ABINIT Discussion Forums

Hello all,

I am trying to understand the Tran-Blaha functional, so I first tested it on an example case. I took the H2 molecule from tutorial 2 and attempted to see how the functional changed the eigenvalues. I set ixc = -208012 and the program ran as if there were no errors, but the output file gives many values of NaN. After a few iterations the Etot and DeltaE all have values of NaN and the difference in forces is equal to 0.000E+00. I was wondering if anyone had any ideas as to why this is happening, or if I am unable to test this molecule with ixc -208012? I am using version 7.03 of ABINIT, and I will attach the bare essentials of my input file.

Thanks Much,
Jesse Watson

Which version of libxc are you using?
Note that there were bugs in the implementation of MGGA functional that have been solved in libxc v.2.0.3 (2014-02-18). See
http://www.tddft.org/programs/octopus/w ... xc_changes