You can measure S+S- correlation functions, which may also be useful. From them (and potentially S=S= and S-S- if there is no U(1) symmetry) you can get the SxSx correlation function
On 05 Sep 2016, at 11:27, Michele Dolfi dolfim@phys.ethz.ch wrote:
Dear Nathanael,
Sx correlations are currently supported only when no quantum numbers are conserved.
This has to do with the internal representation of the site operators, which requires a single change in the the quantum number basis.
Best regards, Michele
-- ETH Zurich Michele Dolfi Institute for Theoretical Physics HIT G 32.4 Wolfgang-Pauli-Str. 27 8093 Zurich Switzerland
dolfim@phys.ethz.ch mailto:dolfim@phys.ethz.ch www.itp.phys.ethz.ch http://www.itp.phys.ethz.ch/
+41 44 633 78 56 phone +41 44 633 11 15 fax
On 05 Sep 2016, at 06:44, Natanael de Carvalho Costa <natanael@if.ufrj.br mailto:natanael@if.ufrj.br> wrote:
Dear All,
I'm using ALPS DMRG (MPS) to study the magnetism of XY model in a frustrated chain.
In order to obtain the Sx correlation functions, I add the following command line: MEASURE_CORRELATIONS[Sx_correlation]= Sx
I expected that the output file presented me all <Sx(i) Sx(j)> correlation functions. However, there were only this line in the output file.
<SCALAR_AVERAGE name="Sx_correlation"><MEAN>6.7066684180584175e-317</MEAN></SCALAR_AVERAGE>
Even when I chose to program make only specifics calculations, such as
MEASURE_LOCAL_AT[SxSx_corr] = "Sx:Sx|(0,1),(0,2),(0,3),(0,4)"
the answer I obtained was zero, no matter of the choice of the external parameters, as showed below.
<VECTOR_AVERAGE name="SxSx_corr"> <SCALAR_AVERAGE indexvalue="( 0 ) -- ( 0 )"><MEAN>0</MEAN></SCALAR_AVERAGE> <SCALAR_AVERAGE indexvalue="( 0 ) -- ( 0 )"><MEAN>0</MEAN></SCALAR_AVERAGE> <SCALAR_AVERAGE indexvalue="( 0 ) -- ( 1 )"><MEAN>0</MEAN></SCALAR_AVERAGE> <SCALAR_AVERAGE indexvalue="( 0 ) -- ( 1 )"><MEAN>0</MEAN></SCALAR_AVERAGE> </VECTOR_AVERAGE>
*(the index are not the same in the input and output file because I'm working with a unit cell with 3 sites.)
Owing to these not good results, I did a test, and then I measured the correlation functions <Splus(i) Splus(j)>, <Splus(i) Sminus(j)> and <Sminus(i) Sminus(j)>, as showed below
MEASURE_LOCAL_AT[SxSx_corr1] = "Splus:Splus|(3,18)" MEASURE_LOCAL_AT[SxSx_corr2] = "Splus:Sminus|(3,18)" MEASURE_LOCAL_AT[SxSx_corr3] = "Sminus:Splus|(3,18)" MEASURE_LOCAL_AT[SxSx_corr4] = "Sminus:Sminus|(3,18)"
and I obtained
<VECTOR_AVERAGE name="SxSx_corr1"> <SCALAR_AVERAGE indexvalue="( 1 ) -- ( 6 )"><MEAN>-0.1565028679350004</MEAN></SCALAR_AVERAGE>
</VECTOR_AVERAGE> <VECTOR_AVERAGE name="SxSx_corr2"> <SCALAR_AVERAGE indexvalue="( 1 ) -- ( 6 )"><MEAN>-0.15912792472820933</MEAN></SCALAR_AVERAGE> </VECTOR_AVERAGE>
<VECTOR_AVERAGE name="SxSx_corr3"> <SCALAR_AVERAGE indexvalue="( 1 ) -- ( 6 )"><MEAN>-0.15912792472820814</MEAN></SCALAR_AVERAGE> </VECTOR_AVERAGE>
<VECTOR_AVERAGE name="SxSx_corr4"> <SCALAR_AVERAGE indexvalue="( 1 ) -- ( 6 )"><MEAN>-0.15650286793499901</MEAN></SCALAR_AVERAGE> </VECTOR_AVERAGE>
which means that <SxSx>'s are really nonzero. Actually, these last results are in good agreement with exact diagonalization for a small chain. (Before you asked me, I correctly defined the onsite operator Sx in the model library.)
Then, probably it could be a bug when the program is summing the terms of SxSx correlations.
I really thank you for this great program.
Regards, Natanael C. Costa
Comp-phys-alps-users Mailing List for the ALPS Project http://alps.comp-phys.org/ http://alps.comp-phys.org/
List info: https://lists.phys.ethz.ch//listinfo/comp-phys-alps-users https://lists.phys.ethz.ch//listinfo/comp-phys-alps-users Archive: https://lists.phys.ethz.ch//pipermail/comp-phys-alps-users https://lists.phys.ethz.ch//pipermail/comp-phys-alps-users
Unsubscribe by writing a mail to comp-phys-alps-users-leave@lists.phys.ethz.ch mailto:comp-phys-alps-users-leave@lists.phys.ethz.ch.
Comp-phys-alps-users Mailing List for the ALPS Project http://alps.comp-phys.org/
List info: https://lists.phys.ethz.ch//listinfo/comp-phys-alps-users Archive: https://lists.phys.ethz.ch//pipermail/comp-phys-alps-users
Unsubscribe by writing a mail to comp-phys-alps-users-leave@lists.phys.ethz.ch.