# [Pw_forum] EPC

ma Yanming yanming_ma at hotmail.com
Mon Apr 28 19:03:20 CEST 2003

Dear Eyvaz

>I have to note that there is some disagreement between
>units. Emax is given in THz, later omega is multiplied
>by 3289.828 in order to convert it to Ry which results
>in zero alpha2F. When I tried to divide both "e" and
>"omega" by 3289.828, I got \lambda \approx 0.19.

I don't think there exist some mismatch between units in lambda.f

Emax is in THz unit. w2(nu) is in Ry**2. So omega is in THz as we know.
It doesn't make sense to divide "e" and "omega" by 3289.828 if e, omega,
and degaussq are in THz unit.

But one thing that I am not quite clear is how to choose degaussq and
ngaussq properly. If we choose the ones used in the EPC calculation, then
the results will be like ghost. If we can not choose those ones then this
will cause the problem that the choice for these values are not unique. For
the purpose of test, I tried to chose different degaussqs and ngaussqs to
get the Spectral function. The shape and smoothness of spectral function is
somehow different. This mean that at this moment PWSCF only can get the
estimated spectral function.

Best Wishes!

Yanming Ma PhD
Steacie Institute for Molecular Sciences,
National Research Councils of Canada.
100 Sussex
K1A 0R6

_________________________________________________________________