Hi all,
I'm trying to do exact diagonalization (with sparsediag) for a spin tube (see lattice description below). Everything goes fine, but energy I get is about twice higher (by absolute value) than expected (I can treat small lattices with other means). This looks suspiciously similar to [1], but that message was left unanswered for two years.
Am I doing anything wrong or is this a bug in sparsediag? For the latter, I can help with debugging. Thank you in advance!
1. https://lists.phys.ethz.ch/pipermail/comp-phys-alps-users/2013/002100.html
Params file:
LATTICE_LIBRARY="./my_lattice.xml"
LATTICE="ladder" W=4
MODEL="spin"; CONSERVED_QUANTUMNUMBERS="Sz"; TRANSLAION_SYMMETRY=false; local_S=1/2; J1=1.0; J2=-0.0625; J3=-0.125; Sz_total=0 {L=6} Sz_total=1 {L=6}
Lattice description is as follows:
<LATTICES>
<LATTICE name="square lattice" dimension="2"> <PARAMETER name="a" default="1"/> <BASIS><VECTOR>a 0</VECTOR><VECTOR>0 a</VECTOR></BASIS> </LATTICE>
<FINITELATTICE name="rectangular half-open" dimension="2"> <LATTICE ref="square lattice"/> <EXTENT dimension="1" size="L"/> <EXTENT dimension="2" size="W"/> <BOUNDARY dimension="1" type="open"/> <BOUNDARY dimension="2" type="periodic"/> </FINITELATTICE>
<UNITCELL name="nnn" dimension="2"> <VERTEX/> <EDGE type="1"> <SOURCE vertex="1"/> <TARGET vertex="1" offset="1 0"/> </EDGE> <EDGE type="2"> <SOURCE vertex="1"/> <TARGET vertex="1" offset="0 2"/> </EDGE> <EDGE type="3"> <SOURCE vertex="1"/> <TARGET vertex="1" offset="0 1"/> </EDGE> </UNITCELL>
<LATTICEGRAPH name="ladder"> <FINITELATTICE ref="rectangular half-open"/> <UNITCELL ref="nnn"/> </LATTICEGRAPH>
</LATTICES>
-- Regards, Valentine Sinitsyn
Hi Valentine
Can you please tell me which energy you would expect for that lattice based on other means? I will then look into it
Matthias
On 24.03.2015, at 14:19, Valentin Sinitsyn valentin.sinitsyn@urfu.ru wrote:
Hi all,
I'm trying to do exact diagonalization (with sparsediag) for a spin tube (see lattice description below). Everything goes fine, but energy I get is about twice higher (by absolute value) than expected (I can treat small lattices with other means). This looks suspiciously similar to [1], but that message was left unanswered for two years.
Am I doing anything wrong or is this a bug in sparsediag? For the latter, I can help with debugging. Thank you in advance!
Params file:
LATTICE_LIBRARY="./my_lattice.xml"
LATTICE="ladder" W=4
MODEL="spin"; CONSERVED_QUANTUMNUMBERS="Sz"; TRANSLAION_SYMMETRY=false; local_S=1/2; J1=1.0; J2=-0.0625; J3=-0.125; Sz_total=0 {L=6} Sz_total=1 {L=6}
Lattice description is as follows:
<LATTICES>
<LATTICE name="square lattice" dimension="2"> <PARAMETER name="a" default="1"/> <BASIS><VECTOR>a 0</VECTOR><VECTOR>0 a</VECTOR></BASIS> </LATTICE>
<FINITELATTICE name="rectangular half-open" dimension="2"> <LATTICE ref="square lattice"/> <EXTENT dimension="1" size="L"/> <EXTENT dimension="2" size="W"/> <BOUNDARY dimension="1" type="open"/> <BOUNDARY dimension="2" type="periodic"/> </FINITELATTICE>
<UNITCELL name="nnn" dimension="2"> <VERTEX/> <EDGE type="1"> <SOURCE vertex="1"/> <TARGET vertex="1" offset="1 0"/> </EDGE> <EDGE type="2"> <SOURCE vertex="1"/> <TARGET vertex="1" offset="0 2"/> </EDGE> <EDGE type="3"> <SOURCE vertex="1"/> <TARGET vertex="1" offset="0 1"/> </EDGE> </UNITCELL>
<LATTICEGRAPH name="ladder"> <FINITELATTICE ref="rectangular half-open"/> <UNITCELL ref="nnn"/> </LATTICEGRAPH>
</LATTICES>
-- Regards, Valentine Sinitsyn
Hi Matthias,
On 25.03.2015 01:38, Matthias Troyer wrote:
Can you please tell me which energy you would expect for that lattice based on other means? I will then look into it
For L=6, as in params below, I'd expect ground state energy around 4.5*J1. However, sparsediag yields -10.1*J1.
Thanks, Valentine
Why do you expect it to be 4.5? You have 20 bonds on which there is a coupling J1.
Matthias
On Mar 24, 2015, at 15:23, Valentin Sinitsyn valentin.sinitsyn@urfu.ru wrote:
Hi Matthias,
On 25.03.2015 01:38, Matthias Troyer wrote: Can you please tell me which energy you would expect for that lattice based on other means? I will then look into it
For L=6, as in params below, I'd expect ground state energy around 4.5*J1. However, sparsediag yields -10.1*J1.
Thanks, Valentine
comp-phys-alps-users@lists.phys.ethz.ch