[Wien] calculating optical effective mass from BoltzTraP software

[Amy Lazicki]

Dear all,

I would like to calculate the optical effective mass from the averaged 
Fermi velocity over the Fermi surface.  On the mailinglist people have 
claimed that you can get the effective mass from the BoltzTraP software, 
but it is still unclear to me exactly how people are getting this 
information out of it.
At least one person suggested that it can be calculated from the Drude 
model:

sigma = n*(e^2)*tau/m*, with (sigma/tau) taken from column 6 of the 
case.trace file at the fermi level (?) 

Can you tell me if this is how others are doing it?  How could I access 
the effective mass tensor more directly?  For example, there is a file 
generated when I run x_trans and BoltzTraP which is called 
case_mass.dat, but it is always empty (as are the other .dat files 
generated).   Should there be something there?  Also, is the value for n 
simply the number of valence electrons per unit cell, or should this be 
taken from the density of states at the fermi level somehow?

Thank you very much for any advice,
Best,
Amy