[Pw_forum] Possibility of obtaining l-dependent potentials from 'atomic'/'ld1'

[Ari P Seitsonen]

Dear Colleagues,

   Does some one know if it is possible to recover the l-dependent 
potentials from the 'atomic'/'ld1' code for educational, or plotting
purpose? This of course only in the norm-conserving, Troullier-Martins 
case. We tried to look up both the options in the input and the items 
written into the .UPF file (after I had already written a Python script 
to extract various fields), but could not find it in either of them. As 
the last resort, I guess that the 'beta' function would be

   beta(r) = [ V_l(r) - V_local(r) ] phi(r) ,

so as phi(r) is given, one could obtain the V_l(r) by inverting the 'beta' 
from this equation? And the normalisation, are the matrix elements D_ij 
contain the (Kleinman-Bylander) factors 1 / < psi(r) | V_l(r) - V_local(r) 
| psi(r) >, or are there some hidden coefficient for normalisations (à 
la 4pi).

   Sorry if this is a FAQ. :(

     Greetings from van Beethoven-city Bonn,

        apsi

-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-
   Ari Paavo Seitsonen / Ari.P.Seitsonen at iki.fi / http://www.iki.fi/~apsi/
     Ecole Normale Supérieure (ENS), Département de Chimie, Paris
     Mobile (F) : +33 789 37 24 25    (CH) : +41 79 71 90 935