Hi, In order to cancel first order errors in d/dx form, and to get better results, each atom is once displaced by +delta and then by -delta. Then Dynamical matrices are calculated. Hence matrix size= 2 times the total degrees of freedom, i.e. twice what you expect.
regards. > Dear PW users, > > After several testing, I found that the phonon density > of state from the PWscf calculations are normalized to > six times the number of atoms in a unit cell. That is > twice of the degrees of freedom. I wonder why don't we > just normalize it to the degrees of freedom or just to > one? > > I would appreciate it very much if someone could give > me an explanation. > > W. YU > > > > > ______________________________________________________ > Click here to donate to the Hurricane Katrina relief effort. > http://store.yahoo.com/redcross-donate3/ > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum >
