it means that one of the orbitals you put in the pseudo configuration, was not computed in the all-electron calculation. Every label that appears in the list of the pseudo-orbitals must also appear in the all-electron list, if for some reason it cannot be computed (i.e. because it is not bound) you can specify negative occupation to skip it.
hth p.s. Molybdenum is a really hard case, I recommend: 1. you use a pseudo from an online library instead, unless you have a really good reason (pslibrary, GBRV,..) 2. you start from an existing input file, i.e. from the pslibrary 3. you brew a big cup of coffee 4. you test it properly On 26/01/18 14:07, [email protected] wrote: > Hello all, > > I am having a hard time understanding this error in ld1.x > > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > Error in routine ld1_setup (4): > no all electron for this ps > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > > I am trying to create a US pseudopotential for Mo. The all-electron > runs ok on its own with the following configuration: > > &input > title='Mo', > zed=42., > rel=1, > config='[Kr] 4d5 5s1', > iswitch=1, > dft='PZ' > / > > However, when I add the cards for PP generation, it fails giving the > previous error. > > I copied the configuration from pslibrary, but I have added the Mo 3d > states in the valence. > > > > > -- Lorenzo Paulatto - Paris _______________________________________________ Pw_forum mailing list [email protected] http://pwscf.org/mailman/listinfo/pw_forum
