d3q Code

Doc updated

A missing ABS in the mean-free path scattering would produce wrong results when using this feature with sample_dir different from zero

Added the mu_initial input parameter for the calculation of final state

When mu_initial is specified, e_initial is automatically set to the energy of the mu-th eigenvalue, unless specified otherwise. In this case, the additional columns of the file will contain the decomposition over the mode of the final state.

If mu_initial is not set, than e_initial must be specified. Note that in this case, the code will compute the final state decomposition of the spectra weight at e_initial, and the additional columns in the file will include the contribution of each modeto the spectra weight. This is quite different from above!

I.e. if you do not set mu_initial that the column of state NU will be equal to the "total" if computed with mu_initial=NU

Added the possibility to convolute with a pseudo-voigt function (as used in ixs experiments) still to be improved. Convolutions is done by FFT which seems to be very efficient, also an alternative subroutine to convolute in reasl space is available, it only convolutes arounf +/- 10 FWHM of the pseudo-voigt which is usually not enough for the long tails.

This also made me realize that the neutron scattering convolution probably contains some mistakes (to be checked) on top of being hard-coded for SnTe

fix

dgradcorr updated to espresso git

command error in script

Final state decomposition redefined in a symmetric form that preservs q2->q3 change for the coalescence part (the X part is unchanged, except for numerical noise which is potentially reduced)

Added the possibility to plot the final state in 3D via xcrysden using eitehr the xsf grid type or bxsf

Hand-made non-analytical contribution replaced with a call to nonanal, I thin kthis works but needs to be checked further as Simoncelli reported a case where LO-TO splitting with causes negative frequencies