEEP=8, MAXSTATES=150, L=600, E0= -0.442833001737 SWEEP=8, MAXSTATES= 500, L=600, E0=- 0.44283300822 SWEEP=8, MAXSTATES= 150, L=700, E0= -0.442877 We see that the result still have not got to -0.443147.maybe we increase the size ,getting to this energy but it takes a lot of time . for periodic boundary condition the energy that achieved is lower. I have a question. Are there some techniques for use of this algorithm such that, for example, for 100 or 200 sites we get to the desire energy?

2 1

DMRG for Antiferromagnetic spin one half chain
by zhian asadzadeh 13 Dec '08

13 Dec '08

Hi Dears I have got following results:" for open boundary condition" SWEEP=8, MAXSTATES=400, L=200, E0= -0.44220777784 SWEEP=50, MAXSTATES=500, L=200, E0= -0.4422077778446504 SWEEP=8, MAXSTATES=150, L=600, E0= -0.442833001737 SWEEP=8, MAXSTATES= 500, L=600, E0=- 0.44283300822 SWEEP=8, MAXSTATES= 150, L=700, E0= -0.442877 We see that the result still have not got to -0.443147.maybe we increase the size ,getting to this energy but it takes a lot of time . for periodic boundary condition the energy that achieved is lower. I have a question. Are there some techniques for use of this algorithm such that, for example, for 100 or 200 sites we get to the desire energy? On Sun, Dec 7, 2008 at 6:00 AM, <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. DMRG For Antiferromagnetic spin one half chain (zhian asadzadeh) > 2. Re: DMRG For Antiferromagnetic spin one half chain > (Matthias Troyer) > 3. Re: DMRG For Antiferromagnetic spin one half chain > (Adrian E. Feiguin) > > > ---------------------------------------------------------------------- > > Message: 1 > Date: Sat, 6 Dec 2008 07:58:46 -0500 > From: "zhian asadzadeh" <zhian.asadzadeh(a)gmail.com> > Subject: [ALPS-users] DMRG For Antiferromagnetic spin one half chain > To: comp-phys-alps-users(a)phys.ethz.ch > Message-ID: > <89e5d6760812060458j27784855jaf87d27a63e2b4c1(a)mail.gmail.com> > Content-Type: text/plain; charset="iso-8859-1" > > Hi Dears > Thanks a lot. I have achieved the ground state energy as follow: > SWEEP=8, L=30,MAXSTATES=100, E0= -0.437045191. > SWEEP=8, L=50,MAXSTATES=140, E0= -0.43944220. > SWEEP=4, L=100,MAXSTATES=100, E0= -0.44127739. > SWEEP=4, L=100,MAXSTATES=300, E0= -0.44127739. > SWEEP=4, L=300,MAXSTATES=100, E0= -0.44127739. > The E0= -0.44127739 is the best enegy that i achieve and when i change the > sweep&M&L,the energy doesn't change.I can't improve the result ,also i > wrote > the model & lattice, but have not changed any thing. > Thanks. > > > > > > On Fri, Dec 5, 2008 at 6:00 AM, > <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. DMRG For Antiferromagnetic Spin one half chain (zhian asadzadeh) > > 2. Re: DMRG For Antiferromagnetic Spin one half chain > > (Matthias Troyer) > > 3. Re: DMRG For Antiferromagnetic Spin one half chain > > (Adrian E. Feiguin) > > > > > > ---------------------------------------------------------------------- > > > > Message: 1 > > Date: Thu, 4 Dec 2008 22:35:51 +0330 > > From: "zhian asadzadeh" <zhian.asadzadeh(a)gmail.com> > > Subject: [ALPS-users] DMRG For Antiferromagnetic Spin one half chain > > To: comp-phys-alps-users(a)phys.ethz.ch > > Message-ID: > > <89e5d6760812041105w5556998du756af1c79c8fb32b(a)mail.gmail.com> > > Content-Type: text/plain; charset="iso-8859-1" > > > > Hi Dear > > Thanks for your consideration.I run the program for different sizes and > > have > > increased it to 500 sites.For up to 100 site the result converges but it > is > > not the result that should it be (*0.443147*),the result that have > > achieved > > earlier from Bethe method and in the work of Mr.white that tests the DMRG > > for showing the powerful of this code. > > Thanks > > zhian > > > > > > On Wed, Dec 3, 2008 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. DMRG for antiferromagnetic spin one half chain (zhian asadzadeh) > > > 2. Re: DMRG for antiferromagnetic spin one half chain > > > (Matthias Troyer) > > > > > > > > > ---------------------------------------------------------------------- > > > > > > Message: 1 > > > Date: Tue, 2 Dec 2008 16:43:12 +0330 > > > From: "zhian asadzadeh" <zhian.asadzadeh(a)gmail.com> > > > Subject: [ALPS-users] DMRG for antiferromagnetic spin one half chain > > > To: comp-phys-alps-users(a)phys.ethz.ch > > > Message-ID: > > > <89e5d6760812020513q686feb0bge51a267945421f2(a)mail.gmail.com> > > > Content-Type: text/plain; charset="iso-8859-1" > > > > > > Hi ADRIAN > > > I run DMRG program for antiferromagnetic S=1/2 chain,but the result > that > > i > > > achieve for groun state energy doesn't concur with the result that > has > > > achieved from bethe ansats method (0.443147). > > > I have changed m&sweep& number of sites(to 1000 site) but my result > just > > 2 > > > digit concur with 0.443147. > > > I don't know what must i do. > > > I will be very appreciate if you can guide me. > > > > > > Thanks > > > zhian > > >

1 0

susceptibility
by Tadeusz Wasiutynski 09 Dec '08

09 Dec '08

Hello, What does it mean "susceptibility" or "uniform susceptibility" in classical tutorial? It is not (<M^2>-<|M|>^2)/T . regards Tadeusz -- Tadeusz Wasiutyński Institute of Nuclear Physics PAN Radzikowskiego 152 PL 31-342 Kraków tel. +48 12 662 8251

1 0

DMRG For Antiferromagnetic spin one half chain
by zhian asadzadeh 07 Dec '08

07 Dec '08

Hi Dears Thanks a lot. I have achieved the ground state energy as follow: SWEEP=8, L=30,MAXSTATES=100, E0= -0.437045191. SWEEP=8, L=50,MAXSTATES=140, E0= -0.43944220. SWEEP=4, L=100,MAXSTATES=100, E0= -0.44127739. SWEEP=4, L=100,MAXSTATES=300, E0= -0.44127739. SWEEP=4, L=300,MAXSTATES=100, E0= -0.44127739. The E0= -0.44127739 is the best enegy that i achieve and when i change the sweep&M&L,the energy doesn't change.I can't improve the result ,also i wrote the model & lattice, but have not changed any thing. Thanks. On Fri, Dec 5, 2008 at 6:00 AM, <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. DMRG For Antiferromagnetic Spin one half chain (zhian asadzadeh) > 2. Re: DMRG For Antiferromagnetic Spin one half chain > (Matthias Troyer) > 3. Re: DMRG For Antiferromagnetic Spin one half chain > (Adrian E. Feiguin) > > > ---------------------------------------------------------------------- > > Message: 1 > Date: Thu, 4 Dec 2008 22:35:51 +0330 > From: "zhian asadzadeh" <zhian.asadzadeh(a)gmail.com> > Subject: [ALPS-users] DMRG For Antiferromagnetic Spin one half chain > To: comp-phys-alps-users(a)phys.ethz.ch > Message-ID: > <89e5d6760812041105w5556998du756af1c79c8fb32b(a)mail.gmail.com> > Content-Type: text/plain; charset="iso-8859-1" > > Hi Dear > Thanks for your consideration.I run the program for different sizes and > have > increased it to 500 sites.For up to 100 site the result converges but it is > not the result that should it be (*0.443147*),the result that have > achieved > earlier from Bethe method and in the work of Mr.white that tests the DMRG > for showing the powerful of this code. > Thanks > zhian > > > On Wed, Dec 3, 2008 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. DMRG for antiferromagnetic spin one half chain (zhian asadzadeh) > > 2. Re: DMRG for antiferromagnetic spin one half chain > > (Matthias Troyer) > > > > > > ---------------------------------------------------------------------- > > > > Message: 1 > > Date: Tue, 2 Dec 2008 16:43:12 +0330 > > From: "zhian asadzadeh" <zhian.asadzadeh(a)gmail.com> > > Subject: [ALPS-users] DMRG for antiferromagnetic spin one half chain > > To: comp-phys-alps-users(a)phys.ethz.ch > > Message-ID: > > <89e5d6760812020513q686feb0bge51a267945421f2(a)mail.gmail.com> > > Content-Type: text/plain; charset="iso-8859-1" > > > > Hi ADRIAN > > I run DMRG program for antiferromagnetic S=1/2 chain,but the result that > i > > achieve for groun state energy doesn't concur with the result that has > > achieved from bethe ansats method (0.443147). > > I have changed m&sweep& number of sites(to 1000 site) but my result just > 2 > > digit concur with 0.443147. > > I don't know what must i do. > > I will be very appreciate if you can guide me. > > > > Thanks > > zhian > >

3 2

DMRG For Antiferromagnetic Spin one half chain
by zhian asadzadeh 04 Dec '08

04 Dec '08

Hi Dear Thanks for your consideration.I run the program for different sizes and have increased it to 500 sites.For up to 100 site the result converges but it is not the result that should it be (*0.443147*),the result that have achieved earlier from Bethe method and in the work of Mr.white that tests the DMRG for showing the powerful of this code. Thanks zhian On Wed, Dec 3, 2008 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. DMRG for antiferromagnetic spin one half chain (zhian asadzadeh) > 2. Re: DMRG for antiferromagnetic spin one half chain > (Matthias Troyer) > > > ---------------------------------------------------------------------- > > Message: 1 > Date: Tue, 2 Dec 2008 16:43:12 +0330 > From: "zhian asadzadeh" <zhian.asadzadeh(a)gmail.com> > Subject: [ALPS-users] DMRG for antiferromagnetic spin one half chain > To: comp-phys-alps-users(a)phys.ethz.ch > Message-ID: > <89e5d6760812020513q686feb0bge51a267945421f2(a)mail.gmail.com> > Content-Type: text/plain; charset="iso-8859-1" > > Hi ADRIAN > I run DMRG program for antiferromagnetic S=1/2 chain,but the result that i > achieve for groun state energy doesn't concur with the result that has > achieved from bethe ansats method (0.443147). > I have changed m&sweep& number of sites(to 1000 site) but my result just 2 > digit concur with 0.443147. > I don't know what must i do. > I will be very appreciate if you can guide me. > > Thanks > zhian >

3 2

Re: [ALPS-users] Comp-phys-alps-users Digest, Vol 33, Issue 1
by zhian asadzadeh 03 Dec '08

03 Dec '08

Hi Dear Thanks for your consideration.I run the program for different sizes and have increased it to 500 sites.For up to 100 site the result converges but it is not the result that should it be (*0.443147*),the result that have achieved earlier from Bethe method and in the work of Mr.white that tests the DMRG for showing the powerful of this code. Thanks zhian

1 0

DMRG for antiferromagnetic spin one half chain
by zhian asadzadeh 02 Dec '08

02 Dec '08

Hi ADRIAN I run DMRG program for antiferromagnetic S=1/2 chain,but the result that i achieve for groun state energy doesn't concur with the result that has achieved from bethe ansats method (0.443147). I have changed m&sweep& number of sites(to 1000 site) but my result just 2 digit concur with 0.443147. I don't know what must i do. I will be very appreciate if you can guide me. Thanks zhian

2 1

Re: [ALPS-users] Comp-phys-alps-users Digest, Vol 32, Issue 6
by Akinlolu Akande 28 Nov '08

28 Nov '08

Hi ADRIAN, Thanks for your response.I did what you suggested but got the below errors.I changed the doubles in the dmrg.h file to complex doubles too.The last change I made was to leave the complex doubles in lines 431 and 472 unchanged in the dmrg.C. Any suggestions? Thanks. Akin g++ -DHAVE_CONFIG_H -I. -I. -I../.. -DNDEBUG -I/home/aakande/ALPS/alps-1.3.3/include -I/home/aakande/boost_1_35_0 -I/usr/local/mpich_quad-gcc/include -I/usr/local/mpich_quad-gcc/include/mpi2c++ -I/home/aakande/ALPS/alps-1.3.3/include -I/home/aakande/lp_solve_4.0 -pthread -ftemplate-depth-150 -O3 -DBOOST_DISABLE_ASSERTS -fexceptions -w -MT dmrg.o -MD -MP -MF .deps/dmrg.Tpo -c -o dmrg.o dmrg.C dmrg.C: In member function 'void DMRGTask::init()': dmrg.C:114: error: no match for 'operator=' in '((DMRGTask*)this)->DMRGTask::conserved_quantumnumber_value.std::vector<_Tp, _Alloc>::operator[] [with _Tp = alps::half_integer<short int>, _Alloc = std::allocator<alps::half_integer<short int> >](((long unsigned int)i)) = alps::evaluate [with U = std::complex<double>](((const std::basic_string<char, std::char_traits<char>, std::allocator<char> >&)((const std::basic_string<char, std::char_traits<char>, std::allocator<char> >*)(& std::basic_string<char, std::char_traits<char>, std::allocator<char> >(((const std::basic_string<char, std::char_traits<char>, std::allocator<char> >&)((const std::basic_string<char, std::char_traits<char>, std::allocator<char> >*)(&((alps::StringValue*)((DMRGTask*)this)->DMRGTask::<anonymous>.alps::scheduler::Task::parms.alps::Parameters::operator[](((const std::string&)((const std::string*)(& std::operator+(const std::basic_string<_CharT, _Traits, _Alloc>&, const _CharT*) [with _CharT = char, _Traits = std::char_traits<char>, _Alloc = std::allocator<char>](((const char*)"_total")))))))->alps::lexical_cast_string<std::basic_string<char, std::char_traits<char>, std::allocator<char> > >::<anonymous>))))))), ((const alps::Parameters&)((const alps::Parameters*)(&((DMRGTask*)this)->DMRGTask::<anonymous>.alps::scheduler::Task::parms))))' /home/aakande/ALPS/alps-1.3.3/include/alps/model/half_integer.h:53: note: candidates are: alps::half_integer<short int>& alps::half_integer<short int>::operator=(const alps::half_integer<short int>&) dmrg.C: In member function 'dmtk::BasicOp<std::complex<double> > DMRGTask::create_site_operator(const std::string&, const alps::SiteOperator&, int)': dmrg.C:403: error: no matching function for call to 'fabs(std::complex<double>&)' /usr/include/bits/mathcalls.h:182: note: candidates are: double fabs(double) /usr/lib/gcc/x86_64-redhat-linux/4.1.2/../../../../include/c++/4.1.2/cmath:230: note: long double std::fabs(long double) /usr/lib/gcc/x86_64-redhat-linux/4.1.2/../../../../include/c++/4.1.2/cmath:226: note: float std::fabs(float) dmrg.C: In member function 'void DMRGTask::build_bond_operator(const alps::BondOperator&, const boost::detail::edge_desc_impl<boost::undirected_tag, long unsigned int>&, dmtk::Hami<std::complex<double> >&)': dmrg.C:472: error: template argument 1 is invalid dmrg.C:472: error: expected unqualified-id before '>' token dmrg.C:487: error: 'coords' was not declared in this scope make: *** [dmrg.o] Error 1 > > > As far as I know the DMRG code does not support complex Hamiltonians, > > but you should ask Adrian. > > > > Matthias > > Look at dmrg.C, replace all the "<double>" by "<complex<double> >", and > recompile. This should do the trick for you. > > Let me know how it works for you. > > Saludos, > <ADRIAN> > > >

1 0

Re: [ALPS-users] complex hopping and site dependent parameters
by Akinlolu Akande 26 Nov '08

26 Nov '08

Thanks Prof. Mattias, Yes I mean hopping t in one direction and the complex conjugate in the other.It actually worked pretty well for the half filled case (Maybe for the wrong reason or it's just by coincidence) at least with my comparison with BA solutions.Other fillings are screwed up. Thanks once again. Adrian, would u like to comment on DMRG for complex Hamiltonian?Thanks Akin >* *>* 1.For a 1D Hubbard model with a complex 't' (hopping) parameter, is *>* it ok to make t complex in the input file?I actually did a run where *>* I used the following input file for the full diagonalization: *>* *>* LATTICE="open chain lattice"; *>* LATTICE_LIBRARY="/home/aakande/alps-applications-1.3.3/tutorial/ *>* lattices.xml" *>* SWEEPS=5 *>* MAXSTATES=30 *>* CONSERVED_QUANTUMNUMBERS="Nup,Ndown" *>* { *>* MODEL="fermion Hubbard"; *>* MODEL_LIBRARY="/home/aakande/alps-applications-1.3.3/tutorial/ *>* models.xml" *>* L=4, Nup_total=1, Ndown_total=1 *>* t=exp(I*5) *>* U=4 *>* NUMBER_EIGENVALUES=1 *>* } * This will actually not work since if all hoppings are t the model is not Hermitian. Do you mean a model with hopping t in one direction and the complex conjugate in the other? Yes.It actually worked pretty well for a half filled case (Maybe for a wrong reason or it's by coincidence). >* 2.How does one adjust the input file for a site dependent parameters *>* model?For example, I am interested in a 1D Hubbard model with 10- *>* site chain for which the U term on the first 5 sites are different *>* from the remaining 5 sites (ie 1st 5 sites have U=U1 and the last 5 *>* sites have U=U2) with a different hopping parameter linking the 2 *>* sets of sites.What will change in the input file for such a site *>* dependent model? * You have to build a graph and not a lattice, and give different types to the various sites (vertices) and bonds (edges). You can then specify different t for the bonds and U for the sites. Matthias

2 1

complex hopping and site dependent parameters
by Akinlolu Akande 26 Nov '08

26 Nov '08

Hi all, I am a new user of the ALPS package.I successfully installed the package on our machines and I have run some examples in the tutorial directory. I have 2 questions: 1.For a 1D Hubbard model with a complex 't' (hopping) parameter, is it ok to make t complex in the input file?I actually did a run where I used the following input file for the full diagonalization: LATTICE="open chain lattice"; LATTICE_LIBRARY="/home/aakande/alps-applications-1.3.3/tutorial/lattices.xml" SWEEPS=5 MAXSTATES=30 CONSERVED_QUANTUMNUMBERS="Nup,Ndown" { MODEL="fermion Hubbard"; MODEL_LIBRARY="/home/aakande/alps-applications-1.3.3/tutorial/models.xml" L=4, Nup_total=1, Ndown_total=1 t=exp(I*5) U=4 NUMBER_EIGENVALUES=1 } which actually gave some results but I just want to confirm if the results obtained with such input are ok. 2.How does one adjust the input file for a site dependent parameters model?For example, I am interested in a 1D Hubbard model with 10-site chain for which the U term on the first 5 sites are different from the remaining 5 sites (ie 1st 5 sites have U=U1 and the last 5 sites have U=U2) with a different hopping parameter linking the 2 sets of sites.What will change in the input file for such a site dependent model? Best regards. Akin

3 4