[Pw_forum] error from q2r.x

[feng.zimin at ireq.ca]

Dear QE community,

 

I have recently met an error message from q2r.x and a search for previous related posts didn't help solve it:

 

nc already filled: wrong q grid or wrong nr

 

The q mesh that I used is automatically generated and in the input of scf calculation I have set 

 

ibrav=8

 

Most importantly, q2r.x didn't report such an error in the first half, rather, it reports this only when it's starting to do a2F. 

 

Another note, during the electron-phonon calculation, the computer crashed for twice. I had to use recover=.true. to finish it. My installation of QE is up-to-date.

 

So I'd like to ask:

1. Why such an error occured; does it have anything to do with the restart?

2. From my experience, deleting all the output files and restart the whole calculation seems to avoid such an error, but do we have another choice when the ph.x is interrupted unexpectedly?

3. For this case, ph.x really takes a long time, so, since the q mesh is automatically generated, is it safe to turn off such a check and go directly to the calculations in q2r.x?

 

Thank you for reading and looking forward to hearing from you soon!

 

Sincerely,

 

Feng Zimin

 

----------------------- q2r.in ------------------------

&input
  zasr='simple',  fildyn='lp.dyn', la2f=.true.,flfrc='lp884.fc'
 /

 

----------------------- q2r.out ------------------------

 

(...omitted...)

 

 q=  -0.50000000 -0.24568141  0.20495143
 q=   0.50000000  0.24568141 -0.20495143
  reading force constants from file lp.dyn18
  nqs=            1
 q=  -0.50000000 -0.24568141 -0.40990287

      q-space grid ok, #points =   32

      fft-check success (sum of imaginary terms < 10^-12)

   Preparing gamma for a2F


%%%%%%%%%%%%%%%%%%%
     Error in routine init (1):
      nc already filled: wrong q grid or wrong nr
%%%%%%%%%%%%%%%%%%%

     stopping ...


-------------- next part --------------
A non-text attachment was scrubbed...
Name: winmail.dat
Type: application/ms-tnef
Size: 4825 bytes
Desc: not available
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20131101/b4963406/attachment.bin>