Strange output

[GoogleCodeExporter]

I've got as an output file that looks very suspicious....

----
 General Geometry - Free Format

                          U :            5.600000    5.600000
                       t_up :            1.000000
                       t_dn :            1.000000
                      mu_up :            1.800000    1.800000
                      mu_dn :            1.800000    1.800000
             Time slice - L :          160
            Number of sites :           75
                       dtau :            0.050000
                       beta :            8.000000
     Number of warmup sweep :         1000
Number of measurement sweep :        10000
   Frequency of measurement :            1
                Random seed :       743334
 Frequency of recomputing G :           15
Global move number of sites :            0
               Accept count :     67737605
               Reject count :     64190494
    Approximate accept rate :            0.513443
                      gamma :            0.549749
           Type of matrix B : Checkerboard
         Type of matrix HSF : discrete
 Sign of equal time measurements:
                   Avg sign :   0.58000000E-02 +-  0.58600341E-02
                Avg up sign :   0.96000000E-02 +-  0.89977775E-02
                Avg dn sign :   0.19400000E-01 +-  0.56797887E-02
============================================================================
 Equal time measurements:
          Up spin occupancy :   0.58002266E+00 +-             NaN
        Down spin occupancy :   0.63798606E+00 +-             NaN
             <U*N_up*N_dn>  :   0.14816933E+01 +-             NaN
             Kinetic energy :  -0.68437824E+01 +-             NaN
               Total energy :  -0.39620891E+01 +-             NaN
                    Density :   0.12180087E+01 +-             NaN
                Chi_thermal :  -0.81783509E+04 +-             NaN
              Specific heat :   0.62542545E+05 +-             NaN
  XX Ferro structure factor :   0.22955761E+02 +-             NaN
  ZZ Ferro structure factor :  -0.37494198E+00 +-             NaN
           Potential energy :  -0.52873111E+00 +-             NaN
             Hopping energy :  -0.12409423E+01 +-             NaN
           Double occupancy :   0.26458809E+00 +-             NaN
      Magnetization squared :   0.17220813E+00 +-             NaN
     XX AF structure factor :   0.12999555E+02 +-             NaN
  Root Mean Square of XX AF :   0.19781772E+04 +-             NaN
     ZZ AF structure factor :  -0.99905699E+00 +-             NaN
  Root Mean Square of ZZ AF :   0.73465390E+02 +-             NaN
============================================================================


---

Original issue reported on code.google.com by iglovi...@gmail.com on 29 Jul 2014 at 6:49

Attachments:

• Lieb_1406581667.75.out

[GoogleCodeExporter]

Can you disable checkerboard decomposition and do the same run again?

Original comment by cxc639 on 29 Jul 2014 at 6:50

[GoogleCodeExporter]

When I disable it output is meaningful. Attached.

May be it is related to the Issue 46 
https://code.google.com/p/quest-qmc/issues/detail?id=46

May be not.

Original comment by iglovi...@gmail.com on 29 Jul 2014 at 10:56

Attachments:

• Lieb_t1_tp1_u5.6_nx5_ny5.geom
• L160.0_mu1.8_1406660086.34.in
• Lieb_1406660086.34_as.out
• Lieb_1406660086.34_ckb.out

[GoogleCodeExporter]

Actually, when I said meaningful, I was bit brave.

Yes, I do not get NaN, when I do not use checkerboard decomposition,

But I get n_up and n_down that are pretty far from each other:

n_up = 0.39
n_dn = 0.91

And I believe this violates SU(2) symmetry of the Hubbard model.

Original comment by iglovi...@gmail.com on 29 Jul 2014 at 11:20

[GoogleCodeExporter]

I see, why is it going like this. Avg sign ~ 0. We should expect non physical 
values, although it does not explain NaN for the checkerboard decomposition.

Original comment by iglovi...@gmail.com on 29 Jul 2014 at 11:22