error during read of case.qdmftup :: x lapw2 -up -qdmft -c -so :: runsp_lapw -qdmft

[gmm05126]

Dear all,

I have started spin-polarized calculations including spin-orbit [runsp_lapw -so -qdmft 4] slightly modifying Ce-gamma.py script from tutorial; changing Uint, JHund and adding to
S = Solver_HubbardI(Beta = Beta, Uint = Uint, JHund = JHund, l = l, Verbosity=2, UseSpinOrbit = SpinOrbit)
The calculations run up to lapw2 -up -qdmft -c -so; please see the list from :log file
(x) lapw0
(x) lapw1 -up
(x) lapw1 -dn
(x) lapwso -up
(x) lapw2 -up -almd -c -so
(x) lapw2 -dn -almd -c -so
(x) lapw2 -up -qdmft -c -so

where collapsed. Here are few very last output lines:

Chemical_Potential found in 52 iterations :
Total Density = 5.790087;Chemical_Potential = -2.045838
Trace of Density Matrix: {'ud': (5.7900872573605131+6.2552863027620289e-17j)}
forrtl: severe (24): end-of-file during read, unit 32, file /home/wien2k/CALC/case/case.qdmftup

What could be problem? I would like just to add that a test non spin-polarized case without spin-orbit went successfully over one iteration.

Thanks in advance for any hint,
Martin Gmitra
Uni Regensburg

[leopo]

If you want to run spin-polarized SO calculations you have to modify the end of Ce-gamma.py, because the example is written for a non-spin-polarized case and it stores the Hubbard energy correction in Ce-gamma.qdmft. Then lapw2 tries to read it from *.qdmftup/-dn and fails.
So the SO-SP case you write the Hubbard energy correction to two separate files *.qdmftup/-dn as follows

f=open(LDAFilename+'.qdmftup','a')
f.write("%.16f\n"%(correnerg*0.5))
f.close()
f=open(LDAFilename+'.qdmftdn','a')
f.write("%.16f\n"%(correnerg*0.5))
f.close()

[gmm05126]

Dear Leonid,
Thanks for you answer. Now the self-consistency is running. As far I see both the qdmftup and qdmftdn files are same. Could you please comment on it why is that? Is it because of spin-orbit coupling? How this writing to the qdmftup/dn files would look like in case without spin-orbit coupling.
Martin