The ABINIT Users Mailing List ( CLOSED )

[Geoffrey Stenuit <geoffrey.stenuit@tyndall.ie>]

Dear Abinit users,

For a given density (XXX_DEN) and a given wavefunction (YYY_WFK), i 
would like to compute the following expectation value:
E=<psi|H|psi> with the Hamiltonian H constructed from the potential 
fixed by the density (XXX_DEN) and psi being the input wavefunction YYY_WFK.
I would like in particular to get the E_n,n=<psi_n|H|psi_n>, at the 
Gamma point, n being the index of the band (YYY_WFK is of course defined 
in my case at the Gamma point).

I tried to compute the expectation value by using the same approach than 
the one used to compute a Band Structure (with "getden -1" and "irdwfk 
1" to read both the density and the wavefunction), but with the number 
of step for the diagonalisation of <psi|H|psi> equal to zero (step 0). 
Unfortunately, the prgm used a new density based on the wavefunction, 
and not the density from XXX_DEN.

Do you know which flag i have to use to avoid this problem ?

Thank you in advance for your help,

Joe

PS I tried also with a number of step not equal to zero and a tolwfr 
very high, so normally exiting the diagonalisation of <psi|H|psi> after 
one step, but unfortunately, the values i got were not correct.