ABINIT Discussion Forums

Dear abinit developers and users,
could you please help me with the following problem:

I performed extensive el-phon calculations (10 Nb atoms, 16x24x6 kpoint-grid, 70 q-vectors) using the procedure described in the latest tutorials (prepgkk 1, prtgkk 1,...). When plugging the results of this lasting-for-weeks job into anaddb, it exits because I just perturbed the first atom (rfatpol 1 1).
My question now is: How can I best make up for the missing perturbations, i.e. with the lowest computational effort?
I guess that a repetition of the same calculation using 'rfatpol 2 10' would be a bad choice, because then abinit could not make use of the knowledge from the response calculations already done.
And, regardless of how I proceed, how can I "force" anaddb to continue the evaluation using the present data in the meanwhile? In the worst case, which line(s) of the source code do I have to change (and how) in order to just ignore the deficit?

Many thanks in advance
Martin

Since nobody seems to have an idea, let me be more concrete: I set up "dummy calculations" (nstep 1, nline 1,..) using "rfatpol 1 10" and "rfdir 1 1 1" for getting full dynamical matrices of my 10 Niobium atomic system. At gamma (qpt 0 0 0) e.g. this results in the following statement in the .out-file:

The list of irreducible perturbations for this q vector is:
1) idir= 1 ipert= 1
2) idir= 2 ipert= 1
3) idir= 3 ipert= 1
4) idir= 1 ipert= 2
5) idir= 2 ipert= 2
6) idir= 3 ipert= 2
7) idir= 1 ipert= 3
8) idir= 2 ipert= 3
9) idir= 3 ipert= 3

================================================================================

A different q-point results in a list of 15 perturbations:

1) idir= 1 ipert= 1
2) idir= 2 ipert= 1
3) idir= 3 ipert= 1
4) idir= 1 ipert= 2
5) idir= 2 ipert= 2
6) idir= 3 ipert= 2
7) idir= 1 ipert= 3
8) idir= 2 ipert= 3
9) idir= 3 ipert= 3
10) idir= 1 ipert= 7
11) idir= 2 ipert= 7
12) idir= 3 ipert= 7
13) idir= 1 ipert= 8
14) idir= 2 ipert= 8
15) idir= 3 ipert= 8

Nevertheless, abinit - at least in the first case, gamma - generates 30 GKK files, i.e. one for every single atom in every direction.

This raises the question: Do I really need these 30 perturbations for every q vector, and what is the meaning of the quoted lists?