Let me make a comment on this, with the caveat to everyone that Peter & I do not completely agree on this (which is OK). A few subtle, but important points.
With DFT we are looking for a fixed point, i.e. the density (and if relevant density matrix or orbital potential) is the same at the start and finish of a single iteration. For simple functionals this is garunteed to exist from the KS theorem. However, with more complex functionals it is not garunteed to exist in all cases. If Wien2k did a variational minimum one would at least get the lowest energy, but a fixed point method is not completely the same. There is a whole world of more complex phenomena which can occur with fixed point methods, and a start is the Wikipedia page on attractors at https://en.m.wikipedia.org/wiki/Attractor . The topic is really rather complex. The take home message is that you have to check not just :DIS, but also other terms to make sure that you really have found the fixed point and are not stuck at some more complex quasi solution. There is an undocumented parameter :MV which can be checked for orbital dependent parameters; it should be less than 1% for a decent solution in my estimation. N.B., the next mixer release will have an additional option "PUSH" which takes some steps to try and reduce some of these issues. Unfortunately if the underlying problem is chaotic (in the mathematical sense) it becomes very complicated and I don't have simple solutions. --- Professor Laurence Marks Department of Materials Science and Engineering Northwestern University http://www.numis.northwestern.edu Corrosion in 4D http://MURI4D.numis.northwestern.edu Co-Editor, Acta Cryst A "Research is to see what everybody else has seen, and to think what nobody else has thought" Albert Szent-Gyorgi On Aug 3, 2015 00:52, "Peter Blaha" <pbl...@theochem.tuwien.ac.at> wrote: > Then the dmat and the scf-files do not fit together ! > > I've noted myself that sometimes the mixer does not properly "mix" the > dmat files, i.e. the new dmat files and the mixed ones disagree and do > not converge to the same numbers. > > To check if this happens in your case: > > compare the dmats of case.scfdmup, case.dmatup/dn_unmixed (they MUST > agree) and case.dmatup/dn. > > If the unmixed dmats contain values close to 1, but the normal (mixed) > dmats don't, you have such a case. Then do: > > save_lapw calc_not_fully_converged > cp case.dmatup_unmixed case.dmatup and dn > another runsp -so > > Of course, l=0 means angular momentum l=0 ("s"). > > > Am 03.08.2015 um 02:01 schrieb Bin Shao: > > The qtl lines for spin up is > > > > :QTL013: 1.0036 2.8012 0.2016 6.8976 0.9330 0.9328 0.9340 0.0384 0.0388 > > 0.0388 0.0436 0.0436 > > Q-s-low E-s-low Q-p-low E-p-low Q-d-low E-d-low Q-f-low > > E-f-low > > > > There is 7 f electrons. Why the occupation number in dmat files is less > > than 7 (about 4)? > > > > The calculation got a magnetic solution with 7 muB and the dmat was from > > a GGA+SOC. > > > > "I usually use the magnetic field only together with l=0." > > > > what's the meaning of l? the quantum number of angular moment? > > > > > > -- > > Bin Shao > > Postdoc > > Department of Physics, Tsinghua University > > Beijing 100084, P. R. China > > Email: binshao1...@gmail.com <mailto:binshao1...@gmail.com> > > > > > > _______________________________________________ > > Wien mailing list > > Wien@zeus.theochem.tuwien.ac.at > > http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien > > SEARCH the MAILING-LIST at: > http://firstname.lastname@example.org/index.html > > > > -- > Peter Blaha > Inst.Materials Chemistry > TU Vienna > Getreidemarkt 9 > A-1060 Vienna > Austria > +43-1-5880115671 > _______________________________________________ > Wien mailing list > Wien@zeus.theochem.tuwien.ac.at > http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien > SEARCH the MAILING-LIST at: > http://email@example.com/index.html >
_______________________________________________ Wien mailing list Wien@zeus.theochem.tuwien.ac.at http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien SEARCH the MAILING-LIST at: http://firstname.lastname@example.org/index.html