by **Daniele Varsano** » Mon Aug 04, 2014 9:29 pm

Dear Samaneh,

I'm not sure I have well understood your question.

The dielectric function is a tensor \eps_ij. The absorption then is given by the trace of the tensor in case of isotropic average (e.g. gas phase molecules or in solution),

(eps_xx+eps_yy+eps_zz)/3.

In Yambo the direction of the field is given by (BLongDir or LongDrXd depending of the used approximation), in output you have the diagonal part of the tensor

corresponding then to the absorption along the chosen direction, if you are interested in different direction of the electric field you need to run different calculations corresponding to different LongDir. This is the same that epsilon.x does (according to the manual, as I'm not an expert of that utility), which gives in output the diagonal part of the dielectric matrix (unless off-diagonal component are requested in input), and I imagine it repeats three calculations for three cartesian direction, in independent particle approximation.

Hope it helps,

Best,

Daniele