[Pw_forum] Example07

2011-10-18 Thread Nicola Marzari 
On 10/17/11 10:35 PM, Paolo Giannozzi wrote:
>
> On Oct 17, 2011, at 20:48 , W2AGZ wrote:
>
>> Has such a calculation of the Eliashberg parameters of aluminum
>> using QE been published
>>
> it has been done (by me: the problem was the behavior of Tc under
> pressure), but never
> published, because the result looked quite dull and did not add
> anything to previous, less sophisticated calculation. It was almost
> 10 years ago. I used something like a 24x24x24 MP
> grid (not translated) for the double-delta integration of the
> electron-phonon coefficients,
> maybe even 32x32x32. This is by far the most critical parameter.


Ciao Paolo, Paul!

Interestingly, we also have done it (i.e. Al Tc under isotropic
or isochoric strain).

We did struggle a lot to get results that were well converged
(in isochoric strain you lose a lot of symmetries), but not not
even 24x24x24 was enough to get a good convergence of T_c (i.e.
changes in broadening would affect Tc very significantly, and we
didn't have enough cpu to find the sweet spot of k-points/broadening ).


nicola



--
Prof Nicola Marzari, Chair of Theory and Simulation of Materials, EPFL

[Pw_forum] QHA runtime error

2011-10-18 Thread Seyed Mojtaba Rezaei Sani 
Dear Prof. Giannozzi

Thank you very much. I must understood that.

Yours

On Tue, Oct 18, 2011 at 5:23 PM, Paolo Giannozzi wrote:

>
> On Oct 17, 2011, at 15:46 , Seyed Mojtaba Rezaei Sani wrote:
>
> > Recalculating omega(q) from C(R)
> > Cannot match namelist object name alas444.fc,
>
> in QHA/Examples/Edit_Me, there is a line
>   FC_file='alas444.fc'
> replace it with
>   FC_file="'alas444.fc'"
>
> Alternatively, in the following group of lines:
>   cat >matdyn.init <   
>  amass(1)=26.98,
>  amass(2)=74.922,
>  asr='crystal',
>  flfrc=$FC_file,
>  flfrq='frequency'
>   
>   EOF
> replace line
>  flfrc=$FC_file,
> with
>  flfrc='$FC_file',
>
> P.
> ---
> Paolo Giannozzi, Dept of Chemistry,
> Univ. Udine, via delle Scienze 208, 33100 Udine, Italy
> Phone +39-0432-558216, fax +39-0432-558222
>
>
>
>
> ___
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
>



-- 
==
Seyed Mojtaba Rezaei Sani
Computational Materials Science Research Lab
Physics Department, Isfahan University of Technology
84156-83111 Isfahan, Iran
Tel lab: +98 311 391 3731
Group page: http://cmsgroup.iut.ac.ir/
s.rezaeisani at ph.iut.ac.ir
===
-- next part --
An HTML attachment was scrubbed...
URL: 
http://www.democritos.it/pipermail/pw_forum/attachments/20111018/f45534e7/attachment.htm

[Pw_forum] BFGS ionic relax: maximum number of steps?

2011-10-18 Thread Xijun Wang 
Hi,

Whenever I run a "restart" relaxation of ions, it always terminated after
one bfgs iteration, and tells me,

   The maximum number of steps has been reached.
   End of BFGS Geometry Optimization
   Begin final coordinates
   ...

Even  I set nstep = 1000 and electron_maxstep = 1000

So, I have to copy all coordinates into a new input to run it
"from_scratch", then it will run normally, terminate after a lot of
iterations.

Anything wrong with my input file?
A compilation error? Or it just means that the convergence has been reached?

Regards,
Xijun Wang


The input file:
*


  calculation   = 'relax',
  restart_mode  = 'restart',
  prefix= 'trip',
  outdir= './'
  pseudo_dir= '/RQusagers/wangxiju/espresso/pseudo'
  etot_conv_thr = 1.0E-6
  forc_conv_thr = 1.0D-5
  tstress   = .true. ,
  tprnfor   = .true. ,
  nstep = 1000 ,
  dt = 150
 /

  ibrav = 0, celldm(1) = 19.964957
  nat = 64, ntyp = 2, ecutwfc = 25, ecutrho = 300
  occupations = 'smearing', smearing = 'cold', degauss = 0.03D0
  nspin = 2, tot_magnetization = 2
  nbnd = 340
  /

  diagonalization = 'cg', electron_maxstep = 1000, conv_thr = 1.D-6
  mixing_mode = 'plain', mixing_beta = 0.1D0, mixing_ndim = 16
 /

  ion_dynamics = 'bfgs', upscale = 100.D0
/
CELL_PARAMETERS hexagonal
   0.992781869   0.0  -0.001946740
   0.0   0.588596312   0.0
  -0.254396859   0.0   1.117512248
ATOMIC_SPECIES
 Zr 91.22400 Zr.pbe-nsp-van.UPF
 N  14.00674  N.pbe-van_bm.UPF
ATOMIC_POSITIONS angstrom
  N 6.239044778 1.554381787 1.967061325
  N 1.558801609 1.554381786 9.815800854
..

-- 
Dept. of Chem and Biochem, Concordia University
7141 Sherbrooke Ouest, Montreal, QC, Canada (H4B 1R6)
Tel: 514-848-2424-#5835 (Lab) TA office: SP175.23
-- next part --
An HTML attachment was scrubbed...
URL: 
http://www.democritos.it/pipermail/pw_forum/attachments/20111018/41b31e43/attachment.htm

[Pw_forum] QHA CODE COMPILATION

2011-10-18 Thread Paolo Giannozzi 

On Oct 17, 2011, at 15:24 , bhabya sahoo wrote:

>  Recalculating omega(q) from C(R)
> Cannot match namelist object name ca.fc

see my previous message

P.
---
Paolo Giannozzi, Dept of Chemistry,
Univ. Udine, via delle Scienze 208, 33100 Udine, Italy
Phone +39-0432-558216, fax +39-0432-558222

[Pw_forum] QHA runtime error

2011-10-18 Thread Paolo Giannozzi 

On Oct 17, 2011, at 15:46 , Seyed Mojtaba Rezaei Sani wrote:

> Recalculating omega(q) from C(R)
> Cannot match namelist object name alas444.fc,

in QHA/Examples/Edit_Me, there is a line
   FC_file='alas444.fc'
replace it with
   FC_file="'alas444.fc'"

Alternatively, in the following group of lines:
   cat >matdyn.init <

[Pw_forum] qdel cann't kill the job.

2011-10-18 Thread Hongsheng Zhao 
On 10/18/2011 02:26 AM, Arles V. Gil Rebaza wrote:
> Hi Hongsheng Zhao, you must be like root and try.
>
>  >> qdel -p 
>
> where XXX is the process number.!!

Thanks a lot, the following command do the trick:

node32:~ # qdel -p 283

Another issue, why I must run it as root?

Regards

>
> Arles V. Gil Rebaza
> IFLP - Argentina
>
> 2011/10/17 ?ric Germaneau  >
>
> login to node32 and kill the process doing a kill -9.
>
> On 10/16/2011 11:56 PM, Axel Kohlmeyer wrote:
>> On Sun, Oct 16, 2011 at 11:41 PM, Hongsheng Zhao> yahoo.com.cn>    wrote:
>>> Hi all,
>>>
>>> I use pbs as the queuing management system.  Currently, I meet the
>>> following strange issue, i.e., the   qdel cann't kill the job, for
>>> detail, see the following:
>>>
>>>   zhaohongsheng at node32:~>  qdel 283
>>> qdel: Server could not connect to MOM283.node32.nxu.edu.cn  
>>> 
>>> zhaohongsheng at node32:~>
>>>
>>> Could you please give me some hints, many thanks in advance.
>> a) this really is a question for the torque mailing list.
>> b) your node's hostname seems to be misconfigured
>>  try using qdel from the login node
>>
>> axel
>>
>>
>>> Regards
>>> --
>>> Hongsheng Zhao  >> yahoo.com.cn>
>>> School of Physics and Electrical Information Science,
>>> Ningxia University, Yinchuan 750021, China
>>> ___
>>> Pw_forum mailing list
>>> Pw_forum at pwscf.org  
>>> http://www.democritos.it/mailman/listinfo/pw_forum
>>>
>
> --
> /Be the change you wish to see in the world
> / ? Mahatma Gandhi ?
>
> Dr. ?ric Germaneau
> 
>
> Graduate University of Chinese Academy of Sciences
> College of Physical Sciences
> Yuquan Road 19A
> Beijing 100049
> China
>
> /Please, if possible, don't send me MS Word or PowerPoint attachments
> Why? See: http://www.gnu.org/philosophy/no-word-attachments.html/
>
>
> ___
> Pw_forum mailing list
> Pw_forum at pwscf.org 
> http://www.democritos.it/mailman/listinfo/pw_forum
>
>
>
>
> --
> ###->   Arles V. <-###
>
>
> ___
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum


-- 
Hongsheng Zhao 
School of Physics and Electrical Information Science,
Ningxia University, Yinchuan 750021, China

[Pw_forum] qdel cann't kill the job.

2011-10-18 Thread Hongsheng Zhao 
On 10/17/2011 11:56 AM, Axel Kohlmeyer wrote:
> On Sun, Oct 16, 2011 at 11:41 PM, Hongsheng Zhao  
> wrote:
>> Hi all,
>>
>> I use pbs as the queuing management system.  Currently, I meet the
>> following strange issue, i.e., the   qdel cann't kill the job, for
>> detail, see the following:
>>
>>   zhaohongsheng at node32:~>  qdel 283
>> qdel: Server could not connect to MOM 283.node32.nxu.edu.cn
>> zhaohongsheng at node32:~>
>>
>> Could you please give me some hints, many thanks in advance.
>
> a) this really is a question for the torque mailing list.

Thanks a lot for you hints.  Do you mean this mailing list: 
torqueusers at supercluster.org?

I've just subscribed to it.  Thanks again.

Regards

> b) your node's hostname seems to be misconfigured
>  try using qdel from the login node
>
> axel
>
>
>>
>> Regards
>> --
>> Hongsheng Zhao
>> School of Physics and Electrical Information Science,
>> Ningxia University, Yinchuan 750021, China
>> ___
>> Pw_forum mailing list
>> Pw_forum at pwscf.org
>> http://www.democritos.it/mailman/listinfo/pw_forum
>>
>
>
>


-- 
Hongsheng Zhao 
School of Physics and Electrical Information Science,
Ningxia University, Yinchuan 750021, China

[Pw_forum] iotk error

2011-10-18 Thread Éric Germaneau 
Dear Professor Giannozzi,

thank you for your suggestion.
First using "-in" does not help, secondly I'm sure of my input since it 
works fine on my desktop.
Now I get this message

/FROM IOTK LIBRARY, VERSION 1.2.0
UNRECOVERABLE ERROR (ierr=1)
ERROR IN: iotk_scan_end (iotk_scan.f90:241)
CVS Revision: 1.23
foundl
ERROR IN: iotk_close_read (iotk_files.f90:746)
CVS Revision: 1.20 /

What does that means?
That happen only on this machine. I can run without in problem a scf 
calculation but for a band calculation for instance I get the error 
above. I use gfortran, woth ifort it's even scf calculation fails.
That's too bad I can not use this machine.

Thank you,

?ric.

On 10/14/2011 08:21 AM, Paolo Giannozzi wrote:
> On Thu, 2011-10-13 at 12:05 -0400, ?ric Germaneau wrote:
>
>> When I execute the pp.x (or band calculation) with ifort 10.1  on
>> several nodes it says
>>  pp.x: error while loading shared libraries: libmpi.so.3.2:
>>  cannot open shared object file: No such file or directory
> http://www.quantum-espresso.org/user_guide/node52.html#SECTION000121010
>
>>  But when I run it on the head node only I get
>>   from postproc : error # 1
>>   reading inputpp namelist
> either there is an error in namelist inputpp, or the code doesn't read
> it, for reasons explained in in the above link at point 11.1.0.3,
> or here: http://www.quantum-espresso.org/user_guide/node19.html
>
>>   from read_rho_xml : error # 1
>>   cannot
>>  open ./SCF/vc-relax_tetragonal.save/charge-density.dat file
>>  for reading
> the file it is not there
>
> P.

-- 
/Be the change you wish to see in the world
/ ? Mahatma Gandhi ?

Dr. ?ric Germaneau 
<http://www.linkedin.com/pub/%C3%A9ric-germaneau/30/931/986>

Graduate University of Chinese Academy of Sciences
College of Physical Sciences
Yuquan Road 19A
Beijing 100049
China

/Please, if possible, don't send me MS Word or PowerPoint attachments
Why? See: http://www.gnu.org/philosophy/no-word-attachments.html /

-- next part --
An HTML attachment was scrubbed...
URL: 
http://www.democritos.it/pipermail/pw_forum/attachments/20111018/287eafc5/attachment.htm

[Pw_forum] BN polarization

2011-10-18 Thread Paolo Giannozzi 

On Oct 17, 2011, at 9:40 , swapnil chandratre wrote:

> Does it mean that instead of improving k points, i need to
> improve geometry, or the unit cell dimensions, etc.

it means that you need to improve your understanding of
what you are calculating and how

P.
---
Paolo Giannozzi, Dept of Chemistry,
Univ. Udine, via delle Scienze 208, 33100 Udine, Italy
Phone +39-0432-558216, fax +39-0432-558222

[Pw_forum] iotk error

2011-10-18 Thread Paolo Giannozzi 

On Oct 18, 2011, at 17:27 , ?ric Germaneau wrote:
>
> FROM IOTK LIBRARY, VERSION 1.2.0 [..]
> What does that means?

it means that your compiler is buggy (and this
must be the 36000th times I say this)

> with ifort it's even scf calculation fails.

with recent versions of ifort everything works

---
Paolo Giannozzi, Dept of Chemistry,
Univ. Udine, via delle Scienze 208, 33100 Udine, Italy
Phone +39-0432-558216, fax +39-0432-558222