Dear Paolo, Dear Lorenzo,
I thank you a lot for your answers: I do understand that I should not worry about these warnings. I played a little bit further with the code. I generated a PP using a meaningless input and, as expected :-) , the "Bessel" test gave me a weird eigenvalue spectrum, and the ghost states could also be nicely seen in the logarithmic derivatives. Lorenzo: I unfortunately am a bit confused by your statement about the fact that the construction of the outer state can be made such that this one has zero or one node. Shouldn't the two pseudo-states be orthogonal ? At this point, Paolo: the distinction you make between NC and NC pseudopotentials add a bit more to my confusion. Could you please elaborate more on these subtleties or point me to a paper dealing with them ? All the best, Max Paolo Giannozzi wrote: > Lorenzo Paulatto wrote: > > >> If you are putting semicore states in valence, you will have two states >> (let's call >> > > them inner and outer) with the same angular momentum. When you are > generating the PP > > you can either ask for the external state to have 0 nodes (e.g. > setting L=S+1) > >> or one node (L=S+2. Both choices are ok, which one is better depends on the >> > details. > > for norm-conserving PP I don't think you can: the inner one must have > 0 nodes, the outer one 1 > > Paolo > -- *********************************************** Latevi Max LAWSON DAKU Departement de chimie physique Universite de Geneve - Sciences II 30, quai Ernest-Ansermet CH-1211 Geneve 4 Switzerland Tel: (41) 22/379 6548 ++ Fax: (41) 22/379 6103 *********************************************** -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090213/2a8f94b2/attachment-0001.htm
