[Wien] lapwdm with ROOT option in case.in2

[Masao ARAI]

Dear Dr. P Novak,

The simple example is an non-magnetic insulator.
I calculated cubic SrTiO3 as a test case.

First, I performed SCF calculations with "runsp_lapw".
(Even though the SrTiO3 is non-magnetic, runsp_lapw is necessary
for lapwdm analysis.)

Keeping the SCF eigenvectors, I made case.indm for analysis of Ti 3d.
Then, I performed following steps non-selfconsitently.

   x lapw2 -up
   x lapw2 -dn
   x lapwdm -up
   x lapwdm -dn

For TETRA case, I obtained the following density matrix:

  product Usym*Coupl after densmat
         0.13610  0.00000  0.00000  0.00000  0.04437
         0.00000  0.09173  0.00000  0.00000  0.00000
         0.00000  0.00000  0.18048  0.00000  0.00000
         0.00000  0.00000  0.00000  0.09173  0.00000
         0.04437  0.00000  0.00000  0.00000  0.13610


This seems to be natural.
On the other hand, if we change the option to ROOT, I obtained
the following:

  product Usym*Coupl after densmat
         0.89133  0.00000  0.00000  0.00000  0.01243
         0.00000  0.87891  0.00000  0.00000  0.00000
         0.00000  0.00000  0.90376  0.00000  0.00000
         0.00000  0.00000  0.00000  0.87891  0.00000
         0.01243  0.00000  0.00000  0.00000  0.89133

These values are somehow strange because they indicate that
3d electrons are mostly filled.

The input and output files for this test case are too large to send to 
this mailing list.
If Dr. Novak or someone are interested, I can send files to their 
personal addresses.

Best regards,
--
NIMS
Masao Arai