[Wien] density matrix

[Tulika Maitra]

Dear Prof. Blaha and Wien users,

I am doing a LDA+U calculation with FLL double counting using U_eff=U-J
as suggested in the userguide. I am using WIEN2k_05 version released in 
october 2005. My system has inversion symmetry and I haven't included the
spin-orbit coupling. The density matrix still has some complex off-diagonal
terms though the imaginary parts exactly cancel out when I calculate the 
partial charges from them. I am giving below the density matrix written 
in scf file in the first iteration and the corresponding partial charges 
(QTL). I have complex off-diagonal elements in the density matrix in all 
the iterations.

-------------------------------------------------------------------------------------------
Density matrix UPUP block, real part.  L= 2
          0.41628  0.00000 -0.09584  0.00000 -0.19734
          0.00000  0.28290  0.00000  0.12019  0.00000
         -0.09584  0.00000  0.47013  0.00000 -0.09584
          0.00000  0.12019  0.00000  0.28290  0.00000
         -0.19734  0.00000 -0.09584  0.00000  0.41628
Density matrix UPUP block, imag part
          0.00000  0.00000 -0.08635  0.00000  0.10447
          0.00000  0.00000  0.00000  0.05663  0.00000
          0.08635  0.00000  0.00000  0.00000 -0.08635
          0.00000 -0.05663  0.00000  0.00000  0.00000
         -0.10447  0.00000  0.08635  0.00000  0.00000

PCS: PARTIAL CHARGES SPHERE =  2 S,P,D,F,PX,PY,PZ,D-Z2,D-X2Y2,D-XY,D-XZ,D-YZ
QTL: 1.0145 2.9123 1.8685 0.0085 0.9708 0.9714 0.9698 0.4701 0.2190 0.6138 0.1628 0.4031
--------------------------------------------------------------------------------------------


Can anyone tell me if this density matrix is reliabe and why it could 
become complex ? Where in the code can I look for the calculation of real 
and imaginary parts of the density matrix?

I know that there was a bug in the U part of the code which has been fixed 
in the newer versions. Since I am using a pretty recent version of WIEN, I 
wonder why I still get complex elements in the density matrix.

Thanking you in advance for any help.

-best regards
Tulika Maitra