From naromero at gmail.com  Mon May  1 22:33:13 2006
From: naromero at gmail.com (Nichols A. Romero)
Date: Mon, 1 May 2006 16:33:13 -0400
Subject: [Pw_forum] Re: vanderwaals
In-Reply-To: <443BD1B3.6090200@ictp.it>
References: <443BD1B3.6090200@ictp.it>
Message-ID: <6ac064b60605011333q212c513fl57ff838656172031@mail.gmail.com>

Sandro (or anyone else for that matter),

Are there any examples files showing how to us the empirical vdW
within CP? I assume there wouldn't be any issue using it in
conjunction with any of the other capabilities of CP, e.g. NVT,
variable-cell MD, BO MD, etc.

On 4/11/06, Sandro Scandolo <scandolo at ictp.it> wrote:
> Nichols,
>
> unfortunately this is not the vdW functional of Langreth et al.
>
> It is a much simpler procedure to correct the DFT forces and stress
> for the missing 1/R^6 tail. It consists of adding to the DFT forces
> and stress a classical interatomic force field that decays at long
> distances like 1/R^6 (with the appropriate C6 coefficients), and
> vanishes at "short" distances, i.e. at interatomic distances where
> one expects DFT to give the correct forces. This is done in a smooth
> fashion, of course. It requires a cut-off distance, a smoothing
> parameter, and the C_ij coefficients for any pair i,j of species.
> The present implementation is system-specific (for a C-H saturated
> system) but can trivially be extended for any system. More details
> can be found in S. Serra et al, Chem Phys Lett 331, 339 (2000)
>
> Regards,
> Sandro
>
> *
>
> From: Nichols A. Romero* <mailto:pw_forum%40pwscf.org>
> /Mon, 3 Apr 2006 10:46:19 -0400/
>
> Hi,
>
> In CP source, there is a file called
> vanderwaals.f90
>
> Is this the vdW functional of D. Langreth et. al?
>
> Bests,
> --
> Nichols A. Romero, Ph.D.
> 1613 Denise Dr. Apt. D
> Forest Hill, MD 21050
> 443-567-8328 (C)
> 410-306-0709 (O)
>
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
>


--
Nichols A. Romero, Ph.D.
1613 Denise Dr. Apt. D
Forest Hill, MD 21050
443-567-8328 (C)
410-306-0709 (O)


From giannozz at nest.sns.it  Tue May  2 17:08:14 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Tue, 2 May 2006 17:08:14 +0200
Subject: [Pw_forum] symmetry problem in vc-relax
In-Reply-To: <BAY19-F72C14AF124FFFD760413DAAB30@phx.gbl>
References: <BAY19-F72C14AF124FFFD760413DAAB30@phx.gbl>
Message-ID: <200605021708.14533.giannozz@nest.sns.it>

On Saturday 29 April 2006 10:38, Ezad Shojaee wrote:

> i am doing vc-relax for TiO2 using symmetry-conserving algorithm
> (cell_dynamics='damp-w' ) but still i have this message:
>
> from checkallsym : error #         5
>   not orthogonal operation

if it happens when you restart, it is a known bug (fixed in the CVS 
version)

Paolo
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From giannozz at nest.sns.it  Wed May  3 11:16:25 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Wed, 3 May 2006 11:16:25 +0200
Subject: [Pw_forum] which k-point
In-Reply-To: <2fd252650604170444w26679ccahccb95d6f9a3cae1d@mail.gmail.com>
References: <2fd252650604170444w26679ccahccb95d6f9a3cae1d@mail.gmail.com>
Message-ID: <200605031116.25977.giannozz@nest.sns.it>

On Monday 17 April 2006 13:44, Dmitry Korotin wrote:

> does the number of current k-point stored in some variable during
> diagonalization of Hamiltonian? (I need to know k-point number 
> inside the call to hpsi procedure).

it is stored in module "wvfct", variable "current_k"

Paolo
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From sodovar at gmail.com  Thu May  4 13:50:16 2006
From: sodovar at gmail.com (Ruslan Minibaev)
Date: Thu, 4 May 2006 15:50:16 +0400
Subject: [Pw_forum] error with pw.x
In-Reply-To: <f530cbd10605040446i6bd80135ud8ac1873572ac800@mail.gmail.com>
References: <f530cbd10605040446i6bd80135ud8ac1873572ac800@mail.gmail.com>
Message-ID: <f530cbd10605040450s69e87b90xaf2a21ab6f32e0bf@mail.gmail.com>

Dear all!

I have the following error:
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
    from d_&_i : error #         1
     nks = 0 for some nodes
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

Could anybody give some hints.

Best regards,
Ruslan Minibaev


From giannozz at nest.sns.it  Thu May  4 14:03:43 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Thu, 4 May 2006 14:03:43 +0200
Subject: [Pw_forum] error with pw.x
In-Reply-To: <f530cbd10605040450s69e87b90xaf2a21ab6f32e0bf@mail.gmail.com>
References: <f530cbd10605040446i6bd80135ud8ac1873572ac800@mail.gmail.com> <f530cbd10605040450s69e87b90xaf2a21ab6f32e0bf@mail.gmail.com>
Message-ID: <200605041403.43432.giannozz@nest.sns.it>

On Thursday 04 May 2006 13:50, Ruslan Minibaev wrote:

>     from d_&_i : error #         1
>      nks = 0 for some nodes

you are parallelizing with more pools than k-points

Paolo
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From ding at sissa.it  Thu May  4 15:47:36 2006
From: ding at sissa.it (Xunlei Ding)
Date: Thu, 04 May 2006 15:47:36 +0200
Subject: [Pw_forum] little suggestion
Message-ID: <445A05F8.2080609@sissa.it>

Dear developer,

In the output file of pwscf of a scf calculation, only the last "total 
energy" has a '!' before it, as
"!    total energy", so we can use grep "!    total energy" to fine it 
easily.
I think it is also convenient to add a "!" also before the last "total 
magnetization" of a scf calculation, isn't it?

Yours,
Ding


From shaposh at isp.nsc.ru  Fri May  5 16:46:25 2006
From: shaposh at isp.nsc.ru (Alexander Shaposhnikov)
Date: Fri, 05 May 2006 21:46:25 +0700
Subject: [Pw_forum] Theoretical XES spectra
Message-ID: <1146840385.3611.18.camel@m00>

I want to use pwscf for the calculation of  theoretical XES spectra
for some materials (SiO2, Si3N4 and others) 
where it is not always possible to use simple model 
of one predominant transition type -for example,
in Si L2,3 spectra (transition to Si 2p states) both Si 3S 
and Si 3D states make contribution. Even worse, some results indicate 
that non-local transitions are also non-negligible.

So i need to explicitly
calculate dipole matrix element for this transition.

Could somebody point me to some theoretical background, specific for
planewaves and pseudopotentials?

Thanks in advance. 




From baroni at sissa.it  Fri May  5 18:29:35 2006
From: baroni at sissa.it (Stefano Baroni)
Date: Fri, 5 May 2006 18:29:35 +0200
Subject: [Pw_forum] Theoretical XES spectra
In-Reply-To: <1146840385.3611.18.camel@m00>
References: <1146840385.3611.18.camel@m00>
Message-ID: <CE6AA35E-8620-4301-8278-1FE4F4D13D51@sissa.it>

Mathieu Taillefumier, Delphine Cabaret, Anne-Marie Flank, and  
Francesco Mauri, "X-ray absorption near-edge structure calculations  
with the pseudopotentials: Application to the K edge in diamond and ?- 
quartz",  Phys. Rev. B 66, 195107 (2002)
Enjoy - SB

On May 5, 2006, at 4:46 PM, Alexander Shaposhnikov wrote:

> I want to use pwscf for the calculation of  theoretical XES spectra
> for some materials (SiO2, Si3N4 and others)
> where it is not always possible to use simple model
> of one predominant transition type -for example,
> in Si L2,3 spectra (transition to Si 2p states) both Si 3S
> and Si 3D states make contribution. Even worse, some results indicate
> that non-local transitions are also non-negligible.
>
> So i need to explicitly
> calculate dipole matrix element for this transition.
>
> Could somebody point me to some theoretical background, specific for
> planewaves and pseudopotentials?
>
> Thanks in advance.
>
>
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum

---
Stefano Baroni - SISSA  &  DEMOCRITOS National Simulation Center -  
Trieste
[+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype)

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: /pipermail/attachments/20060505/c0767bef/attachment.htm 

From giannozz at nest.sns.it  Fri May  5 18:31:03 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Fri, 5 May 2006 18:31:03 +0200
Subject: [Pw_forum] little suggestion
In-Reply-To: <445A05F8.2080609@sissa.it>
References: <445A05F8.2080609@sissa.it>
Message-ID: <200605051831.03154.giannozz@nest.sns.it>

On Thursday 04 May 2006 15:47, Xunlei Ding wrote:

> In the output file of pwscf of a scf calculation, only the last "total
> energy" has a '!' before it, as
> "!    total energy", so we can use grep "!    total energy" to fine it
> easily.
> I think it is also convenient to add a "!" also before the last "total
> magnetization" of a scf calculation, isn't it?

it could be easily done, but it could also break scripts that
use grep to find the total energy. Maybe it is time to think
a little bit on how to write a better-organized output that is 
both human-readable and easy to parse.

Anyway, if you like it, you just have to replace in 
PW/electrons.f90 the line
     IF ( lsda ) WRITE( stdout, 9017 ) magtot, absmag
with (for instance)
     IF ( lsda ) THEN
        IF ( conv_elec ) THEN
           WRITE( stdout, 9017 ) magtot, absmag
        ELSE
           WRITE( stdout, 9019 ) magtot, absmag
        END IF
     END IF

9019 FORMAT(/'!     total magnetization       =', F9.2,' Bohr mag/cell', &
            /'!     absolute magnetization    =', F9.2,' Bohr mag/cell' )

Paolo
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From dkorotin at gmail.com  Sat May  6 06:56:53 2006
From: dkorotin at gmail.com (Dmitry Korotin)
Date: Sat, 6 May 2006 10:56:53 +0600
Subject: [Pw_forum] ld1.x and actinides
Message-ID: <2fd252650605052156t691f5fd8i8bf0e4bf73689237@mail.gmail.com>

Dear all,
is it possible to generate ultrasoft pseudopotential for actinides
usin ld1.x code? Are there any  restrictions for l=3 (f-electrons)
channel?

Thank you in advance,
Dmitry Korotin.


From sodovar at gmail.com  Sat May  6 08:34:30 2006
From: sodovar at gmail.com (Ruslan Minibaev)
Date: Sat, 6 May 2006 10:34:30 +0400
Subject: [Pw_forum] Re: error with pw.x
Message-ID: <f530cbd10605052334v5e1286fas4b43260cf17b5b1@mail.gmail.com>

>>     from d_&_i : error #         1
>>      nks = 0 for some nodes

> you are parallelizing with more pools than k-points

Thank you, Paolo
-------------- next part --------------
An HTML attachment was scrubbed...
URL: /pipermail/attachments/20060506/a321be0e/attachment.htm 

From dkorotin at gmail.com  Sat May  6 09:29:16 2006
From: dkorotin at gmail.com (Dmitry Korotin)
Date: Sat, 6 May 2006 13:29:16 +0600
Subject: [Pw_forum] Pu pseudopotential generation
Message-ID: <2fd252650605060029w314dc2fctc7c8a8f4ff311afc@mail.gmail.com>

Dear ESPRESSO users,
I'm trying to generate ultrasoft pseudopotential for Pu (with ld1.x code).
This is my first attempt to generate some pseudo, so there are could
be some stupid errors in input file.
My input file is:
&input
	title = 'Pu',
	zed = 94.0,
	rel = 2,
	iswitch = 3,
	xmin=-7
	beta=0.1
	rmax=250
	dft='PBE'
	prefix = 'Pu'
	config = '[Rn] 5f5 6d1 7s0'
/
&inputp
	pseudotype = 3
	file_pseudopw = 'Pu.rel.us.upf'
	lloc = -1
	rcloc = 1.5
	rho0 = 0.01
	nlcc = .true.
/
14
6S  1  0  2.00  0.00  1.40  1.40  0.0
6P  2  1  2.00  0.00  1.70  1.70  0.5
6P  2  1  0.00  0.05  1.70  1.70  0.5
6P  2  1  4.00  0.00  1.70  1.70  1.5
6P  2  1  0.00  0.10  1.70  1.70  1.5
6D  3  2  1.00  0.00  2.35  2.35  1.5
6D  3  2  0.00  0.10  2.35  2.35  1.5
6D  3  2  0.00  0.00  2.30  2.30  2.5
6D  3  2  0.00  0.10  2.30  2.30  2.5
5F  4  3  5.00  0.00  1.10  1.10  2.5
5F  4  3  0.00  0.10  1.10  1.10  2.5
5F  4  3  0.00  0.00  1.10  1.10  3.5
5F  4  3  0.00  0.10  1.10  1.10  3.5
7S  5  0  0.00  -.10  3.20  3.20  0.0

And the error is:

      Wfc   6P  rcut= 1.700  Estimated cut-off energy=       61.49 Ry
     l=   1 ns=   2 Node at 0.42494532
      This function has    1 nodes for 0 < r <    1.702

 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
     from gener_pseudo : error #         1
     too many nodes
 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

     stopping ...

Could anyone tell me, is there any critical mistakes in my input and
how should I solve this problem?
Or may be someone have input files for other actinides (as examples)?

Thank you in advance,
Dmitry Korotin.


From akohlmey at vitae.cmm.upenn.edu  Sat May  6 13:19:59 2006
From: akohlmey at vitae.cmm.upenn.edu (Axel Kohlmeyer)
Date: Sat, 6 May 2006 07:19:59 -0400 (EDT)
Subject: [Pw_forum] Pu pseudopotential generation
In-Reply-To: <2fd252650605060029w314dc2fctc7c8a8f4ff311afc@mail.gmail.com>
Message-ID: <Pine.LNX.4.44.0605060706120.11970-100000@localhost.localdomain>

On Sat, 6 May 2006, Dmitry Korotin wrote:

DK> Dear ESPRESSO users,
DK> I'm trying to generate ultrasoft pseudopotential for Pu (with ld1.x code).
DK> This is my first attempt to generate some pseudo, so there are could
DK> be some stupid errors in input file.

dear dmitry,

generation of pseudopotentials is process that - while usually not
complicated in the individual steps - requires a lot of experience
and care to get right for complicated elements. i would strongly
suggest you start with working through some of the simple examples
in the quantum esperesso code, but especially with a tutorial like
the one from the fhi98PP homepage at
http://www.fhi-berlin.mpg.de/th/fhi98md/fhi98PP/
before you create potential files from scratch.

once you've digested that, you'll easily understand the problem
with your input, i.e. what is a good value for the cutoff radius
and how you can find it out.

in general, you are in for some trouble, because DFT is notoriously
bad at getting complicated spin states right for d- and f-electrons.
so i can only recommend to carefully check whether you are getting
useful results at all. i also would recommend, you start testing
with a simpler to create potential type, e.g. a scalar-relativistic
norm-conserving pseudopotential.

best regards,
    axel


DK> My input file is:
DK> &input
DK> 	title = 'Pu',
DK> 	zed = 94.0,
DK> 	rel = 2,
DK> 	iswitch = 3,
DK> 	xmin=-7
DK> 	beta=0.1
DK> 	rmax=250
DK> 	dft='PBE'
DK> 	prefix = 'Pu'
DK> 	config = '[Rn] 5f5 6d1 7s0'
DK> /
DK> &inputp
DK> 	pseudotype = 3
DK> 	file_pseudopw = 'Pu.rel.us.upf'
DK> 	lloc = -1
DK> 	rcloc = 1.5
DK> 	rho0 = 0.01
DK> 	nlcc = .true.
DK> /
DK> 14
DK> 6S  1  0  2.00  0.00  1.40  1.40  0.0
DK> 6P  2  1  2.00  0.00  1.70  1.70  0.5
DK> 6P  2  1  0.00  0.05  1.70  1.70  0.5
DK> 6P  2  1  4.00  0.00  1.70  1.70  1.5
DK> 6P  2  1  0.00  0.10  1.70  1.70  1.5
DK> 6D  3  2  1.00  0.00  2.35  2.35  1.5
DK> 6D  3  2  0.00  0.10  2.35  2.35  1.5
DK> 6D  3  2  0.00  0.00  2.30  2.30  2.5
DK> 6D  3  2  0.00  0.10  2.30  2.30  2.5
DK> 5F  4  3  5.00  0.00  1.10  1.10  2.5
DK> 5F  4  3  0.00  0.10  1.10  1.10  2.5
DK> 5F  4  3  0.00  0.00  1.10  1.10  3.5
DK> 5F  4  3  0.00  0.10  1.10  1.10  3.5
DK> 7S  5  0  0.00  -.10  3.20  3.20  0.0
DK> 
DK> And the error is:
DK> 
DK>       Wfc   6P  rcut= 1.700  Estimated cut-off energy=       61.49 Ry
DK>      l=   1 ns=   2 Node at 0.42494532
DK>       This function has    1 nodes for 0 < r <    1.702
DK> 
DK>  %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
DK>      from gener_pseudo : error #         1
DK>      too many nodes
DK>  %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
DK> 
DK>      stopping ...
DK> 
DK> Could anyone tell me, is there any critical mistakes in my input and
DK> how should I solve this problem?
DK> Or may be someone have input files for other actinides (as examples)?
DK> 
DK> Thank you in advance,
DK> Dmitry Korotin.
DK> _______________________________________________
DK> Pw_forum mailing list
DK> Pw_forum at pwscf.org
DK> http://www.democritos.it/mailman/listinfo/pw_forum
DK> 

-- 
=======================================================================
Axel Kohlmeyer   akohlmey at cmm.chem.upenn.edu   http://www.cmm.upenn.edu
   Center for Molecular Modeling   --   University of Pennsylvania
Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323
tel: 1-215-898-1582,  fax: 1-215-573-6233,  office-tel: 1-215-898-5425
=======================================================================
If you make something idiot-proof, the universe creates a better idiot.




From emenendez at macul.ciencias.uchile.cl  Sun May  7 18:38:38 2006
From: emenendez at macul.ciencias.uchile.cl (Eduardo Ariel Menendez P)
Date: Sun, 7 May 2006 12:38:38 -0400 (CLT)
Subject: [Pw_forum] getting restart files from local scratch directories in parallel
 runs
Message-ID: <Pine.LNX.4.58.0605071231380.23010@macul.ciencias.uchile.cl>

Hello,
Is it possible to write the wafefunctions, densities, and potential files
in local directories during the calculation, and then write the
prefix.save in
the user NFS directory? As I understand, all files are saved in the
"outdir"  directory, which should be local, not an NFS directory. However,
doing that I lost the possibility of restart.
Maybe one only needs to retrieve the $outdir/prefix.save file from
 "any one" or from a "specific one"
of the local directories using a command of the particular queue system.
Is that the case?

Eduardo A. Menendez Proupin
Department of Physics
Faculty of Science
University of Chile
Las Palmeras 3425
?u?oa, Santiago
Chile
Phone: 56+2+978 74 11
http://fisica.ciencias.uchile.cl/~emenendez/


From konstantin_kudin at yahoo.com  Mon May  8 00:16:15 2006
From: konstantin_kudin at yahoo.com (Konstantin Kudin)
Date: Sun, 7 May 2006 15:16:15 -0700 (PDT)
Subject: [Pw_forum] getting restart files from local scratch directories in parallel runs
In-Reply-To: <Pine.LNX.4.58.0605071231380.23010@macul.ciencias.uchile.cl>
Message-ID: <20060507221615.7355.qmail@web52006.mail.yahoo.com>

 Hi Eduardo,

 With PWSCF 3.0 you have an option to write wavefunction files (usually
the largest) to a separate directory, specified by wfcdir='path', and
these could be local to the nodes. 

 On the other hand, outdir can be set to some NFS storage. Basically,
for an efficient restart one only needs the density file *.rho, and the
"prefix.save" stuff. That is why these go to a more permanent place
such as 'outdir'.

 Ultimately what you really want to do is to minimize the NFS traffic,
and making *.wfc files local to the nodes is often sufficient for that.

 If keeping 'outdir' on the NFS filesystem is still too demanding, you
could copy the required stuff to the first node, set your 'outdir' to
that, and once the job is finished copy it back. But that requires a
bit more scripting.

 I hope this helps.

 Kostya

 

 

--- Eduardo Ariel Menendez P <emenendez at macul.ciencias.uchile.cl>
wrote:

> 
> Hello,
> Is it possible to write the wafefunctions, densities, and potential
> files
> in local directories during the calculation, and then write the
> prefix.save in
> the user NFS directory? As I understand, all files are saved in the
> "outdir"  directory, which should be local, not an NFS directory.
> However,
> doing that I lost the possibility of restart.
> Maybe one only needs to retrieve the $outdir/prefix.save file from
>  "any one" or from a "specific one"
> of the local directories using a command of the particular queue
> system.
> Is that the case?
> 
> Eduardo A. Menendez Proupin
> Department of Physics
> Faculty of Science
> University of Chile
> Las Palmeras 3425
> ?u?oa, Santiago
> Chile
> Phone: 56+2+978 74 11
> http://fisica.ciencias.uchile.cl/~emenendez/
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
> 


__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 


From giannozz at nest.sns.it  Mon May  8 10:19:35 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Mon, 8 May 2006 10:19:35 +0200
Subject: [Pw_forum] Pu pseudopotential generation
In-Reply-To: <2fd252650605060029w314dc2fctc7c8a8f4ff311afc@mail.gmail.com>
References: <2fd252650605060029w314dc2fctc7c8a8f4ff311afc@mail.gmail.com>
Message-ID: <200605081019.35462.giannozz@nest.sns.it>

On Saturday 06 May 2006 09:29, Dmitry Korotin wrote:

> I'm trying to generate ultrasoft pseudopotential for Pu (with ld1.x code).
> This is my first attempt to generate some pseudo

it is a rather impressive first attempt: ultrasoft full-relativistic
with 14 projectors! I would start with something simpler 
(which, in the case of Pu, may still be rather complex).

The specific error:

>      from gener_pseudo : error #         1
>      too many nodes

means that the code couldn't manage to produce a nodeless
pseudowavefunction

Paolo
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From giannozz at nest.sns.it  Mon May  8 10:33:08 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Mon, 8 May 2006 10:33:08 +0200
Subject: [Pw_forum] getting restart files from local scratch directories in parallel runs
In-Reply-To: <Pine.LNX.4.58.0605071231380.23010@macul.ciencias.uchile.cl>
References: <Pine.LNX.4.58.0605071231380.23010@macul.ciencias.uchile.cl>
Message-ID: <200605081033.08513.giannozz@nest.sns.it>

On Sunday 07 May 2006 18:38, Eduardo Ariel Menendez P wrote:

> Is it possible to write the wafefunctions, densities, and potential 
> files in local directories during the calculation, and then write the
> prefix.save in the user NFS directory? 

it might be possible but it requires some changes in the code and
in the way input variables are handled, i.e. you need to specify 
separate "scratch" and "output" directories. There is presently 
some confusion in I/O: which code writes which files in which 
directory, but fixing this requires some time and some care. It
will be done but not right now.

Paolo
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From giannozz at nest.sns.it  Tue May  9 11:12:13 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Tue, 9 May 2006 11:12:13 +0200
Subject: [Pw_forum] Tutorial: Simulating matter at the nano-scale
Message-ID: <200605091112.13678.giannozz@nest.sns.it>

               CECAM-Psi-k-MolSimu tutorial:

"Simulating matter at the nano-scale using density-functional theory,
             pseudopotentials and plane waves"
                Lyon, November 13-17, 2006
!!! please note change of date^^^^^

This tutorial is aimed at students and young researchers who need 
an introduction to the field of electronic structure computations. 
It is also directed to non-specialists (e.g. experimentalists) who 
need to use the results of computer simulations for their work.

The tutorial will consist in morning lectures, followed by afternoon
practical hand-on sessions. In the lectures, emphasis will be put on 
the fundamental concepts, illustrated by a variety of examples showing 
the merits, but also the limitations, of the various DFT implementations. 
In addition to standard ground state properties (equilibrium structure,
cohesive energy, etc.) and Kohn-Sham orbitals analysis, an introduction
to modern linear response theories (dielectric response, vibrational 
properties) will be provided. Finally, the concepts and tools of molecular
dynamics (Verlet algorithm, thermostats, etc.) will be presented.  
In the hands-on sessions, examples of practical applications will be
shown, using the Quantum-ESPRESSO distribution as computational tool.

The number of participants is limited to 30. 
Some background in condensed-matter physics or theoretical chemistry, 
and a basic knowledge of Unix (Linux), are required. 
Cecam and MolSimu will provide funding covering travel and living 
expenses for eligible participants. 
Interested people should apply via the web site http://www.cecam.fr
(you should require a user and a password if you do not have any;
please check that your application is "submitted", not just "open").
Notifications of acceptance will be sent starting in September 
(please note change wrt previous announcements). It will
still be possible to apply until all slots are filled.

Organizers:
Paolo Giannozzi, SNS Pisa and Democritos Trieste
Xavier Blase, Universit? Lyon 1 and CNRS
Ralph Gebauer, ICTP and Democritos Trieste

-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From lukethulin at netscape.net  Tue May  9 14:20:50 2006
From: lukethulin at netscape.net (Luke Thulin)
Date: Tue, 09 May 2006 08:20:50 -0400
Subject: [Pw_forum] titania crystal stucture
Message-ID: <44608922.8040701@netscape.net>

Hello,

I'd like to model the band structure of anatase titania, but I'm having 
more difficulty than I expected just finding the coordinates for the 
atoms in the unit cell.  So far I've only been able to find out that it 
has some type of tetragonal or body centered tetragonal structure with a 
= 3.785A and c=9.514A.  Can anyone help me fill in the rest of what I 
need to enter the structure into QE so I can get a good visual of it in 
XCrysDen?  Or for future reference, does anyone know of any resources, 
online or otherwise, of the crystal structures of some common compounds 
such as this?

Thank you very much,
Luke Thulin


From aaron at chips.ncsu.edu  Tue May  9 15:54:39 2006
From: aaron at chips.ncsu.edu (aaron at chips.ncsu.edu)
Date: Tue, 9 May 2006 09:54:39 -0400 (EDT)
Subject: [Pw_forum] titania crystal stucture
In-Reply-To: <44608922.8040701@netscape.net>
References: <44608922.8040701@netscape.net>
Message-ID: <Pine.LNX.4.58.0605090953060.14164@chips.ncsu.edu>

Hi Luke,

Wikipedia has quite a bit of info on Anatase Titania.

Here's a link given to the the crystal structure:

http://ruby.colorado.edu/~smyth/min/tio2.html

I hope this helps.

Best,

Aaron


On Tue, 9 May 2006, Luke Thulin wrote:

> Hello,
>
> I'd like to model the band structure of anatase titania, but I'm having
> more difficulty than I expected just finding the coordinates for the
> atoms in the unit cell.  So far I've only been able to find out that it
> has some type of tetragonal or body centered tetragonal structure with a
> = 3.785A and c=9.514A.  Can anyone help me fill in the rest of what I
> need to enter the structure into QE so I can get a good visual of it in
> XCrysDen?  Or for future reference, does anyone know of any resources,
> online or otherwise, of the crystal structures of some common compounds
> such as this?
>
> Thank you very much,
> Luke Thulin
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
>


From ezadshojaee at hotmail.com  Tue May  9 16:59:50 2006
From: ezadshojaee at hotmail.com (Ezad Shojaee)
Date: Tue, 09 May 2006 14:59:50 +0000
Subject: [Pw_forum] question about   band structure
Message-ID: <BAY19-F20CDDC3D499CB2513BE99AAA90@phx.gbl>

hi everyone
i want to plot a band structure along a path  in the unit cell of TiO2  & i 
have this error :

from potinit : error #         1
     starting and expected charges differ

what is wrong with my calculation ?
this is my input file for pw.x  :

&control
    calculation='nscf'
    prefix='TiO2',
    pseudo_dir = '/root/pseudo/',
    outdir='/root/tmp/'
/
&system
    ibrav= 7,  celldm(1) =3.785,   celldm(3) =2.5136,
    nat=  6, ntyp= 2,
    ecutwfc =200.0,
    nbnd = 12,
/
&electrons
/
ATOMIC_SPECIES
Ti  47.867     Ti.vdb.UPF
O   15.9994    O.vdb.UPF
ATOMIC_POSITIONS  { bohr }
Ti       0.000000000  -0.946250000  -1.189250000
Ti       0.000000000   0.946250000   1.189250000
O       0.000000000   0.946250000  -2.593319231
O       0.000000000  -0.946250000   2.593319231
O       1.892500000   0.946250000  -0.214759231
O      -1.892500000  -0.946250000   0.214759231

K_POINTS {tpiba}
20
   0.0 0.0 0.00  1
   0.0 0.0 0.10  1
   0.0 0.0 0.15  1
   0.0 0.0 0.20  1
   0.0 0.0 0.25  1
   0.0 0.0 0.30  1
   0.0 0.0 0.35  1
   0.0 0.0 0.40  1
   0.0 0.0 0.45  1
   0.0 0.0 0.50  1
   0.0 0.0 0.55  1
   0.0 0.0 0.60  1
   0.0 0.0 0.65  1
   0.0 0.0 0.70  1
   0.0 0.0 0.75  1
   0.0 0.0 0.80  1
   0.0 0.0 0.85  1
   0.0 0.0 0.90  1
   0.0 0.0 0.95  1
   0.0 0.0 1.00  1

_________________________________________________________________
Express yourself instantly with MSN Messenger! Download today it's FREE! 
http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/



From giannozz at nest.sns.it  Tue May  9 17:16:22 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Tue, 9 May 2006 17:16:22 +0200
Subject: [Pw_forum] question about   band structure
In-Reply-To: <BAY19-F20CDDC3D499CB2513BE99AAA90@phx.gbl>
References: <BAY19-F20CDDC3D499CB2513BE99AAA90@phx.gbl>
Message-ID: <200605091716.22442.giannozz@nest.sns.it>

On Tuesday 09 May 2006 16:59, Ezad Shojaee wrote:

> from potinit : error #         1
>      starting and expected charges differ

the starting charge (produced by a preceding self
consistent calculation) is either bad or inconsistent 
with the data for the non-self-consistent calculation

>     ecutwfc =200.0,

200Ry for ultrasoft PP's? you have a lot of computer time
to waste

Paolo
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From suriano at sissa.it  Tue May  9 17:26:29 2006
From: suriano at sissa.it (Antonio Suriano)
Date: Tue, 9 May 2006 17:26:29 +0200
Subject: [Pw_forum] pools and neb
Message-ID: <DB87AA05-89C8-41A1-A749-EB7B021ABB61@sissa.it>

Hi

pw 2.*  sometimes crashes when we use k points paralization with neb.
My questions:

1) How to use image paralization? (some example of input file)
2) is 3.* affected by this bug?
3) Is it possible to convert 2.* path file to 3.* path files for a  
restart?
4) Is 4 processors for pool and as pools as k-points the best  
performance achieving setup?





Chiacchiera con i tuoi amici in tempo reale! 
 http://it.yahoo.com/mail_it/foot/*http://it.messenger.yahoo.com 


From brsahu at physics.utexas.edu  Tue May  9 17:38:18 2006
From: brsahu at physics.utexas.edu (Bhagawan Sahu)
Date: Tue, 9 May 2006 10:38:18 -0500 (CDT)
Subject: [Pw_forum] compilation error in iotk
Message-ID: <Pine.LNX.4.44.0605091020460.6409-100000@linux2.ph.utexas.edu>

Hi,

 I am trying to compile espresso-3.0 on a linux 64 bit cluster (intel 
xeon's) intel 8.1 compiler. 

The ./configure step correctly identifies the linux EMT64 architechture

checking build system type... i686-pc-linux-gnu
checking architecture... linux64
checking for mpif90... mpif90
checking for Fortran 77 compiler default output... a.out
checking whether the Fortran 77 compiler works... yes
checking whether we are cross compiling... yes
checking for suffix of executables... 
checking for suffix of object files... o
checking whether we are using the GNU Fortran 77 compiler... no
checking whether mpif90 accepts -g... yes
checking version of mpif90... ifort 8.1
checking for ifort... ifort
checking whether we are using the GNU Fortran 77 compiler... no
checking whether ifort accepts -g... yes
setting F90... ifort
......
.....

Parallel environment detected successfully.
Configured for compilation of parallel executables.

For more info, read the ESPRESSO User's Guide (Doc/users-guide.tex).
--------------------------------------------------------------------
configure: success

 make.sys file is also generated. 

However, invoking 'make all' gives the following error in 'iotk'

mpif90 -O2 -assume byterecl -nomodule -fpp -D__LINUX64 -D__INTEL -D__FFTW 
-D__USE_INTERNAL_FFTW -D__MPI -D__PARA -I../include  -I. -I../Modules 
-I../PW -I../PH -I../iotk/src -c iotk_attr+COMPLEX1_0.f90
iotk_config.h(73): #error:  unknown Linux64 compiler, please add to 
iotk_config.h

iotk_config.h(73): #error:  unknown Linux64 compiler, please add to 
iotk_config.h

make[2]: *** [iotk_attr+COMPLEX1_0.o] Error 2
make[2]: Leaving directory 
`/work/eefz340/sahu/pwscf/espresso-3.0/iotk/src'
make[1]: *** [libiotk.a] Error 2
make[1]: Leaving directory `/work/eefz340/sahu/pwscf/espresso-3.0/iotk'
make: *** [libiotk] Error 2

Pl. note I downloaded the latest esperesso-3.0 package from pwscf web 
today.

Sahu



From giannozz at nest.sns.it  Tue May  9 17:49:28 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Tue, 9 May 2006 17:49:28 +0200
Subject: [Pw_forum] compilation error in iotk
In-Reply-To: <Pine.LNX.4.44.0605091020460.6409-100000@linux2.ph.utexas.edu>
References: <Pine.LNX.4.44.0605091020460.6409-100000@linux2.ph.utexas.edu>
Message-ID: <200605091749.28992.giannozz@nest.sns.it>

On Tuesday 09 May 2006 17:38, Bhagawan Sahu wrote:

> iotk_config.h(73): #error:  unknown Linux64 compiler, please add to
> iotk_config.h

please replace iotk/include/iotk_config.h with the attached version
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy
-------------- next part --------------
A non-text attachment was scrubbed...
Name: iotk_config.h
Type: text/x-chdr
Size: 2654 bytes
Desc: not available
Url : /pipermail/attachments/20060509/d1e67f21/attachment.h 

From jcconesa at icp.csic.es  Tue May  9 18:00:47 2006
From: jcconesa at icp.csic.es (=?ISO-8859-1?Q?Jos=E9_C._Conesa?=)
Date: Tue, 09 May 2006 18:00:47 +0200
Subject: [Pw_forum] titania crystal stucture
In-Reply-To: <44608922.8040701@netscape.net>
Message-ID: <4460D8CF.20782.40FEA62A@localhost>

On 9 May 2006 at 8:20, Luke Thulin wrote:

> Hello,
> 
> I'd like to model the band structure of anatase titania, but I'm having 
> more difficulty than I expected just finding the coordinates for the 
> atoms in the unit cell.  So far I've only been able to find out that it 
> has some type of tetragonal or body centered tetragonal structure with a 
> = 3.785A and c=9.514A.  Can anyone help me fill in the rest of what I 
> need to enter the structure into QE so I can get a good visual of it in 
> XCrysDen?  Or for future reference, does anyone know of any resources, 
> online or otherwise, of the crystal structures of some common compounds 
> such as this?
> 
There are a number of crystallography databases where you can search structures 
online. A great one is ICSD, but only a small fraction of the structures is available to 
nonsuscribed institutions; even in that case, a good set of the most fundamental ones 
can be retrieved freely there through the access at
http://icsdweb.fiz-karlsruhe.de/index.php 
Other interesting structure servers are:

http://rruff.geo.arizona.edu/AMS/amcsd.php
http://cst-www.nrl.navy.mil/lattice/

but surely there are more. Perhaps if yopu look with google for "crystal structure 
database" you would find some of them. Of course, knowing how to handle a crystal 
description (with space group, taking into account settings, etc.) is quite impoertant to 
know what to do with the structure descriptions. For such structure handling, and also 
visualization and symmetry operations, a good program is PowderCell for Windows 
(now at version 2.4; free to download), see:
http://www.ccp14.ac.uk/ccp/web-mirrors/powdcell/a_v/v_1/powder/e_cell.html
Good luck,

Jose C. Conesa
Instituto de Catalisis y Petroleoquimica, CSIC
Campus de Cantoblanco
28049 Madrid  -  Spain
Phone Nr. 34-91-5854766
Fax Nr. 34-91-5854760



From giannozz at nest.sns.it  Tue May  9 18:05:46 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Tue, 9 May 2006 18:05:46 +0200
Subject: [Pw_forum] pools and neb
In-Reply-To: <DB87AA05-89C8-41A1-A749-EB7B021ABB61@sissa.it>
References: <DB87AA05-89C8-41A1-A749-EB7B021ABB61@sissa.it>
Message-ID: <200605091805.46417.giannozz@nest.sns.it>

On Tuesday 09 May 2006 17:26, Antonio Suriano wrote:

> pw 2.*  sometimes crashes when we use k points paralization 
> with neb. 

"sometimes"="in a deterministic way", or "randomly" ?

> 2) is 3.* affected by this bug?

please try: http://web1.sns.it/~giannozz/public/espresso-3.1.tar.gz
if you do not want to download the CVS version

Paolo
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From lukethulin at netscape.net  Tue May  9 18:11:06 2006
From: lukethulin at netscape.net (Luke Thulin)
Date: Tue, 09 May 2006 12:11:06 -0400
Subject: [Pw_forum] titania crystal stucture
In-Reply-To: <4460D8CF.20782.40FEA62A@localhost>
References: <4460D8CF.20782.40FEA62A@localhost>
Message-ID: <4460BF1A.4000005@netscape.net>

Thanks for your help.  I eventually did stumble across the 
http://cst-www.nrl.navy.mil/lattice/struk/c5.html site, it is a good one.

jcconesa at icp.csic.es wrote:

>On 9 May 2006 at 8:20, Luke Thulin wrote:
>
>  
>
>>Hello,
>>
>>I'd like to model the band structure of anatase titania, but I'm having 
>>more difficulty than I expected just finding the coordinates for the 
>>atoms in the unit cell.  So far I've only been able to find out that it 
>>has some type of tetragonal or body centered tetragonal structure with a 
>>= 3.785A and c=9.514A.  Can anyone help me fill in the rest of what I 
>>need to enter the structure into QE so I can get a good visual of it in 
>>XCrysDen?  Or for future reference, does anyone know of any resources, 
>>online or otherwise, of the crystal structures of some common compounds 
>>such as this?
>>
>>    
>>
>There are a number of crystallography databases where you can search structures 
>online. A great one is ICSD, but only a small fraction of the structures is available to 
>nonsuscribed institutions; even in that case, a good set of the most fundamental ones 
>can be retrieved freely there through the access at
>http://icsdweb.fiz-karlsruhe.de/index.php 
>Other interesting structure servers are:
>
>http://rruff.geo.arizona.edu/AMS/amcsd.php
>http://cst-www.nrl.navy.mil/lattice/
>
>but surely there are more. Perhaps if yopu look with google for "crystal structure 
>database" you would find some of them. Of course, knowing how to handle a crystal 
>description (with space group, taking into account settings, etc.) is quite impoertant to 
>know what to do with the structure descriptions. For such structure handling, and also 
>visualization and symmetry operations, a good program is PowderCell for Windows 
>(now at version 2.4; free to download), see:
>http://www.ccp14.ac.uk/ccp/web-mirrors/powdcell/a_v/v_1/powder/e_cell.html
>Good luck,
>
>Jose C. Conesa
>Instituto de Catalisis y Petroleoquimica, CSIC
>Campus de Cantoblanco
>28049 Madrid  -  Spain
>Phone Nr. 34-91-5854766
>Fax Nr. 34-91-5854760
>
>_______________________________________________
>Pw_forum mailing list
>Pw_forum at pwscf.org
>http://www.democritos.it/mailman/listinfo/pw_forum
>  
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: /pipermail/attachments/20060509/b4db5dca/attachment.htm 

From suriano at sissa.it  Tue May  9 18:11:54 2006
From: suriano at sissa.it (Antonio Suriano)
Date: Tue, 9 May 2006 18:11:54 +0200
Subject: [Pw_forum] pools and neb
In-Reply-To: <200605091805.46417.giannozz@nest.sns.it>
References: <DB87AA05-89C8-41A1-A749-EB7B021ABB61@sissa.it> <200605091805.46417.giannozz@nest.sns.it>
Message-ID: <254F74E8-430A-4BD9-B10D-C481E07E244C@sissa.it>

Il giorno 09/mag/06, alle ore 18:05, Paolo Giannozzi ha scritto:

> On Tuesday 09 May 2006 17:26, Antonio Suriano wrote:
>
>> pw 2.*  sometimes crashes when we use k points paralization
>> with neb.
>
> "sometimes"="in a deterministic way", or "randomly" ?


Sometimes in the end of the THIRD scf calculation (in my case the  
fifth image)  before calculating forces.
It just waits for something and does nothing.
It does not stop or crash. Simply waits and runs doing nothing.

Chiacchiera con i tuoi amici in tempo reale! 
 http://it.yahoo.com/mail_it/foot/*http://it.messenger.yahoo.com 


From brsahu at physics.utexas.edu  Tue May  9 18:27:13 2006
From: brsahu at physics.utexas.edu (Bhagawan Sahu)
Date: Tue, 9 May 2006 11:27:13 -0500 (CDT)
Subject: [Pw_forum] compilation error in iotk
In-Reply-To: <200605091749.28992.giannozz@nest.sns.it>
Message-ID: <Pine.LNX.4.44.0605091118220.7454-100000@linux2.ph.utexas.edu>

Hi,

 It worked. But it gives error in momery.f90.

___________________________________________________________
/Modules/path_opt_routines.o ../Modules/path_io_routines.o 
../Modules/path_reparametrisation.o ../Modules/parallel_include.o 
../Modules/parameters.o ../Modules/parser.o ../Modules/printout_base.o 
../Modules/pseudo_types.o ../Modules/read_cards.o 
../Modules/read_namelists.o ../Modules/readpseudo.o ../Modules/recvec.o 
../Modules/shmem_include.o ../Modules/splinelib.o ../Modules/stick_base.o 
../Modules/uspp.o ../Modules/version.o ../Modules/wavefunctions.o 
../Modules/wave_base.o ../Modules/timestep.o ../Modules/xml_io_base.o 
../flib/ptools.a ../flib/flib.a ../clib/clib.a ../iotk/src/libiotk.a 
-L/opt/intel/mkl721/lib/em64t -lmkl_lapack -L/opt/intel/mkl721/lib/em64t 
-lmkl_em64t -lguide -lpthread -L/opt/apps/fftw/fftw-3.0.1/lib -lfftw3  
/opt/intel/mkl721/lib/em64t/libmkl_lapack.a: could not read symbols: File 
format not recognized
make[1]: *** [memory.x] Error 1
make[1]: Leaving directory `/work/eefz340/sahu/espresso-3.0/PW'
make: *** [pw] Error 2
_____________________


Sahu

On Tue, 9 May 2006, Paolo Giannozzi wrote:

> On Tuesday 09 May 2006 17:38, Bhagawan Sahu wrote:
> 
> > iotk_config.h(73): #error:  unknown Linux64 compiler, please add to
> > iotk_config.h
> 
> please replace iotk/include/iotk_config.h with the attached version
> 



From akohlmey at cmm.upenn.edu  Tue May  9 18:46:26 2006
From: akohlmey at cmm.upenn.edu (Axel Kohlmeyer)
Date: Tue, 9 May 2006 12:46:26 -0400
Subject: [Pw_forum] compilation error in iotk
In-Reply-To: <Pine.LNX.4.44.0605091118220.7454-100000@linux2.ph.utexas.edu>
References: <200605091749.28992.giannozz@nest.sns.it>
	 <Pine.LNX.4.44.0605091118220.7454-100000@linux2.ph.utexas.edu>
Message-ID: <7b6913e90605090946y3da9a369jee0a9f5b4f1fb4e0@mail.gmail.com>

On 5/9/06, Bhagawan Sahu <brsahu at physics.utexas.edu> wrote:
>
> Hi,
>
>  It worked. But it gives error in momery.f90.

[...]

> ../Modules/wave_base.o ../Modules/timestep.o ../Modules/xml_io_base.o
> ../flib/ptools.a ../flib/flib.a ../clib/clib.a ../iotk/src/libiotk.a
> -L/opt/intel/mkl721/lib/em64t -lmkl_lapack -L/opt/intel/mkl721/lib/em64t
> -lmkl_em64t -lguide -lpthread -L/opt/apps/fftw/fftw-3.0.1/lib -lfftw3
> /opt/intel/mkl721/lib/em64t/libmkl_lapack.a: could not read symbols: File
> format not recognized
> make[1]: *** [memory.x] Error 1
> make[1]: Leaving directory `/work/eefz340/sahu/espresso-3.0/PW'
> make: *** [pw] Error 2

seems like you have the 32-bit version of the compiler installed
and thus you need to change your make.sys file and the library
paths.

to confirm the compiler type try, e.g.:
file Modules/path_base.o

with 32bit you'd get:
Modules/path_base.o: ELF 32-bit LSB relocatable, Intel 80386, version
1 (GNU/Linux), not stripped

on opteron you'd get instead:
Modules/path_base.o: ELF 64-bit LSB relocatable, AMD x86-64, version 1
(GNU/Linux), not stripped

regards,
     axel.

> _____________________
>
>
> Sahu
>


--
=======================================================================
Axel Kohlmeyer   akohlmey at cmm.chem.upenn.edu   http://www.cmm.upenn.edu
  Center for Molecular Modeling   --   University of Pennsylvania
Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323
tel: 1-215-898-1582,  fax: 1-215-573-6233,  office-tel: 1-215-898-5425
=======================================================================
If you make something idiot-proof, the universe creates a better idiot.


From ferretti.andrea at unimore.it  Tue May  9 19:06:10 2006
From: ferretti.andrea at unimore.it (Andrea Ferretti)
Date: Tue, 9 May 2006 19:06:10 +0200 (CEST)
Subject: [Pw_forum] compilation error in iotk
In-Reply-To: <7b6913e90605090946y3da9a369jee0a9f5b4f1fb4e0@mail.gmail.com>
References: <200605091749.28992.giannozz@nest.sns.it> 
 <Pine.LNX.4.44.0605091118220.7454-100000@linux2.ph.utexas.edu>
 <7b6913e90605090946y3da9a369jee0a9f5b4f1fb4e0@mail.gmail.com>
Message-ID: <Pine.LNX.4.58.0605091855080.19896@bamboo.fisica.unimo.it>


 On 5/9/06, Bhagawan Sahu <brsahu at physics.utexas.edu> wrote:
 
 > ../Modules/wave_base.o ../Modules/timestep.o ../Modules/xml_io_base.o
 > ../flib/ptools.a ../flib/flib.a ../clib/clib.a ../iotk/src/libiotk.a
 > -L/opt/intel/mkl721/lib/em64t -lmkl_lapack -L/opt/intel/mkl721/lib/em64t
 > -lmkl_em64t -lguide -lpthread -L/opt/apps/fftw/fftw-3.0.1/lib -lfftw3
 > /opt/intel/mkl721/lib/em64t/libmkl_lapack.a: could not read symbols: File
 > format not recognized
 > make[1]: *** [memory.x] Error 1
 > make[1]: Leaving directory `/work/eefz340/sahu/espresso-3.0/PW'
 > make: *** [pw] Error 2
 
letting aside the main compilation error (already tackled by Axel), 
it seems to me you are also using fftw v3.x which are not compatible with the 
std version of espresso 3.0 (as far as know)... you should use fftw 2.x instead

andrea






From giannozz at nest.sns.it  Tue May  9 19:06:23 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Tue, 9 May 2006 19:06:23 +0200
Subject: [Pw_forum] pools and neb
In-Reply-To: <254F74E8-430A-4BD9-B10D-C481E07E244C@sissa.it>
References: <DB87AA05-89C8-41A1-A749-EB7B021ABB61@sissa.it> <200605091805.46417.giannozz@nest.sns.it> <254F74E8-430A-4BD9-B10D-C481E07E244C@sissa.it>
Message-ID: <200605091906.23825.giannozz@nest.sns.it>

On Tuesday 09 May 2006 18:11, Antonio Suriano wrote:

> Sometimes in the end of the THIRD scf calculation (in my case 
> the fifth image)  before calculating forces.
> It just waits for something and does nothing.
> It does not stop or crash. Simply waits and runs doing nothing.

it is waiting another processor, or group of processors. Typically
this happens because one processor or groups of processors has
either crashed or gone in a different direction wrt to the others.
It helps if you let all processors print to file, flushing the output 
as frequently as you can (otherwise the interesting part will stay
in the buffer)

Paolo
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From giannozz at nest.sns.it  Tue May  9 19:09:08 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Tue, 9 May 2006 19:09:08 +0200
Subject: [Pw_forum] compilation error in iotk
In-Reply-To: <Pine.LNX.4.58.0605091855080.19896@bamboo.fisica.unimo.it>
References: <200605091749.28992.giannozz@nest.sns.it> <7b6913e90605090946y3da9a369jee0a9f5b4f1fb4e0@mail.gmail.com> <Pine.LNX.4.58.0605091855080.19896@bamboo.fisica.unimo.it>
Message-ID: <200605091909.08580.giannozz@nest.sns.it>

On Tuesday 09 May 2006 19:06, Andrea Ferretti wrote:

> letting aside the main compilation error (already tackled by Axel),
> it seems to me you are also using fftw v3.x which are not compatible 
> with the std version of espresso 3.0 (as far as know)... you should use 
> fftw 2.x instead

I was going to write the same, but I didn't since in the first message 
I saw -D__USE_INTERNAL_FFTW, so it doesn't matter which libraries 
one loads: they aren't used. 

FFTW v.3 is sort of implemented in the CVS version, though

Paolo
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From brsahu at physics.utexas.edu  Tue May  9 19:14:31 2006
From: brsahu at physics.utexas.edu (Bhagawan Sahu)
Date: Tue, 9 May 2006 12:14:31 -0500 (CDT)
Subject: [Pw_forum] compilation error in iotk
In-Reply-To: <7b6913e90605090946y3da9a369jee0a9f5b4f1fb4e0@mail.gmail.com>
Message-ID: <Pine.LNX.4.44.0605091207570.8086-100000@linux2.ph.utexas.edu>

Alex,

It does give

 Modules/path_base.o: ELF 32-bit LSB relocatable, Intel 80386, version
1 (GNU/Linux), not stripped
 
I used 

setenv ARCH linux64 before invoking the ./configure step. It does pick up
the linux64 architecture.

checking build system type... i686-pc-linux-gnu
checking architecture... linux64
checking for mpif90... mpif90
checking for Fortran 77 compiler default output... a.out
checking whether the Fortran 77 compiler works... yes
checking whether we are cross compiling... yes
checking for suffix of executables... 
checking for suffix of object files... o
checking whether we are using the GNU Fortran 77 compiler... no
checking whether mpif90 accepts -g... yes
checking version of mpif90... ifort 8.1
checking for ifort... ifort
checking whether we are using the GNU Fortran 77 compiler... no
checking whether ifort accepts -g... yes
setting F90... ifort
setting MPIF90... mpif90
checking for mpicc... mpicc
checking whether we are using the GNU C compiler... yes
checking whether mpicc accepts -g... yes
checking for mpicc option to accept ANSI C... none needed
checking for mpif77... mpif77

At this I am not sure whetner the f77 f90 compilers it picks up under 
LINUX 64 architechture is 32 bit or 64 bit. but path.o shows it to be 32 
bit.

I am checking with the sysad about this. I guess if I compile with 32 bit
then the problem will not be there

Sahu

On Tue, 9 May 2006, Axel Kohlmeyer wrote:

> On 5/9/06, Bhagawan Sahu <brsahu at physics.utexas.edu> wrote:
> >
> > Hi,
> >
> >  It worked. But it gives error in momery.f90.
> 
> [...]
> 
> > ../Modules/wave_base.o ../Modules/timestep.o ../Modules/xml_io_base.o
> > ../flib/ptools.a ../flib/flib.a ../clib/clib.a ../iotk/src/libiotk.a
> > -L/opt/intel/mkl721/lib/em64t -lmkl_lapack -L/opt/intel/mkl721/lib/em64t
> > -lmkl_em64t -lguide -lpthread -L/opt/apps/fftw/fftw-3.0.1/lib -lfftw3
> > /opt/intel/mkl721/lib/em64t/libmkl_lapack.a: could not read symbols: File
> > format not recognized
> > make[1]: *** [memory.x] Error 1
> > make[1]: Leaving directory `/work/eefz340/sahu/espresso-3.0/PW'
> > make: *** [pw] Error 2
> 
> seems like you have the 32-bit version of the compiler installed
> and thus you need to change your make.sys file and the library
> paths.
> ~
> to confirm the compiler type try, e.g.:
> file Modules/path_base.o
> 
> with 32bit you'd get:
> Modules/path_base.o: ELF 32-bit LSB relocatable, Intel 80386, version
> 1 (GNU/Linux), not stripped
> 
> on opteron you'd get instead:
> Modules/path_base.o: ELF 64-bit LSB relocatable, AMD x86-64, version 1
> (GNU/Linux), not stripped
> 
> regards,
>      axel.
> 
> > _____________________
> >
> >
> > Sahu
> >
> 
> 
> --
> =======================================================================
> Axel Kohlmeyer   akohlmey at cmm.chem.upenn.edu   http://www.cmm.upenn.edu
>   Center for Molecular Modeling   --   University of Pennsylvania
> Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323
> tel: 1-215-898-1582,  fax: 1-215-573-6233,  office-tel: 1-215-898-5425
> =======================================================================
> If you make something idiot-proof, the universe creates a better idiot.
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
> 



From uccaati at ucl.ac.uk  Tue May  9 19:21:44 2006
From: uccaati at ucl.ac.uk (Antonio Tilocca)
Date: Tue, 9 May 2006 18:21:44 +0100 (BST)
Subject: [Pw_forum] forces & energy conservation in cp
Message-ID: <Pine.GSO.4.58.0605091747420.27642@socrates-a.ucl.ac.uk>

dear users,

I was running some tests with the current (May, 7) cvs version of the cp
code, using the same input file, pseudos, etc as in a previous MD run with the
old (pre-qEspresso) cpv code.

Although energies and eigenvalues after the initial (fixed-ions)
electronic minimization are ok,
I have noticed some serious problems with total energy (constant
of motion econt) conservation in a subsequent NVE MD trajectory, which
were not present in the same run with the old code, and are not affected by the
choice of time step, fictitious mass, etc.

The comparison of the output files seems to point to a problem with the
force calculation: the forces on most ions match closely with the forces
obtained with the previous code, except for 2-3 ions whose z-component of
force calculated with the cvs version is up to 100 times larger than
what it should be.

I was wondering if someone else experienced similar problems?

thanks,

Antonio Tilocca




From brsahu at physics.utexas.edu  Tue May  9 20:21:02 2006
From: brsahu at physics.utexas.edu (Bhagawan Sahu)
Date: Tue, 9 May 2006 13:21:02 -0500 (CDT)
Subject: [Pw_forum] compilation error in iotk
In-Reply-To: <7b6913e90605090946y3da9a369jee0a9f5b4f1fb4e0@mail.gmail.com>
Message-ID: <Pine.LNX.4.44.0605091318380.8755-100000@linux2.ph.utexas.edu>

Dear Alex and Paolo,

 with 32-bit ARCH, the compilation and linking is done. thanx paolo for 
providing the iotk_config.h file and alex for hinting at 32 bit/64 bit 
conflict.

with Andrea/Paolo's comment I removed -D_USE_INTERNAL_FFTW and forced the 
code to use the machine compiled FFTW/2.1.5.

Sahu

On Tue, 9 May 2006, Axel Kohlmeyer wrote:

> On 5/9/06, Bhagawan Sahu <brsahu at physics.utexas.edu> wrote:
> >
> > Hi,
> >
> >  It worked. But it gives error in momery.f90.
> 
> [...]
> 
> > ../Modules/wave_base.o ../Modules/timestep.o ../Modules/xml_io_base.o
> > ../flib/ptools.a ../flib/flib.a ../clib/clib.a ../iotk/src/libiotk.a
> > -L/opt/intel/mkl721/lib/em64t -lmkl_lapack -L/opt/intel/mkl721/lib/em64t
> > -lmkl_em64t -lguide -lpthread -L/opt/apps/fftw/fftw-3.0.1/lib -lfftw3
> > /opt/intel/mkl721/lib/em64t/libmkl_lapack.a: could not read symbols: File
> > format not recognized
> > make[1]: *** [memory.x] Error 1
> > make[1]: Leaving directory `/work/eefz340/sahu/espresso-3.0/PW'
> > make: *** [pw] Error 2
> 
> seems like you have the 32-bit version of the compiler installed
> and thus you need to change your make.sys file and the library
> paths.
> 
> to confirm the compiler type try, e.g.:
> file Modules/path_base.o
> 
> with 32bit you'd get:
> Modules/path_base.o: ELF 32-bit LSB relocatable, Intel 80386, version
> 1 (GNU/Linux), not stripped
> 
> on opteron you'd get instead:
> Modules/path_base.o: ELF 64-bit LSB relocatable, AMD x86-64, version 1
> (GNU/Linux), not stripped
> 
> regards,
>      axel.
> 
> > _____________________
> >
> >
> > Sahu
> >
> 
> 
> --
> =======================================================================
> Axel Kohlmeyer   akohlmey at cmm.chem.upenn.edu   http://www.cmm.upenn.edu
>   Center for Molecular Modeling   --   University of Pennsylvania
> Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323
> tel: 1-215-898-1582,  fax: 1-215-573-6233,  office-tel: 1-215-898-5425
> =======================================================================
> If you make something idiot-proof, the universe creates a better idiot.
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
> 



From giannozz at nest.sns.it  Tue May  9 20:20:38 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Tue, 9 May 2006 20:20:38 +0200
Subject: [Pw_forum] forces & energy conservation in cp
In-Reply-To: <Pine.GSO.4.58.0605091747420.27642@socrates-a.ucl.ac.uk>
References: <Pine.GSO.4.58.0605091747420.27642@socrates-a.ucl.ac.uk>
Message-ID: <200605092020.38779.giannozz@nest.sns.it>

On Tuesday 09 May 2006 19:21, Antonio Tilocca wrote:

> I was running some tests with the current (May, 7) cvs version of the cp
> code, using the same input file, pseudos, etc as in a previous MD run with
> the old (pre-qEspresso) cpv code [...]
> I have noticed some serious problems with total energy (constant
> of motion econt) conservation in a subsequent NVE MD trajectory, which
> were not present in the same run with the old code, and are not affected by
> the choice of time step, fictitious mass, etc.

please provide a test job
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From giannozz at nest.sns.it  Tue May  9 20:24:27 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Tue, 9 May 2006 20:24:27 +0200
Subject: [Pw_forum] compilation error in iotk
In-Reply-To: <Pine.LNX.4.44.0605091318380.8755-100000@linux2.ph.utexas.edu>
References: <Pine.LNX.4.44.0605091318380.8755-100000@linux2.ph.utexas.edu>
Message-ID: <200605092024.27636.giannozz@nest.sns.it>

On Tuesday 09 May 2006 20:21, Bhagawan Sahu wrote:

> with Andrea/Paolo's comment I removed -D_USE_INTERNAL_FFTW and forced the
> code to use the machine compiled FFTW/2.1.5.

not a good idea: using the internal FFTW is the safest way

P.
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From sbraccia at sissa.it  Tue May  9 23:58:47 2006
From: sbraccia at sissa.it (carlo sbraccia)
Date: Tue, 09 May 2006 17:58:47 -0400
Subject: [Pw_forum] pools and neb
In-Reply-To: <DB87AA05-89C8-41A1-A749-EB7B021ABB61@sissa.it>
References: <DB87AA05-89C8-41A1-A749-EB7B021ABB61@sissa.it>
Message-ID: <44611097.1030703@sissa.it>

Dear Antonio,

> pw 2.*  sometimes crashes when we use k points paralization with neb.

you should provide the input that makes the code to crash. I have been 
using neb with parallelization on k-points for a long time and I've 
never had troubles.

> My questions:
>
> 1) How to use image paralization? (some example of input file)

for image parallelization just use

mpirun -np N pw.x -nimage M -in input > output

Of course M < N and M < num_of_images. Much better if M is a whole 
divisor of num_of_images.

> 2) is 3.* affected by this bug?

Maybe, since the "bug" has never been spotted before.

> 3) Is it possible to convert 2.* path file to 3.* path files for a  
> restart?

I think so: just remove the section "QUICK-MIN FIELDS" in the path file 
and restart.

> 4) Is 4 processors for pool and as pools as k-points the best  
> performance achieving setup?

It is impossible to give a general answer: it depends on both the system 
size and on the computer you are using.

However, I already posted a similar message a couple of months ago 
(http://www.democritos.it/pipermail/pw_forum/2006-March/003803.html) .

carlo


From vdiep at ictp.it  Wed May 10 10:01:21 2006
From: vdiep at ictp.it (Diep Quang Vinh)
Date: Wed, 10 May 2006 10:01:21 +0200 (CEST)
Subject: [Pw_forum] Using another basis set
Message-ID: <32867.10.104.33.121.1147248081.squirrel@webmail3.ictp.trieste.it>

Dear All,

In the code pw, I know that we use plane wave as the basis set, is it possible for us to
import different set and run the code? Also since we use pseudopotential, do you have an
idea on how one calculates the KB projector within the code?

Thank you very much

Vinh




From giannozz at nest.sns.it  Wed May 10 15:10:33 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Wed, 10 May 2006 15:10:33 +0200
Subject: [Pw_forum] Using another basis set
In-Reply-To: <32867.10.104.33.121.1147248081.squirrel@webmail3.ictp.trieste.it>
References: <32867.10.104.33.121.1147248081.squirrel@webmail3.ictp.trieste.it>
Message-ID: <200605101510.33912.giannozz@nest.sns.it>

On Wednesday 10 May 2006 10:01, Diep Quang Vinh wrote:

> In the code pw, I know that we use plane wave as the basis set, 
> is it possible for us to import different set and run the code? 

"to import results obtained from different basis sets" is in
principle possible, even if not trivial. To use a different
basis set, you may want to use a different code.

> do you have an idea on how one calculates the KB projector
> within the code?

see attached, eq. 41 . Integrals are performed numerically 
on a radial grid.

P.
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy
-------------- next part --------------
A non-text attachment was scrubbed...
Name: plane_waves.pdf
Type: application/pdf
Size: 36120 bytes
Desc: not available
Url : /pipermail/attachments/20060510/6c60d922/attachment.pdf 

From sodovar at gmail.com  Wed May 10 16:27:13 2006
From: sodovar at gmail.com (Ruslan Minibaev)
Date: Wed, 10 May 2006 18:27:13 +0400
Subject: [Pw_forum] USPP of In
Message-ID: <f530cbd10605100727x7b5d5f54rdeeba91d00f9442a@mail.gmail.com>

Dear all
I am looking for a ultrasoft pseudopotential of In (Indium).
Can anyone help me? Thanks.
Best regards, Ruslan Minibaev
-------------- next part --------------
An HTML attachment was scrubbed...
URL: /pipermail/attachments/20060510/f14ea6fa/attachment.htm 

From lukethulin at netscape.net  Wed May 10 19:51:53 2006
From: lukethulin at netscape.net (Luke Thulin)
Date: Wed, 10 May 2006 13:51:53 -0400
Subject: [Pw_forum] Confused on nat definition
Message-ID: <44622839.4030206@netscape.net>

Hello,

This may seem silly, but I'm confused as to what exactly the number of 
atoms in a unit cell is (nat).  For example, the Silicon example says 
that nat is only two, yet a diamond structure such as this should have 
much more than two atoms per unit cell.  Is the definition of nat the 
number of basis vectors?


From eyvaz_isaev at yahoo.com  Wed May 10 20:51:03 2006
From: eyvaz_isaev at yahoo.com (Eyvaz Isaev)
Date: Wed, 10 May 2006 11:51:03 -0700 (PDT)
Subject: [Pw_forum] Confused on nat definition
In-Reply-To: <44622839.4030206@netscape.net>
Message-ID: <20060510185103.53008.qmail@web60316.mail.yahoo.com>

Hi, 

> This may seem silly, but I'm confused as to what
> exactly the number of 
> atoms in a unit cell is (nat).  For example, the
> Silicon example says 
> that nat is only two, yet a diamond structure such
> as this should have 
> much more than two atoms per unit cell.  

Let us consider the diamond case. If you choose as
basis vectors next 3 vectors (which are the standard
choice)

1/2, 1/2, 0
1/2, 0  , 1/2
0  , 1/2, 1/2

you have only 2 atoms in the unit cell
(parallelepiped) spanned by these vectors:
0,     0, 0
1/4, 1/4, 1/4

If you decide to choose as basis vectors next 3 ones

1 0 0
0 1 0
0 0 1

you have 8 atoms in the unit cell which is now a cub.
If your choice is the latter  for CaF2 structure you
will have 12 atoms, but using the former - only 3. 

So, number of atoms (nat) in a unit cell depends on
your unit cell choice defined by 3 basis vectors.

> Is the definition of nat the  number of basis 
> vectors?
To me it is not so clear, but see above.

Bests,
Eyvaz.

> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
> 


__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 


From hqzhou at nju.edu.cn  Thu May 11 06:41:33 2006
From: hqzhou at nju.edu.cn (Huiqun Zhou)
Date: Thu, 11 May 2006 12:41:33 +0800
Subject: [Pw_forum] Confused on nat definition
References: <20060510185103.53008.qmail@web60316.mail.yahoo.com>
Message-ID: <005f01c674b5$2e0ffc60$1d00a8c0@solarflare>

Hmm ..., because there are 4 lattice points, that are (0,0,0) and 3 end 
points of the
basis vectors as stated, in a diamond structure, and there are two primitive 
elements
which are located at (0,0,0) and (1/4,1/4,1/4) for this structure, so there 
should be
4 x 2 = 8 atoms in a conventional unit cell of diamond structure.

But because of periodicity, theoretically you can choose either only the 
primitive
elements (2 atoms) in an unit cell with appropriatelly selected large number 
of k points
for calculation, or the whole member (8 atoms) in an unit cell with less k 
points for
calculation. I may be wrong, please correct me.

Huiqun Zhou

----- Original Message ----- 
From: "Eyvaz Isaev" <eyvaz_isaev at yahoo.com>
To: <pw_forum at pwscf.org>
Sent: Thursday, May 11, 2006 2:51 AM
Subject: Re: [Pw_forum] Confused on nat definition


> Hi,
>
>> This may seem silly, but I'm confused as to what
>> exactly the number of
>> atoms in a unit cell is (nat).  For example, the
>> Silicon example says
>> that nat is only two, yet a diamond structure such
>> as this should have
>> much more than two atoms per unit cell.
>
> Let us consider the diamond case. If you choose as
> basis vectors next 3 vectors (which are the standard
> choice)
>
> 1/2, 1/2, 0
> 1/2, 0  , 1/2
> 0  , 1/2, 1/2
>
> you have only 2 atoms in the unit cell
> (parallelepiped) spanned by these vectors:
> 0,     0, 0
> 1/4, 1/4, 1/4
>
> If you decide to choose as basis vectors next 3 ones
>
> 1 0 0
> 0 1 0
> 0 0 1
>
> you have 8 atoms in the unit cell which is now a cub.
> If your choice is the latter  for CaF2 structure you
> will have 12 atoms, but using the former - only 3.
>
> So, number of atoms (nat) in a unit cell depends on
> your unit cell choice defined by 3 basis vectors.
>
>> Is the definition of nat the  number of basis
>> vectors?
> To me it is not so clear, but see above.
>
> Bests,
> Eyvaz.
>
>> _______________________________________________
>> Pw_forum mailing list
>> Pw_forum at pwscf.org
>> http://www.democritos.it/mailman/listinfo/pw_forum
>>
>
>
> __________________________________________________
> Do You Yahoo!?
> Tired of spam?  Yahoo! Mail has the best spam protection around
> http://mail.yahoo.com
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
>
> 



From baroni at sissa.it  Thu May 11 07:29:45 2006
From: baroni at sissa.it (Stefano Baroni)
Date: Thu, 11 May 2006 07:29:45 +0200
Subject: [Pw_forum] Confused on nat definition
In-Reply-To: <005f01c674b5$2e0ffc60$1d00a8c0@solarflare>
References: <20060510185103.53008.qmail@web60316.mail.yahoo.com> <005f01c674b5$2e0ffc60$1d00a8c0@solarflare>
Message-ID: <4402C34F-C6E2-4D3D-A203-05EB51232708@sissa.it>

OK. Let me try to shed some light on this unexpectedly obscure point.

On May 11, 2006, at 6:41 AM, Huiqun Zhou wrote:

> Hmm ..., because there are 4 lattice points, that are (0,0,0) and 3  
> end points of the
> basis vectors as stated, in a diamond structure, and there are two  
> primitive elements
> which are located at (0,0,0) and (1/4,1/4,1/4) for this structure,  
> so there should be
> 4 x 2 = 8 atoms in a conventional unit cell of diamond structure.

I am sorry, Huiqun, but I am afraid that this argument is wrong and  
confusing. What do you mean by "4 lattice points"? Of course, every  
triplet of vectors originating at (0,0,0) has one common point (the  
origin), and 3 end points. AND SO WHAT?

> But because of periodicity, theoretically you can choose either  
> only the primitive
> elements (2 atoms) in an unit cell with appropriatelly selected  
> large number of k points
> for calculation, or the whole member (8 atoms) in an unit cell with  
> less k points for
> calculation. I may be wrong, please correct me.

Although it is certainly true that, for a same system, doing a  
calculation with a larger conventional unit cell would require a  
smaller number of k points, I feel that it is confusing to try to  
explain a property of the real-space lattice with something that has  
only to do with the way calculations are technically done (the number  
of k points).

Let me try to clarify the matter as much as I can.

1) The Bravais lattice of the diamond structure is face-centered  
cubic (FCC). A basis for this Bravais lattice is, e.g., (1/2,1/2,0),  
(1/2,0,1/2), (0,1/2,1/2).

2) Each elementary cell of this FCC lattice contains two equivalent  
atoms, located at (000)+R(l,m,n) and (1/4,1/4,1/4)+R(l,m,n), where R 
(l,m,n) = l*(1/2,1/2,0) + m*(1/2,0,1/2) + n*(0,1/2,1/2) is a generic  
point of the Bravais lattice (l,m,n being integer numbers).

3) The above basis for the Bravais lattice defines a unit cell of  
*minimum volume*. Of ourse, one is free to choose as a basis any  
triplet of integer linear (and linearly independent) combinations of  
the minimum basis. For instance, one could choose: (1,0,0) =  
(1/2,1/2,0)+(1/2,0,1/2)-(0,1/2,1/2), (0,1,0)= [guess what?], and  
(0,0,1)=[guess what?]. These three vectors form a basis for the  
simple cubic (SC) Bravais lattice. Calculate the volume of the  
original FCC unit cell (defined at point 1 above) and compare it with  
the volume of the SC unit cell considered here (hint: the volume of a  
parallelepiped is the absolute value of the determinant of the 3x3  
matrix whose columns are the coordinates of the three vectors which  
form the edges of the parallelepiped). You will find that the the  
volume of the SC cell is 4 times larger than the volume of the FCC  
cell. This means that if you choose to describe the diamond lattice  
as SC instead of FCC (which you are free to do), the unit cell will  
contain 4x2 instead of 2 atoms. You see where the factor 4 comes  
from? Nothing to do with the "number of lattice points" (whatever  
these lattice points may be).

4) Now, the volume of the Brillouin zone (BZ) is (2\pi)^3 divided by  
the volume of the unit cell. Hence, the volume of the SC BZ is 1/4  
the volume of the FCC BZ. That's why, in order to sample a SC BZ with  
the same accuracy of an FCC cell you will need 1/4 as many k points.  
But this is another story which has little to do with the main point  
of this thread.

Hope this clarifies a little bit the muddy waters

Stefano Baroni

>
> Huiqun Zhou
>
> ----- Original Message ----- From: "Eyvaz Isaev"  
> <eyvaz_isaev at yahoo.com>
> To: <pw_forum at pwscf.org>
> Sent: Thursday, May 11, 2006 2:51 AM
> Subject: Re: [Pw_forum] Confused on nat definition
>
>
>> Hi,
>>
>>> This may seem silly, but I'm confused as to what
>>> exactly the number of
>>> atoms in a unit cell is (nat).  For example, the
>>> Silicon example says
>>> that nat is only two, yet a diamond structure such
>>> as this should have
>>> much more than two atoms per unit cell.
>>
>> Let us consider the diamond case. If you choose as
>> basis vectors next 3 vectors (which are the standard
>> choice)
>>
>> 1/2, 1/2, 0
>> 1/2, 0  , 1/2
>> 0  , 1/2, 1/2
>>
>> you have only 2 atoms in the unit cell
>> (parallelepiped) spanned by these vectors:
>> 0,     0, 0
>> 1/4, 1/4, 1/4
>>
>> If you decide to choose as basis vectors next 3 ones
>>
>> 1 0 0
>> 0 1 0
>> 0 0 1
>>
>> you have 8 atoms in the unit cell which is now a cub.
>> If your choice is the latter  for CaF2 structure you
>> will have 12 atoms, but using the former - only 3.
>>
>> So, number of atoms (nat) in a unit cell depends on
>> your unit cell choice defined by 3 basis vectors.
>>
>>> Is the definition of nat the  number of basis
>>> vectors?
>> To me it is not so clear, but see above.
>>
>> Bests,
>> Eyvaz.
>>
>>> _______________________________________________
>>> Pw_forum mailing list
>>> Pw_forum at pwscf.org
>>> http://www.democritos.it/mailman/listinfo/pw_forum
>>>
>>
>>
>> __________________________________________________
>> Do You Yahoo!?
>> Tired of spam?  Yahoo! Mail has the best spam protection around
>> http://mail.yahoo.com
>> _______________________________________________
>> Pw_forum mailing list
>> Pw_forum at pwscf.org
>> http://www.democritos.it/mailman/listinfo/pw_forum
>>
>
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum

---
Stefano Baroni - SISSA  &  DEMOCRITOS National Simulation Center -  
Trieste
[+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype)

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: /pipermail/attachments/20060511/460554a1/attachment.htm 

From degironc at sissa.it  Thu May 11 09:22:56 2006
From: degironc at sissa.it (degironc)
Date: Thu, 11 May 2006 09:22:56 +0200
Subject: [Pw_forum] Confused on nat definition
In-Reply-To: <44622839.4030206@netscape.net>
References: <44622839.4030206@netscape.net>
Message-ID: <4462E650.9050603@sissa.it>

nat = number of atoms in the simulation cell.
silicon has two atoms in the  (fcc) Wigner-Seitz cell.
In a simple cubic (a supercell),  nat should be 8.

stefano
 

Luke Thulin wrote:

> Hello,
>
> This may seem silly, but I'm confused as to what exactly the number of 
> atoms in a unit cell is (nat).  For example, the Silicon example says 
> that nat is only two, yet a diamond structure such as this should have 
> much more than two atoms per unit cell.  Is the definition of nat the 
> number of basis vectors?
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum





From jyraty at ulg.ac.be  Thu May 11 10:00:39 2006
From: jyraty at ulg.ac.be (Jean-Yves Raty)
Date: Thu, 11 May 2006 10:00:39 +0200
Subject: [Pw_forum] Using another basis set
In-Reply-To: <200605101510.33912.giannozz@nest.sns.it>
Message-ID: <000a01c674d0$febebb10$04f4a8c0@Jeanyvesraty>

Dear Professor Gianozzi,

From giannozz at nest.sns.it  Thu May 11 10:17:18 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Thu, 11 May 2006 10:17:18 +0200
Subject: [Pw_forum] Using another basis set
In-Reply-To: <000a01c674d0$febebb10$04f4a8c0@Jeanyvesraty>
References: <000a01c674d0$febebb10$04f4a8c0@Jeanyvesraty>
Message-ID: <200605111017.18686.giannozz@nest.sns.it>

On Thursday 11 May 2006 10:00, Jean-Yves Raty wrote:

> From which reference book is the attached pdf file taken from ? 

from personal notes I started to write many years ago and never
finished... Paolo
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy
-------------- next part --------------
A non-text attachment was scrubbed...
Name: plane_waves.tex
Type: text/x-tex
Size: 27341 bytes
Desc: not available
Url : /pipermail/attachments/20060511/123008e9/attachment.tex 

From lukethulin at netscape.net  Thu May 11 13:08:12 2006
From: lukethulin at netscape.net (Luke Thulin)
Date: Thu, 11 May 2006 07:08:12 -0400
Subject: [Pw_forum] Confused on nat definition
In-Reply-To: <4402C34F-C6E2-4D3D-A203-05EB51232708@sissa.it>
References: <20060510185103.53008.qmail@web60316.mail.yahoo.com> <005f01c674b5$2e0ffc60$1d00a8c0@solarflare> <4402C34F-C6E2-4D3D-A203-05EB51232708@sissa.it>
Message-ID: <44631B1C.9090906@netscape.net>

Thank you for the clarification, it makes fine sense to me.  The 
original problem I am having is that I'm trying to input the titania 
antase structure.  It is fairly complicated, but I found a source, 
http://cst-www.nrl.navy.mil/lattice/struk/c5.html,  that spells it out 
clearly.  It says it's a body-centered tetragonal and lists six basis 
vectors, presumably to describe the peroidicity of the space group that 
it is in.  So I figured that ibrav = 7 for bct, nat = 6, and use the 
basis vectors to enter the atomic coordinates.  Yet when I view my 
structure with XCrysDen, it looks nothing like anatase.  I then started 
looking at the examples and how they are input and starting thinking 
that my choice for nat was wrong (based on all the discussion about 
Si).  I did see that the internet source and PWscf do not use the same 
vectors to describe the bct lattice, I thought that they were equivalent 
and resulted in the same volume of the unit cell, though I will double 
check.  Sorry about all this, but I wouldn't have asked if I hadn't 
spend so much time trying to enter this structure already.

Thanks for your help,
Luke
.
baroni at sissa.it wrote:

> OK. Let me try to shed some light on this unexpectedly obscure point.
>
> On May 11, 2006, at 6:41 AM, Huiqun Zhou wrote:
>
>> Hmm ..., because there are 4 lattice points, that are (0,0,0) and 3 
>> end points of the
>> basis vectors as stated, in a diamond structure, and there are two 
>> primitive elements
>> which are located at (0,0,0) and (1/4,1/4,1/4) for this structure, so 
>> there should be
>> 4 x 2 = 8 atoms in a conventional unit cell of diamond structure.
>
>
> I am sorry, Huiqun, but I am afraid that this argument is wrong and 
> confusing. What do you mean by "4 lattice points"? Of course, every 
> triplet of vectors originating at (0,0,0) has one common point (the 
> origin), and 3 end points. AND SO WHAT?
>
>> But because of periodicity, theoretically you can choose either only 
>> the primitive
>> elements (2 atoms) in an unit cell with appropriatelly selected large 
>> number of k points
>> for calculation, or the whole member (8 atoms) in an unit cell with 
>> less k points for
>> calculation. I may be wrong, please correct me.
>
>
> Although it is certainly true that, for a same system, doing a 
> calculation with a larger conventional unit cell would require a 
> smaller number of k points, I feel that it is confusing to try to 
> explain a property of the real-space lattice with something that has 
> only to do with the way calculations are technically done (the number 
> of k points).
>
> Let me try to clarify the matter as much as I can.
>
> 1) The Bravais lattice of the diamond structure is face-centered cubic 
> (FCC). A basis for this Bravais lattice is, e.g., (1/2,1/2,0), 
> (1/2,0,1/2), (0,1/2,1/2).
>
> 2) Each elementary cell of this FCC lattice contains two equivalent 
> atoms, located at (000)+R(l,m,n) and (1/4,1/4,1/4)+R(l,m,n), where 
> R(l,m,n) = l*(1/2,1/2,0) + m*(1/2,0,1/2) + n*(0,1/2,1/2) is a generic 
> point of the Bravais lattice (l,m,n being integer numbers).
>
> 3) The above basis for the Bravais lattice defines a unit cell of 
> *minimum volume*. Of ourse, one is free to choose as a basis any 
> triplet of integer linear (and linearly independent) combinations of 
> the minimum basis. For instance, one could choose: (1,0,0) = 
> (1/2,1/2,0)+(1/2,0,1/2)-(0,1/2,1/2), (0,1,0)= [guess what?], and 
> (0,0,1)=[guess what?]. These three vectors form a basis for the simple 
> cubic (SC) Bravais lattice. Calculate the volume of the original FCC 
> unit cell (defined at point 1 above) and compare it with the volume of 
> the SC unit cell considered here (hint: the volume of a parallelepiped 
> is the absolute value of the determinant of the 3x3 matrix whose 
> columns are the coordinates of the three vectors which form the edges 
> of the parallelepiped). You will find that the the volume of the SC 
> cell is 4 times larger than the volume of the FCC cell. This means 
> that if you choose to describe the diamond lattice as SC instead of 
> FCC (which you are free to do), the unit cell will contain 4x2 instead 
> of 2 atoms. You see where the factor 4 comes from? Nothing to do with 
> the "number of lattice points" (whatever these lattice points may be).
>
> 4) Now, the volume of the Brillouin zone (BZ) is (2\pi)^3 divided by 
> the volume of the unit cell. Hence, the volume of the SC BZ is 1/4 the 
> volume of the FCC BZ. That's why, in order to sample a SC BZ with the 
> same accuracy of an FCC cell you will need 1/4 as many k points. But 
> this is another story which has little to do with the main point of 
> this thread.
>
> Hope this clarifies a little bit the muddy waters
>
> Stefano Baroni
>
>>
>> Huiqun Zhou
>>
>> ----- Original Message ----- From: "Eyvaz Isaev" 
>> <eyvaz_isaev at yahoo.com <mailto:eyvaz_isaev at yahoo.com>>
>> To: <pw_forum at pwscf.org <mailto:pw_forum at pwscf.org>>
>> Sent: Thursday, May 11, 2006 2:51 AM
>> Subject: Re: [Pw_forum] Confused on nat definition
>>
>>
>>> Hi,
>>>
>>>> This may seem silly, but I'm confused as to what
>>>> exactly the number of
>>>> atoms in a unit cell is (nat).  For example, the
>>>> Silicon example says
>>>> that nat is only two, yet a diamond structure such
>>>> as this should have
>>>> much more than two atoms per unit cell.
>>>
>>>
>>> Let us consider the diamond case. If you choose as
>>> basis vectors next 3 vectors (which are the standard
>>> choice)
>>>
>>> 1/2, 1/2, 0
>>> 1/2, 0  , 1/2
>>> 0  , 1/2, 1/2
>>>
>>> you have only 2 atoms in the unit cell
>>> (parallelepiped) spanned by these vectors:
>>> 0,     0, 0
>>> 1/4, 1/4, 1/4
>>>
>>> If you decide to choose as basis vectors next 3 ones
>>>
>>> 1 0 0
>>> 0 1 0
>>> 0 0 1
>>>
>>> you have 8 atoms in the unit cell which is now a cub.
>>> If your choice is the latter  for CaF2 structure you
>>> will have 12 atoms, but using the former - only 3.
>>>
>>> So, number of atoms (nat) in a unit cell depends on
>>> your unit cell choice defined by 3 basis vectors.
>>>
>>>> Is the definition of nat the  number of basis
>>>> vectors?
>>>
>>> To me it is not so clear, but see above.
>>>
>>> Bests,
>>> Eyvaz.
>>>
>>>> _______________________________________________
>>>> Pw_forum mailing list
>>>> Pw_forum at pwscf.org <mailto:Pw_forum at pwscf.org>
>>>> http://www.democritos.it/mailman/listinfo/pw_forum
>>>>
>>>
>>>
>>> __________________________________________________
>>> Do You Yahoo!?
>>> Tired of spam?  Yahoo! Mail has the best spam protection around
>>> http://mail.yahoo.com
>>> _______________________________________________
>>> Pw_forum mailing list
>>> Pw_forum at pwscf.org <mailto:Pw_forum at pwscf.org>
>>> http://www.democritos.it/mailman/listinfo/pw_forum
>>>
>>
>> _______________________________________________
>> Pw_forum mailing list
>> Pw_forum at pwscf.org <mailto:Pw_forum at pwscf.org>
>> http://www.democritos.it/mailman/listinfo/pw_forum
>
>
> ---
> Stefano Baroni - SISSA  &  DEMOCRITOS National Simulation Center - Trieste
> [+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype)
>
> 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: /pipermail/attachments/20060511/27f53ab1/attachment.htm 

From tone.kokalj at ijs.si  Thu May 11 13:17:40 2006
From: tone.kokalj at ijs.si (Tone Kokalj)
Date: Thu, 11 May 2006 13:17:40 +0200
Subject: [Pw_forum] Confused on nat definition
In-Reply-To: <44631B1C.9090906@netscape.net>
References: <20060510185103.53008.qmail@web60316.mail.yahoo.com>
	 <005f01c674b5$2e0ffc60$1d00a8c0@solarflare>
	 <4402C34F-C6E2-4D3D-A203-05EB51232708@sissa.it>
	 <44631B1C.9090906@netscape.net>
Message-ID: <1147346260.5912.26.camel@localhost.localdomain>

On Thu, 2006-05-11 at 07:08 -0400, Luke Thulin wrote:
> Thank you for the clarification, it makes fine sense to me.  The
> original problem I am having is that I'm trying to input the titania
> antase structure.  It is fairly complicated, but I found a source,
> http://cst-www.nrl.navy.mil/lattice/struk/c5.html,  that spells it out
> clearly.  It says it's a body-centered tetragonal and lists six basis
> vectors, presumably to describe the peroidicity of the space group
> that it is in.  So I figured that ibrav = 7 for bct, nat = 6, and use
> the basis vectors to enter the atomic coordinates. Yet when I view my
> structure with XCrysDen, it looks nothing like anatase....

If you have the lattice basis vectors, you can choose ibrav=0 (means
custom lattice) and with CELL_PARAMETERS you input your basis vectors.
By doing that
you can use precisely the same "settings" as used by your source
mentioned above.

Regards, Tone



From ding at sissa.it  Thu May 11 15:55:12 2006
From: ding at sissa.it (Xunlei Ding)
Date: Thu, 11 May 2006 15:55:12 +0200
Subject: [Pw_forum] questions on wf_collect
Message-ID: <44634240.7020904@sissa.it>

Dear All,

I have some questions on wf_collect.
The calculations I did are usually first "relax", then "scf", then some 
PP calculations.
In order to use CPU more effectively, I perfer to use, say 4 CPU for the 
"relax" calculations,
2 for the scf and 1 for the PP calculations. To read the wavefile of the 
previous calculation, wf_collect= .true. should be set  for all
these calculations. Am I right?
Another question is, what is the short-coming for setting wf_collect= 
.true. , since wf_collect= .false. is the default setting? And are there 
anything should
be paid attention when using wf_collect= .true. ?

Thank you very much!

Yours sincerely,
Ding




From naromero at gmail.com  Thu May 11 16:36:55 2006
From: naromero at gmail.com (Nichols A. Romero)
Date: Thu, 11 May 2006 10:36:55 -0400
Subject: [Pw_forum] parallel scaling in PWSCF
In-Reply-To: <444FDAD3.4050002@sissa.it>
References: <6ac064b60604260941u15471f47oe6d45fee2ff6ce8@mail.gmail.com>
	 <7b6913e90604261038t25d0bb43le5c8a740e793d981@mail.gmail.com>
	 <003b01c66965$e4178d80$fcc2db80@ornl.gov>
	 <Pine.LNX.4.58.0604262209040.5061@bamboo.fisica.unimo.it>
	 <444FDAD3.4050002@sissa.it>
Message-ID: <6ac064b60605110736n6775c245yeae3ff68ed525542@mail.gmail.com>

Hi,

As I was digging through PWSCF and I noticed that in para.f90, the
maximum number of processors is coded to 128. But that there is a
switch that can be used to set the maximum to 2048.

-D__QK_USER__

This is going to sound like a dumb question, but is this safe. Will
the diagonalization routines on a gamma point calculation have issues
if one goes up to 256 or 512. I you had many k-points and
spin-polarization that wouldn't so much of an issue I think.

Any comments?

On 4/26/06, carlo sbraccia <sbraccia at sissa.it> wrote:
> Hi,
>
> beware using the CVS version of PWSCF: the parallel Davidson has not yet
> been fully tested and seems to be buggy.
> For this reason we are going to disable it until we shall be able to
> solve the problem. To avoid the serial bottleneck in the diagonalization
> you can use conjugate-gradient.
>
> carlo
>
> Andrea Ferretti wrote:
>
> >Hi everybody,
> >
> >I am currently running a Copper surface with 140 Cu atoms + a molecule...
> >the system has 1642 electrons and (due to metallicity) the calculation is
> >performed for 985 bands (few kpt, like 4)...
> >due to the 11 electrons for each Cu atom, I have a huge number of bands in
> >a (relatively) small cell, and so a (relatively) low number of PWs respect
> >to nbnd.
> >taking a look at the dimension of wfc, no problem with memory in
> >principle, even if, due to the weird
> >dimensions of the system, non-scalable memory is quite large, around 1Gb.
> >
> >on a IBM Sp5 machine I observed a severe limit in the scaling passing from
> >32 to 64 procs using both espresso 2.1.x and espresso 3.0...
> >( anyway, I succeeded in performing a "relax" calculation for the system
> >!!!! )
> >
> >as far as I know, this problem might be connected to a serial part in the
> >diagonalization which has been parallelized in the current CVS version
> >(as already pointed out by Axel)...
> >At the moment I am testing this CVS version against my system, I will let
> >you know the results as soon as possible...
> >
> >cheers
> >andrea
> >
> >
> >
> >
>
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
>


-- 
Nichols A. Romero, Ph.D.
1613 Denise Dr. Apt. D
Forest Hill, MD 21050
443-567-8328 (C)
410-306-0709 (O)


From giannozz at nest.sns.it  Thu May 11 17:09:23 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Thu, 11 May 2006 17:09:23 +0200
Subject: [Pw_forum] questions on wf_collect
In-Reply-To: <44634240.7020904@sissa.it>
References: <44634240.7020904@sissa.it>
Message-ID: <200605111709.23560.giannozz@nest.sns.it>

On Thursday 11 May 2006 15:55, Xunlei Ding wrote:

> [...] I prefer to use, say 4 CPU for the "relax" calculations, 2 for the
> scf and 1 for the PP calculations. To read the wavefile of the previous
> calculation, wf_collect= .true. should be set  for all these calculations.

wf_collect should be set when writing wavefunctions; the code should
be able to recognize the correct format when reading

> Another question is, what is the short-coming for setting 
> wf_collect=.true. , since wf_collect= .false. is the default 
> setting?

more disk usage, communications, I/O 

Paolo
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From fcleri at yahoo.it  Thu May 11 17:09:23 2006
From: fcleri at yahoo.it (fabrizio cleri)
Date: Thu, 11 May 2006 17:09:23 +0200 (CEST)
Subject: [Pw_forum] Espresso on Mac OSX 10.4
Message-ID: <20060511150923.53846.qmail@web26708.mail.ukl.yahoo.com>

Ciao to everybody,

A warning for those compiling with xlf under Mac OSX 10.4 ("Tiger"). You will most likely get the following linkage error:

 ld: Undefined symbols:
 _sprintf$LDBLStub
 _fprintf$LDBLStub
 _printf$LDBLStub

This is becoming a known problem in the Apple Developers' community, it seems to happen because these functions in Tiger are actually macros which translate into symbols in the System stub library which is not included on the link line.

In the case of a F90 compilation (I am using the IBM xlf compiler) the way around this problem is to link explicitly also the library /usr/lib/libSystemstubs.a by adding

-lSystemStubs

to the LDFLAGS in make.sys.

have fun,
fc

_______________________________________________
Fabrizio Cleri
Institut d'Electronique, Micro?lectronique et Nanotechnologie
Universit? de Science et Technologie de Lille
59652 Villeneuve d'Ascq, France
tel.: +33-320-197928
e-mail: fabrizio.cleri at isen.fr



Chiacchiera con i tuoi amici in tempo reale! 
 http://it.yahoo.com/mail_it/foot/*http://it.messenger.yahoo.com 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: /pipermail/attachments/20060511/2c17f5fd/attachment.htm 

From naromero at gmail.com  Thu May 11 17:16:00 2006
From: naromero at gmail.com (Nichols A. Romero)
Date: Thu, 11 May 2006 11:16:00 -0400
Subject: [Pw_forum] problems with cvs repository
Message-ID: <6ac064b60605110816i3f3596e3qdef56e83d613f35b@mail.gmail.com>

Hi,

Hopefully I am not doing something stupid.

I am unable to checkout the CVS version of Espresso. Has the CVS
repository been recently reconfigured somehow?

I am receiving the following error:

Logging in to :pserver:cvsanon at democritos.sissa.it:2401/home/cvs/
CVS password:
/home/cvs/: no such repository

Bests,
-- 
Nichols A. Romero, Ph.D.
1613 Denise Dr. Apt. D
Forest Hill, MD 21050
443-567-8328 (C)
410-306-0709 (O)


From giannozz at nest.sns.it  Thu May 11 17:32:36 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Thu, 11 May 2006 17:32:36 +0200
Subject: [Pw_forum] problems with cvs repository
In-Reply-To: <6ac064b60605110816i3f3596e3qdef56e83d613f35b@mail.gmail.com>
References: <6ac064b60605110816i3f3596e3qdef56e83d613f35b@mail.gmail.com>
Message-ID: <200605111732.36228.giannozz@nest.sns.it>

On Thursday 11 May 2006 17:16, Nichols A. Romero wrote:

> I am unable to checkout the CVS version of Espresso. Has the CVS
> repository been recently reconfigured somehow?

it has been reconfigured (2n+1) times, but not in the last weeks.

> Logging in to :pserver:cvsanon at democritos.sissa.it:2401/home/cvs/
> CVS password:
> /home/cvs/: no such repository

it works for me. Which repository are you trying to check out?

Paolo
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From lukethulin at netscape.net  Thu May 11 22:44:44 2006
From: lukethulin at netscape.net (Luke Thulin)
Date: Thu, 11 May 2006 16:44:44 -0400
Subject: [Pw_forum] question about   band structure
In-Reply-To: <BAY19-F20CDDC3D499CB2513BE99AAA90@phx.gbl>
References: <BAY19-F20CDDC3D499CB2513BE99AAA90@phx.gbl>
Message-ID: <4463A23C.3020303@netscape.net>

I am also looking to plot TiO2 band structure so I noticed a major problem:

celldim(1) = 3.785

That is the TiO2 lattice constant in angstroms, celldim(1) is defined in 
Bohr, so if you visualize your input structure you should probably 
notice that your lattice constants are short by a factor of .529.  The 
same issue appears for your definition of the atomic positions.
Luke

ezadshojaee at hotmail.com wrote:

> hi everyone
> i want to plot a band structure along a path  in the unit cell of 
> TiO2  & i have this error :
>
> from potinit : error #         1
>     starting and expected charges differ
>
> what is wrong with my calculation ?
> this is my input file for pw.x  :
>
> &control
>    calculation='nscf'
>    prefix='TiO2',
>    pseudo_dir = '/root/pseudo/',
>    outdir='/root/tmp/'
> /
> &system
>    ibrav= 7,  celldm(1) =3.785,   celldm(3) =2.5136,
>    nat=  6, ntyp= 2,
>    ecutwfc =200.0,
>    nbnd = 12,
> /
> &electrons
> /
> ATOMIC_SPECIES
> Ti  47.867     Ti.vdb.UPF
> O   15.9994    O.vdb.UPF
> ATOMIC_POSITIONS  { bohr }
> Ti       0.000000000  -0.946250000  -1.189250000
> Ti       0.000000000   0.946250000   1.189250000
> O       0.000000000   0.946250000  -2.593319231
> O       0.000000000  -0.946250000   2.593319231
> O       1.892500000   0.946250000  -0.214759231
> O      -1.892500000  -0.946250000   0.214759231
>
> K_POINTS {tpiba}
> 20
>   0.0 0.0 0.00  1
>   0.0 0.0 0.10  1
>   0.0 0.0 0.15  1
>   0.0 0.0 0.20  1
>   0.0 0.0 0.25  1
>   0.0 0.0 0.30  1
>   0.0 0.0 0.35  1
>   0.0 0.0 0.40  1
>   0.0 0.0 0.45  1
>   0.0 0.0 0.50  1
>   0.0 0.0 0.55  1
>   0.0 0.0 0.60  1
>   0.0 0.0 0.65  1
>   0.0 0.0 0.70  1
>   0.0 0.0 0.75  1
>   0.0 0.0 0.80  1
>   0.0 0.0 0.85  1
>   0.0 0.0 0.90  1
>   0.0 0.0 0.95  1
>   0.0 0.0 1.00  1
>
> _________________________________________________________________
> Express yourself instantly with MSN Messenger! Download today it's 
> FREE! http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/
>
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum



From eylvisaker at physics.ucdavis.edu  Fri May 12 00:46:30 2006
From: eylvisaker at physics.ucdavis.edu (Erik Ylvisaker)
Date: Thu, 11 May 2006 15:46:30 -0700
Subject: [Pw_forum] Error running pw.x
Message-ID: <00ae01c6754c$bfa84c80$17f4a8c0@ErikLaptop>

Hi, I am trying to run pw on an Intel Itanium system.  I was able to compile it without errors, but when I run it I get the following message:

##################
     nbndx  =     7  nbnd   =     7  natomwfc =    16  npwx   =     736
     nelec  =   5.00  nkb   =    13  ngl    =      92

     Initial potential from superposition of free atoms

     starting charge    4.99685, renormalised to    5.00000
     Starting wfc are atomic

 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
     from cfts_3 : error #         1
     routine called by wrong architecture
 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

     stopping ...
##################                                                  

from looking at the  cfts_3.f90 file this routine is apparently meant for  IBM : essl library or NEC sx4/5: GPFA library.  Does anyone know what I need to modify in the make.sys (or another?) file to get it to not include this file?

I tried removing the compile flag -D__USE_INTERNAL_FFTW but that gave an error on an #include <fftw.h> line.  Also __LINUX64 is defined, which I think is correct.

~Erik Ylvisaker

-------------- next part --------------
An HTML attachment was scrubbed...
URL: /pipermail/attachments/20060511/5439c235/attachment.htm 

From akohlmey at cmm.upenn.edu  Fri May 12 01:06:09 2006
From: akohlmey at cmm.upenn.edu (Axel Kohlmeyer)
Date: Thu, 11 May 2006 19:06:09 -0400
Subject: [Pw_forum] Error running pw.x
In-Reply-To: <00ae01c6754c$bfa84c80$17f4a8c0@ErikLaptop>
References: <00ae01c6754c$bfa84c80$17f4a8c0@ErikLaptop>
Message-ID: <7b6913e90605111606k2752addesec44bebb2f0662cd@mail.gmail.com>

On 5/11/06, Erik Ylvisaker <eylvisaker at physics.ucdavis.edu> wrote:

erik,

this is a (well) known problem. please check out the manual or the
following URL:
http://www.democritos.it/pipermail/pw_forum/2006-February/003662.html

regards,
    axel.
>
> Hi, I am trying to run pw on an Intel Itanium system.  I was able to compile
> it without errors, but when I run it I get the following message:
>
> ##################
>      nbndx  =     7  nbnd   =     7  natomwfc =    16  npwx   =     736
>      nelec  =   5.00  nkb   =    13  ngl    =      92
>
>      Initial potential from superposition of free atoms
>
>      starting charge    4.99685, renormalised to    5.00000
>      Starting wfc are atomic
>
> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
>      from cfts_3 : error #         1
>      routine called by wrong architecture
> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
>
>      stopping ...
> ##################
>
>
> from looking at the cfts_3.f90 file this routine is apparently meant for
> IBM : essl library or NEC sx4/5: GPFA library.  Does anyone know what I need
> to modify in the make.sys (or another?) file to get it to not include this
> file?
>
> I tried removing the compile flag -D__USE_INTERNAL_FFTW but that gave an
> error on an #include <fftw.h> line.  Also __LINUX64 is defined, which I
> think is correct.
>
> ~Erik Ylvisaker
>
>


-- 
=======================================================================
Axel Kohlmeyer   akohlmey at cmm.chem.upenn.edu   http://www.cmm.upenn.edu
  Center for Molecular Modeling   --   University of Pennsylvania
Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323
tel: 1-215-898-1582,  fax: 1-215-573-6233,  office-tel: 1-215-898-5425
=======================================================================
If you make something idiot-proof, the universe creates a better idiot.


From eylvisaker at physics.ucdavis.edu  Fri May 12 01:29:42 2006
From: eylvisaker at physics.ucdavis.edu (Erik Ylvisaker)
Date: Thu, 11 May 2006 16:29:42 -0700
Subject: [Pw_forum] Error running pw.x
References: <00ae01c6754c$bfa84c80$17f4a8c0@ErikLaptop> <7b6913e90605111606k2752addesec44bebb2f0662cd@mail.gmail.com>
Message-ID: <00c301c67552$c83a4d70$17f4a8c0@ErikLaptop>

Axel,

Thanks for the quick reply!  That fix worked.

~Erik

----- Original Message ----- 
From: "Axel Kohlmeyer" <akohlmey at vitae.cmm.upenn.edu>
To: <pw_forum at pwscf.org>
Sent: Thursday, May 11, 2006 4:06 PM
Subject: Re: [Pw_forum] Error running pw.x


> On 5/11/06, Erik Ylvisaker <eylvisaker at physics.ucdavis.edu> wrote:
>
> erik,
>
> this is a (well) known problem. please check out the manual or the
> following URL:
> http://www.democritos.it/pipermail/pw_forum/2006-February/003662.html
>
> regards,
>    axel.
>>
>> Hi, I am trying to run pw on an Intel Itanium system.  I was able to 
>> compile
>> it without errors, but when I run it I get the following message:
>>
>> ##################
>>      nbndx  =     7  nbnd   =     7  natomwfc =    16  npwx   =     736
>>      nelec  =   5.00  nkb   =    13  ngl    =      92
>>
>>      Initial potential from superposition of free atoms
>>
>>      starting charge    4.99685, renormalised to    5.00000
>>      Starting wfc are atomic
>>
>> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
>>      from cfts_3 : error #         1
>>      routine called by wrong architecture
>> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
>>
>>      stopping ...
>> ##################
>>
>>
>> from looking at the cfts_3.f90 file this routine is apparently meant for
>> IBM : essl library or NEC sx4/5: GPFA library.  Does anyone know what I 
>> need
>> to modify in the make.sys (or another?) file to get it to not include 
>> this
>> file?
>>
>> I tried removing the compile flag -D__USE_INTERNAL_FFTW but that gave an
>> error on an #include <fftw.h> line.  Also __LINUX64 is defined, which I
>> think is correct.
>>
>> ~Erik Ylvisaker
>>
>>
>
>
> -- 
> =======================================================================
> Axel Kohlmeyer   akohlmey at cmm.chem.upenn.edu   http://www.cmm.upenn.edu
>  Center for Molecular Modeling   --   University of Pennsylvania
> Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323
> tel: 1-215-898-1582,  fax: 1-215-573-6233,  office-tel: 1-215-898-5425
> =======================================================================
> If you make something idiot-proof, the universe creates a better idiot.
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum 



From sclauzer at sissa.it  Fri May 12 14:09:30 2006
From: sclauzer at sissa.it (Gabriele Sclauzero)
Date: Fri, 12 May 2006 14:09:30 +0200
Subject: [Pw_forum] read_file: problem reading .save files
Message-ID: <44647AFA.6030104@sissa.it>

Dear users,

     while using the cvs version of pwscf (althought not the latest) I 
encountered the following problem. When I run a job with pw.x saving 
files (like .wfc*, .igk*, .save ...) locally in the scratch dirs of each 
node (not accesible from the other nodes), each node will have his .wfc 
and .igk, but only one node will write and then be able to access the 
.save directory tree.
     If I subsequently run a post processing program (like pp.x, 
projwfc.x , pwcond.x ...) that uses the read_file subroutine to access 
these data files in the .save directory, ALL the processors try to read 
some file in that directory (but obviously not all of them will be able 
to see that file, if I use more than one node). The error given is:

     from read_file  : error #         1
     problem reading file /local_scratch/sclauzer/tmp/prefix.save

I also followed the suggestion (from one of pwscf developers) to print 
some debug info.
 From some processors I get:

  lexist =  T
  /local_scratch/sclauzer/tmp/prefix.save/data-file.xml

 From the others:

  lexist =  F
  /local_scratch/sclauzer/tmp/prefix.save/data-file.xml

Where the second string is the file every processor is trying to access, 
I suppose.

Gabriele



From giannozz at nest.sns.it  Fri May 12 14:42:59 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Fri, 12 May 2006 14:42:59 +0200
Subject: [Pw_forum] read_file: problem reading .save files
In-Reply-To: <44647AFA.6030104@sissa.it>
References: <44647AFA.6030104@sissa.it>
Message-ID: <200605121442.59988.giannozz@nest.sns.it>

On Friday 12 May 2006 14:09, Gabriele Sclauzero wrote:

>      from read_file  : error #         1
>      problem reading file /local_scratch/sclauzer/tmp/prefix.save 

I am quite sure this has been fixed in the meantime. Note that

- there is no guarantee that the processor reading the data file is 
the same that wrote it in a previous run. One has no control on
which processor holds which MPI process.

- you need to set wf_collect to .true. in order to save wavefunctions
into a single file

- if I remember correctly, pseudopotentials are still read by every
processor. There is no special reason for this, other than the time 
and boredom needed to implement it in the same way as the rest 
of I/O (i.e. read on one processor, broadcast all variables to all
other processors)

- there is no way to write things to disk in such a way that it 
satisfies all requirements.

Notice to everybody: when reporting problems in the CVS version,
please check out the very latest version first

Paolo
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From konstantin_kudin at yahoo.com  Fri May 12 18:11:16 2006
From: konstantin_kudin at yahoo.com (Konstantin Kudin)
Date: Fri, 12 May 2006 09:11:16 -0700 (PDT)
Subject: [Pw_forum] read_file: problem reading .save files
In-Reply-To: <200605121442.59988.giannozz@nest.sns.it>
Message-ID: <20060512161116.47825.qmail@web52008.mail.yahoo.com>

--- Paolo Giannozzi <giannozz at nest.sns.it> wrote:

> On Friday 12 May 2006 14:09, Gabriele Sclauzero wrote:
> 
> >      from read_file  : error #         1
> >      problem reading file /local_scratch/sclauzer/tmp/prefix.save 

> I am quite sure this has been fixed in the meantime.

 Actually, I encountered the same issue too with one of the very recent
CVS versions (2 days ago). While PW itself proceeds fine from the
distributed *.wfc files, the post-processing codes crash (pw2wannier90
for me). I tracked it a bit to the subroutine read_file.f90, where
things crash in these calls
  CALL pw_readfile( 'reset', ierr ) 
  CALL pw_readfile( 'dim',   ierr ) 
even when the *.save directory itself is perfectly visible. Like when
running on 2 cpus of a dual box. I ran under x86-64 Linux with mpich
1.2.7 and ifort 9.0


> Note that
> 
> - there is no guarantee that the processor reading the data file is 
> the same that wrote it in a previous run. One has no control on
> which processor holds which MPI process.

 Some queueing systems read the cpu list from a node file (
$PBS_NODEFILE under PBS ), and so the chance that the order is the same
is high.

> - you need to set wf_collect to .true. in order to save wavefunctions
> into a single file

 But then is post-processing smart enough to read the collected file
and distribute it? I could not verify this due to crashes.

> - if I remember correctly, pseudopotentials are still read by every
> processor. There is no special reason for this, other than the time 
> and boredom needed to implement it in the same way as the rest 
> of I/O (i.e. read on one processor, broadcast all variables to all
> other processors)

 But these come from a separate pseudo-directory, do they not?

 Kostya

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 


From cesards at msi.umn.edu  Fri May 12 18:16:09 2006
From: cesards at msi.umn.edu (cesards at msi.umn.edu)
Date: Fri, 12 May 2006 11:16:09 -0500 (CDT)
Subject: [Pw_forum] Re:Espresso on Mac OSX 10.4
Message-ID: <56388.128.101.191.226.1147450569.squirrel@www.msi.umn.edu>

There is an easy way to compile pwscf with xlf in Mac OS X 10.4: Set gcc
version to 3.3. This is done with the command

sudo gcc_select 3.3

If you get the message "Error trying to determine current cc version (got
)" change the order of directory in you PATH variable in order to make
/opt/ibm/... to appear at its end. The xlc alias to cc will stop working,
but ass soon you have set gcc version you can change PATH to its normal
direcory order.

Cesar.



From proffess at yandex.ru  Fri May 12 18:21:28 2006
From: proffess at yandex.ru (Sergey Lisenkov)
Date: Fri, 12 May 2006 20:21:28 +0400 (MSD)
Subject: [Pw_forum] numbering the wavefunction files
Message-ID: <4464B608.000007.04358@pantene.yandex.ru>

 Dear developers,

  Sorry for the posting silly question. I see that if the total number of cpus less than 9, wfc/atwfc files are numbered as *.wfc1, *.wfc2, ... , but when cpus > 10 , than all files are numbered as *.wfc01, *.wfc02 , ... . What cause this thing? I use my script for copying the wfc,atwfc files to each node , and that thing make me unhappy.

 Thanks,
   Sergey


From giannozz at nest.sns.it  Fri May 12 18:49:13 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Fri, 12 May 2006 18:49:13 +0200
Subject: [Pw_forum] numbering the wavefunction files
In-Reply-To: <4464B608.000007.04358@pantene.yandex.ru>
References: <4464B608.000007.04358@pantene.yandex.ru>
Message-ID: <200605121849.13598.giannozz@nest.sns.it>

On Friday 12 May 2006 18:21, Sergey Lisenkov wrote:

>   I see that if the total number of  cpus less than 9, 
>   wfc/atwfc files are numbered as *.wfc1, *.wfc2, ... , 
>   but  when cpus > 10 , than all files are numbered as 
>   *.wfc01, *.wfc02 , ... .  What cause this thing? 

explicit instructions. I think I actually did it this way,
for no special reasons other than aestethics.

>   I use my script for copying the wfc,atwfc files to 
> each node , and that thing make me unhappy.

use a better script, or instruct the code to collect
wavefunctions into a single file, or change the way
the suffix is calculated: see variable nd_nmbr in
PW/startup.f90

Paolo
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From giannozz at nest.sns.it  Fri May 12 19:32:59 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Fri, 12 May 2006 19:32:59 +0200
Subject: [Pw_forum] read_file: problem reading .save files
In-Reply-To: <20060512161116.47825.qmail@web52008.mail.yahoo.com>
References: <20060512161116.47825.qmail@web52008.mail.yahoo.com>
Message-ID: <200605121932.59769.giannozz@nest.sns.it>

On Friday 12 May 2006 18:11, Konstantin Kudin wrote:

>  Actually, I encountered the same issue too with one of the very 
>  recent CVS versions (2 days ago).

very recent = 2 minutes ago, not 2 days ago

> things crash in these calls
>   CALL pw_readfile( 'reset', ierr )
>   CALL pw_readfile( 'dim',   ierr )
> even when the *.save directory itself is perfectly visible.

so it is a different problem from the one I was referring to (that occurs
when the .save directory is on a local filesystem, not visible to other
processors). What happens exactly and under which circumstances?

> > - you need to set wf_collect to .true. in order to save wavefunctions
> > into a single file
>
>  But then is post-processing smart enough to read the collected file
> and distribute it? I could not verify this due to crashes.

last time I tried it was

> > - [...] pseudopotentials are still read by every processor [...]
>
>  But these come from a separate pseudo-directory, do they not?

if you restart from the .save directory, they don't; they come from
the PP files saved in the restart directory

Paolo
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From akohlmey at vitae.cmm.upenn.edu  Sat May 13 00:07:44 2006
From: akohlmey at vitae.cmm.upenn.edu (Axel Kohlmeyer)
Date: Fri, 12 May 2006 18:07:44 -0400 (EDT)
Subject: [Pw_forum] parallel scaling in PWSCF
In-Reply-To: <6ac064b60605110736n6775c245yeae3ff68ed525542@mail.gmail.com>
Message-ID: <Pine.LNX.4.44.0605111047490.12008-100000@localhost.localdomain>

On Thu, 11 May 2006, Nichols A. Romero wrote:

NR> Hi,

hi nichols,

NR> As I was digging through PWSCF and I noticed that in para.f90, the
NR> maximum number of processors is coded to 128. But that there is a
NR> switch that can be used to set the maximum to 2048.
NR> 
NR> -D__QK_USER__

this flag is for the cray xt3 (and red storm) machines.

NR> This is going to sound like a dumb question, but is this safe. Will
NR> the diagonalization routines on a gamma point calculation have issues
NR> if one goes up to 256 or 512. I you had many k-points and
NR> spin-polarization that wouldn't so much of an issue I think.

exactly. however, the xt3 is a special beast. since you have no
swap, no local disc, things are very different. the diagonalization
(at least when i cranked up the max nodes number) didn't work well
with too many nodes per k-point, but due to the machine design
issues, you need to go across many nodes, you you have additional
memory to buffer the wavefunction file access (and get a decent
performance at all).

for most machines however it makes no sense to go up to so many
nodes, hence the define. nevertheless, the beauty of open source
software lies in the fact, that you have access to the source code
and can tune it to your own needs. feel free to crank up the 
maximum allowed number of processors and tell up about your 
experiences.

best regards,
    axel.


NR> 
NR> Any comments?


NR> 
NR> On 4/26/06, carlo sbraccia <sbraccia at sissa.it> wrote:
NR> > Hi,
NR> >
NR> > beware using the CVS version of PWSCF: the parallel Davidson has not yet
NR> > been fully tested and seems to be buggy.
NR> > For this reason we are going to disable it until we shall be able to
NR> > solve the problem. To avoid the serial bottleneck in the diagonalization
NR> > you can use conjugate-gradient.
NR> >
NR> > carlo
NR> >
NR> > Andrea Ferretti wrote:
NR> >
NR> > >Hi everybody,
NR> > >
NR> > >I am currently running a Copper surface with 140 Cu atoms + a molecule...
NR> > >the system has 1642 electrons and (due to metallicity) the calculation is
NR> > >performed for 985 bands (few kpt, like 4)...
NR> > >due to the 11 electrons for each Cu atom, I have a huge number of bands in
NR> > >a (relatively) small cell, and so a (relatively) low number of PWs respect
NR> > >to nbnd.
NR> > >taking a look at the dimension of wfc, no problem with memory in
NR> > >principle, even if, due to the weird
NR> > >dimensions of the system, non-scalable memory is quite large, around 1Gb.
NR> > >
NR> > >on a IBM Sp5 machine I observed a severe limit in the scaling passing from
NR> > >32 to 64 procs using both espresso 2.1.x and espresso 3.0...
NR> > >( anyway, I succeeded in performing a "relax" calculation for the system
NR> > >!!!! )
NR> > >
NR> > >as far as I know, this problem might be connected to a serial part in the
NR> > >diagonalization which has been parallelized in the current CVS version
NR> > >(as already pointed out by Axel)...
NR> > >At the moment I am testing this CVS version against my system, I will let
NR> > >you know the results as soon as possible...
NR> > >
NR> > >cheers
NR> > >andrea
NR> > >
NR> > >
NR> > >
NR> > >
NR> >
NR> > _______________________________________________
NR> > Pw_forum mailing list
NR> > Pw_forum at pwscf.org
NR> > http://www.democritos.it/mailman/listinfo/pw_forum
NR> >
NR> 
NR> 
NR> 

-- 
=======================================================================
Axel Kohlmeyer   akohlmey at cmm.chem.upenn.edu   http://www.cmm.upenn.edu
   Center for Molecular Modeling   --   University of Pennsylvania
Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323
tel: 1-215-898-1582,  fax: 1-215-573-6233,  office-tel: 1-215-898-5425
=======================================================================
If you make something idiot-proof, the universe creates a better idiot.




From lanhaiping at gmail.com  Sat May 13 11:46:36 2006
From: lanhaiping at gmail.com (lan haiping)
Date: Sat, 13 May 2006 17:46:36 +0800
Subject: [Pw_forum] crystal structures' transformation
Message-ID: <c92002fa0605130246w2e70ec21t615a3f7a8e368417@mail.gmail.com>

Hi,All.

 There are several  choices for us to describe   crystal structure  in
PWSCF. So i would like to know
whether there are options or tools for  the crystal structures
transformation , e.g . crystal -->  angstrom ?

 Regards,

Hai-Ping
-------------- next part --------------
An HTML attachment was scrubbed...
URL: /pipermail/attachments/20060513/5063310b/attachment.htm 

From hsd22 at hermes.cam.ac.uk  Sat May 13 18:27:47 2006
From: hsd22 at hermes.cam.ac.uk (H.S.Domingos)
Date: Sat, 13 May 2006 17:27:47 +0100 (BST)
Subject: [Pw_forum] Number of iterations
In-Reply-To: <200605091112.13678.giannozz@nest.sns.it>
References: <200605091112.13678.giannozz@nest.sns.it>
Message-ID: <Pine.LNX.4.64.0605131725470.9437@hermes-1.csi.cam.ac.uk>

Dear All,

could you please tell me how to change the maximum number of iterations 
that vc-relax will do.

at line :

Entering Dynamics;  it = ...

I am getting allways 50 iterations. How can I change that ?

Thank you very much in advance,

Helder


=======================================================================
|  Dr. Helder S. Domingos                                             |
|                                                                     |
| INESC Microsyst & Nanotechnol, Lisbon, P-1000 Portugal              |
| and                                                                 |
| R&D unit for  Molecular Chemical Physics                            |
| Chemistry Department, University of Coimbra                         |
=======================================================================




From eyvaz_isaev at yahoo.com  Sat May 13 18:44:11 2006
From: eyvaz_isaev at yahoo.com (Eyvaz Isaev)
Date: Sat, 13 May 2006 09:44:11 -0700 (PDT)
Subject: [Pw_forum] Number of iterations
In-Reply-To: <Pine.LNX.4.64.0605131725470.9437@hermes-1.csi.cam.ac.uk>
Message-ID: <20060513164411.41228.qmail@web60322.mail.yahoo.com>

Hi,

Try nstep=100 in &control section

Bests,
Eyvaz.

--- "H.S.Domingos" <hsd22 at hermes.cam.ac.uk> wrote:

> Dear All,
> 
> could you please tell me how to change the maximum
> number of iterations 
> that vc-relax will do.
> 
> at line :
> 
> Entering Dynamics;  it = ...
> 
> I am getting allways 50 iterations. How can I change
> that ?
> 
> Thank you very much in advance,
> 
> Helder
> 
> 
>
=======================================================================
> |  Dr. Helder S. Domingos                           
>                  |
> |                                                   
>                  |
> | INESC Microsyst & Nanotechnol, Lisbon, P-1000
> Portugal              |
> | and                                               
>                  |
> | R&D unit for  Molecular Chemical Physics          
>                  |
> | Chemistry Department, University of Coimbra       
>                  |
>
=======================================================================
> 
> 
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
> 


__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 


From giannozz at nest.sns.it  Sun May 14 22:40:33 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Sun, 14 May 2006 22:40:33 +0200
Subject: [Pw_forum] crystal structures' transformation
In-Reply-To: <c92002fa0605130246w2e70ec21t615a3f7a8e368417@mail.gmail.com>
References: <c92002fa0605130246w2e70ec21t615a3f7a8e368417@mail.gmail.com>
Message-ID: <200605142240.33157.giannozz@nest.sns.it>

On Saturday 13 May 2006 11:46, lan haiping wrote:

>  There are several  choices for us to describe   crystal structure
>  in  PWSCF. So i would like to know whether there are options or
>  tools for  the crystal structures transformation , e.g . 
>  crystal -->  angstrom ? 

one "transformation" is that at the startup the code will print 
coordinates in units of the lattice parameter, no matter how you
read them. 

Paolo
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From ding at sissa.it  Mon May 15 15:41:54 2006
From: ding at sissa.it (Xunlei Ding)
Date: Mon, 15 May 2006 15:41:54 +0200
Subject: [Pw_forum] How to fix some atoms in the ph.x calculation
Message-ID: <44688522.5090303@sissa.it>

Dear all,

An molecule is adsorbed on a surface. I just want to calculate the 
vibrational frequencies of the molecule (and a few atoms in the first 
layer).
How can I fix the other atoms in the ph.x calculation? I found "nrapp", 
"nat_todo" in the INPUT_PH, but I don't know whether they can be used 
for this purpose, and how to use them. Thank you!

Yours sincerely,
Ding


From brsahu at physics.utexas.edu  Mon May 15 21:54:30 2006
From: brsahu at physics.utexas.edu (Bhagawan Sahu)
Date: Mon, 15 May 2006 14:54:30 -0500 (CDT)
Subject: [Pw_forum] SOC
In-Reply-To: <1144912293.4002.7.camel@dhpc-5-26.sissa.it>
Message-ID: <Pine.LNX.4.44.0605151428300.6412-100000@linux2.ph.utexas.edu>

Andrea,

 In the relativistic PP generation, How the the orbitals, to be pseudozied 
, are filled in in case of Pt?

I am trying to generate the Bi PP using the example of Pt given with the 
distribution.

for example, in pt.in 

7
5D  3  2  4.00  0.00  2.10  2.40  1.50
5D  3  2  0.00 -0.20  2.10  2.40  1.50
5D  3  2  4.00  0.00  2.10  2.40  2.50
5D  3  2  0.00 -0.20  2.10  2.40  2.50
6P  2  1  0.00 -0.00  3.30  3.30  0.50
6P  2  1  0.00 -0.00  3.40  3.40  1.50
6S  1  0  2.00  0.00  2.60  2.60  0.50

6s is local so is placed at the end.

four 5d states---> for j=3/2, there are 4 states so max occupation in 
these states is 4.0 (for up-spin) so down-spin occupation is zero. 
These counts for first two 5d orbitals (one for up and one for down, 
j=3/2).

for j=5/2 there are 6 states, so maximum occupation in these states is 6.0 
but we have 4 remaining electrons so next two 5d states are for j=5/2 
(one up and one down j=5/2).

What I do not understand is the 6P states. for l=1 we have j=1/2 or 3/2 

for j=1/2, we have two states with maximum occupation 2.0 
for j=3/2, we have four states with maxmum occupation 4.0 

For Bi, there are 3 p-electrons  so 
the p config is

6P  2  1  2.00 -0.00  3.30  3.30  0.50
6P  2  1  0.00 -0.00  3.40  3.40  0.50
6P  2  1  1.00 -0.00  3.30  3.30  1.50
6P  2  1  0.00 -0.00  3.40  3.40  1.50

Is this true?

the first two are for j=1/2 and last two are for j=3/2

Also, since pseudotype =3 (ultrasoft), are the entries for rcut(meant for 
norm conserving PP's) in Pt.in meaningful? I guess they are ignored in the 
ultrasoft generation.

Sahu


four 5d states--> why there are four of these?

 

On Thu, 13 Apr 2006, Andrea Dal Corso wrote:

> Yes the SO term is only in the pseudo-potential but valence states
> interact with a fully relativistic pseudo-potential at each scf cycle. 
> We make an approximation because we neglect the spin-orbit coupling
> outside the core radius where however the effect is expected to be
> small.
> 
> Andrea
> 
> 
> On Wed, 2006-04-12 at 12:43 -0500, Bhagawan Sahu wrote:
> > Dear pwscf users,
> > 
> >   Is it true that
> > 
> > the  bulk spin-orbit coupling (SOC) calculation is done using a 
> > relativistic 
> > PP for the atom/ion (RRKJ scheme and Andrea Dal Carso's PRB paper) and the 
> > SOC effect is not included self-consistently on the valence states during 
> > the scf cycle?
> > 
> > Sahu 
> > 
> > _______________________________________________
> > Pw_forum mailing list
> > Pw_forum at pwscf.org
> > http://www.democritos.it/mailman/listinfo/pw_forum
> 



From Conor.Hogan at roma2.infn.it  Mon May 15 23:42:16 2006
From: Conor.Hogan at roma2.infn.it (Conor Hogan)
Date: Mon, 15 May 2006 22:42:16 +0100 (GMT+0100)
Subject: [Pw_forum] Dipole correction in vacuum, polar slabs
In-Reply-To: <Pine.LNX.4.64.0605131725470.9437@hermes-1.csi.cam.ac.uk>
Message-ID: <Pine.OSF.3.95a.1060515224106.25821A-100000@axroma2.roma2.infn.it>

Dear developers/users,

I have a few questions and observations on the dipole correction to the
spurious macroscopic field that arises in polar slabs+vacuum calculations,
as implemented in PWSCF. As far as I can see, the only place this has been
discussed before has been in this forum:
http://www.democritos.it/pipermail/pw_forum/2004-July/001135.html , in
some papers (e.g. Journal of Chemical Physics (2004) 120, p9934), and in
the source code PW/add_efield.f90 PW/compute_dip.f90.

1) Reference to this facility seem to be completely removed from the
documentation (distinct, of course, from the facility to add an external
electric field, tefield =.true.). In particular, the fact that the flag
dipfield = .true. is needed is only mentioned in the source (took me a
while to find that out!). Is this a conscious decision? Is the dipole
correction now "unsupported"? Was it ever "supported"? Is the routine
known to be buggy? One of the main reasons I started using PWSCF is for
this reason...

2) I see some weird behaviour when the dipole field is used, and I'd like
to know if it's a bug, or if it arises from a real physical mechanism.
What I've observed is that convergence depends on the relative position of
the slab within the supercell: if the slab is roughly symmetric about
zero, and the dipole layer located at about 0.5 of the cell, the
correction works, and convergence is achieved; if the slab and layer are
shifted, the convergence is never reached and the total energy shoots to
large positive values.
At first appearance, this to me is a bug. However, it also occurred to me
that it might just be related to the old problematic definition of the
dipole moment within the infinite lattice. Has anyone any idea why this
behaviour arises? If the potential converges, can I trust the calculation?

Thanks, and best regards to all.
Conor


----

Dr. Conor Hogan
Dipartimento di Fisica e CNR-INFM
Universita' di Roma "Tor Vergata"
Tel: +39 06 72594548
Fax: +39 06 2023507
http://www.fisica.uniroma2.it/~cmtheo-group/

The early bird gets the worm, but the second mouse gets the cheese. - S. Wright

If you go through a lot of hammers each month, I don't think it necessarily 
means you're a hard worker. It may just mean that you have a lot to learn 
about proper hammer maintenance - J. Handey



From giannozz at nest.sns.it  Tue May 16 09:30:07 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Tue, 16 May 2006 09:30:07 +0200
Subject: [Pw_forum] SOC
In-Reply-To: <Pine.LNX.4.44.0605151428300.6412-100000@linux2.ph.utexas.edu>
References: <Pine.LNX.4.44.0605151428300.6412-100000@linux2.ph.utexas.edu>
Message-ID: <200605160930.07475.giannozz@nest.sns.it>

On Monday 15 May 2006 21:54, Bhagawan Sahu wrote:

> for example, in pt.in
>
> 7
> 5D  3  2  4.00  0.00  2.10  2.40  1.50
> 5D  3  2  0.00 -0.20  2.10  2.40  1.50
> 5D  3  2  4.00  0.00  2.10  2.40  2.50
> 5D  3  2  0.00 -0.20  2.10  2.40  2.50
> 6P  2  1  0.00 -0.00  3.30  3.30  0.50
> 6P  2  1  0.00 -0.00  3.40  3.40  1.50
> 6S  1  0  2.00  0.00  2.60  2.60  0.50
>
> 6s is local so is placed at the end.
>
> four 5d states---> for j=3/2, there are 4 states so max occupation in
> these states is 4.0 (for up-spin) so down-spin occupation is zero.
> These counts for first two 5d orbitals (one for up and one for down,
> j=3/2).
> for j=5/2 there are 6 states, so maximum occupation in these states is 6.0
> but we have 4 remaining electrons so next two 5d states are for j=5/2
> (one up and one down j=5/2).

note that there are two 5d bound states (pseudized at the energy 
of the bound state), one for j=3/2 (4 electrons) and one for j=5/2 
(also 4 electrons); plus two unbound states, pseudized at the specified 
energy (-0.2 Ry), fifth column, as explained in the documentation:

 ener = energy (Ry) used to pseudize the corresponding state
             if 0.d0, use the one-electron energy of the all-electron state
             Do not use 0.d0 for unbound states!

> What I do not understand is the 6P states. for l=1 we have j=1/2 or 3/2
>
> for j=1/2, we have two states with maximum occupation 2.0
> for j=3/2, we have four states with maxmum occupation 4.0
>
> For Bi, there are 3 p-electrons  so
> the p config is
>
> 6P  2  1  2.00 -0.00  3.30  3.30  0.50
> 6P  2  1  0.00 -0.00  3.40  3.40  0.50
> 6P  2  1  1.00 -0.00  3.30  3.30  1.50
> 6P  2  1  0.00 -0.00  3.40  3.40  1.50
>
> Is this true?

if you remove the second and fourth lines, this is the ground 
state configuration. The unbound states may not be needed;
if needed, you have to specify an appropriate pseudization
energy, not 0.0 (see above)

> Also, since pseudotype =3 (ultrasoft), are the entries for rcut (meant 
> for norm conserving PP's) in Pt.in meaningful? I guess they are ignored 
> in the ultrasoft generation.

this code generate first a norm-conserving PP, and on top of that,
an ultrasoft PP, so rcut is definitely meaningful.

Paolo
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From dalcorso at sissa.it  Tue May 16 09:32:13 2006
From: dalcorso at sissa.it (Andrea Dal Corso)
Date: Tue, 16 May 2006 09:32:13 +0200
Subject: [Pw_forum] SOC
In-Reply-To: <Pine.LNX.4.44.0605151428300.6412-100000@linux2.ph.utexas.edu>
References: <Pine.LNX.4.44.0605151428300.6412-100000@linux2.ph.utexas.edu>
Message-ID: <1147764733.4006.19.camel@dhpc-5-41.sissa.it>

On Mon, 2006-05-15 at 14:54 -0500, Bhagawan Sahu wrote:
> Andrea,
> 
>  In the relativistic PP generation, How the the orbitals, to be pseudozied 
> , are filled in in case of Pt?
> 
> I am trying to generate the Bi PP using the example of Pt given with the 
> distribution.
> 
> for example, in pt.in 
> 
> 7
> 5D  3  2  4.00  0.00  2.10  2.40  1.50
> 5D  3  2  0.00 -0.20  2.10  2.40  1.50
> 5D  3  2  4.00  0.00  2.10  2.40  2.50
> 5D  3  2  0.00 -0.20  2.10  2.40  2.50
> 6P  2  1  0.00 -0.00  3.30  3.30  0.50
> 6P  2  1  0.00 -0.00  3.40  3.40  1.50
> 6S  1  0  2.00  0.00  2.60  2.60  0.50
> 
> 6s is local so is placed at the end.
> 
> four 5d states---> for j=3/2, there are 4 states so max occupation in 
> these states is 4.0 (for up-spin) so down-spin occupation is zero. 
> These counts for first two 5d orbitals (one for up and one for down, 
> j=3/2).

There are not up or down states. The states are eigenfunctions of J^2
and J_z (total angular momentum). They are not eigenfunctions of the
spin operator. When j=3/2, j_z can be -3/2, -1/2, 1/2, -3/2 so there are
four states that can contain a maximum of 4 electrons. The other 4
electrons go in j=5/2 states. 

> 
> for j=5/2 there are 6 states, so maximum occupation in these states is 6.0 
> but we have 4 remaining electrons so next two 5d states are for j=5/2 
> (one up and one down j=5/2).
> 
> What I do not understand is the 6P states. for l=1 we have j=1/2 or 3/2 
> 
> for j=1/2, we have two states with maximum occupation 2.0 
> for j=3/2, we have four states with maxmum occupation 4.0 
> 
> For Bi, there are 3 p-electrons  so 
> the p config is
> 
> 6P  2  1  2.00 -0.00  3.30  3.30  0.50
> 6P  2  1  0.00 -0.00  3.40  3.40  0.50
> 6P  2  1  1.00 -0.00  3.30  3.30  1.50
> 6P  2  1  0.00 -0.00  3.40  3.40  1.50
> 

I think here the configuration is OK. However in US there are two states
for each channel (these are not spin-up and spin-down but two different
energies). Here you set all energies to 0.00 (that means: take the
eigenvalue) and this is not correct. The second energy should be
different from zero:

6P  2  1  2.00 -0.00  3.30  3.30  0.50
6P  2  1  0.00 Energy  3.40  3.40  0.50
6P  2  1  1.00 -0.00  3.30  3.30  1.50
6P  2  1  0.00 Energy  3.40  3.40  1.50

Which energy I do not know. It depends on the atom and is a matter of
cooking. Often an energy slightly higher than the eigenvalue works. In
other cases you need to look at the logarithmic derivative to guess a
reasonable value.


> Is this true?
> 
> the first two are for j=1/2 and last two are for j=3/2
> 
> Also, since pseudotype =3 (ultrasoft), are the entries for rcut(meant for 
> norm conserving PP's) in Pt.in meaningful? I guess they are ignored in the 
> ultrasoft generation.
> 

They are not ignored. The ultrasoft pseudization is done starting from
the norm conserving wavefunctions.


> Sahu
> 
> 
> four 5d states--> why there are four of these?
> 
>  

two values of j and two energies for each j.


> 
> On Thu, 13 Apr 2006, Andrea Dal Corso wrote:
> 
> > Yes the SO term is only in the pseudo-potential but valence states
> > interact with a fully relativistic pseudo-potential at each scf cycle. 
> > We make an approximation because we neglect the spin-orbit coupling
> > outside the core radius where however the effect is expected to be
> > small.
> > 
> > Andrea
> > 
> > 
> > On Wed, 2006-04-12 at 12:43 -0500, Bhagawan Sahu wrote:
> > > Dear pwscf users,
> > > 
> > >   Is it true that
> > > 
> > > the  bulk spin-orbit coupling (SOC) calculation is done using a 
> > > relativistic 
> > > PP for the atom/ion (RRKJ scheme and Andrea Dal Carso's PRB paper) and the 
> > > SOC effect is not included self-consistently on the valence states during 
> > > the scf cycle?
> > > 
> > > Sahu 
> > > 
> > > _______________________________________________
> > > Pw_forum mailing list
> > > Pw_forum at pwscf.org
> > > http://www.democritos.it/mailman/listinfo/pw_forum
> > 
> 
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
-- 
Andrea Dal Corso                    Tel. 0039-040-3787428
SISSA, Via Beirut 2/4               Fax. 0039-040-3787528
34014 Trieste (Italy)               e-mail: dalcorso at sissa.it



From guido.fratesi at unimi.it  Tue May 16 12:35:31 2006
From: guido.fratesi at unimi.it (Guido Fratesi)
Date: Tue, 16 May 2006 12:35:31 +0200 (CEST)
Subject: [Pw_forum] Re: Dipole correction in vacuum, polar slabs
Message-ID: <Pine.LNX.4.64.0605161235060.26957@localhost.localdomain>

> 1) Reference to this facility seem to be completely removed from the
> documentation (distinct, of course, from the facility to add an external
> electric field, tefield =.true.). In particular, the fact that the flag
> dipfield = .true. is needed is only mentioned in the source (took me a
> while to find that out!). Is this a conscious decision? Is the dipole
> correction now "unsupported"? Was it ever "supported"? Is the routine
> known to be buggy? One of the main reasons I started using PWSCF is for
> this reason...

I don't know about "supported"/"unsupported".
However, I'm currently using dipole correction for calculations of adatom 
adsorption on a single side of a slab, and I found workfunction changes in good 
agreement (that is, slightly overestimated) with experimental findings.

To use it, I'm setting
&CONTROL
   tefield  = .true.  !!this to add some kind of electric field
   dipfield = .true.  !!this to add also the field given by the dipole
/
&SYSTEM
   edir=3       !!field along z (I'v never tried x or y)
   eamp=0       !!add only dipole field, no additional electric field
   emaxpos=0.45 !!position of the maximum of the added potential
                !!depends on how I've built the supercell
   eopreg =0.10 !!see below
/

The added potential is increasing from [emaxpos+eamp-1]*celldm(edir)*alat to 
[emaxpos]*celldm(edir)*alat with slope eamp, then decreasing to 
[emaxpos+eamp]*celldm(edir)*alat to recover periodic boundary conditions for 
the potential. The region in which the potential is rapidly decreasing to 
conserve periodicity should be in vacuum where an additional electric field 
takes no significant effect.

> 2) I see some weird behaviour when the dipole field is used, and I'd like
> to know if it's a bug, or if it arises from a real physical mechanism.
> What I've observed is that convergence depends on the relative position of
> the slab within the supercell: if the slab is roughly symmetric about
> zero, and the dipole layer located at about 0.5 of the cell, the
> correction works, and convergence is achieved; if the slab and layer are
> shifted, the convergence is never reached and the total energy shoots to
> large positive values.

Have you checked the position of the fast decrease of the potential, if it is 
in vacuum or not?
Notice: the electric field can also be plotted by using plot_num=12 in the 
postprocessing program pp.x.

> At first appearance, this to me is a bug. However, it also occurred to me
> that it might just be related to the old problematic definition of the
> dipole moment within the infinite lattice. Has anyone any idea why this
> behaviour arises? If the potential converges, can I trust the calculation?

The dipole along z is given correctly, since it is computed in 
PW/compute_dip.f90 as

1/Omega \int \rho(r) (r-r0) d^3r

where r0=(0,0,z0) and z0 is in the middle of the region in which the electric 
field is applied (which should therefore be in the slab region). As far as see 
from the code, the dipole field correction is implemented only for field 
direction along z.

Hope this helps.

Regards,
Guido Fratesi



From gaoguoying1981 at yahoo.com.cn  Tue May 16 15:10:43 2006
From: gaoguoying1981 at yahoo.com.cn (=?gb2312?q?=D3=A2=B9=FA=20=B8=DF?=)
Date: Tue, 16 May 2006 21:10:43 +0800 (CST)
Subject: [Pw_forum] some problems on the phonon
Message-ID: <20060516131043.22781.qmail@web15201.mail.cnb.yahoo.com>

Dear all,
  I'm learning on the PWSCF and meet with some problems when I calculate the phonon spectrum except Gramm.
  It is succesfull when calculating one qpoint but wrong many qpoints.Errors appear as follows 
  In nscf file
  task # 1
  from phq_init:error# 1
  wrong order of k points
or  
  from read_namelists:error # 225
  reading namelist cell
   
    The second question, when I calculate the phonon frequencies under pressure,it appears error in nscf file like this :
  from ptoinit :error
  starting and expected charges differ  
  
  I'm thirsty for your reply.Thanks.
Best regards

		
---------------------------------
????????-3.5G???20M??? 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: /pipermail/attachments/20060516/2d966f51/attachment.htm 

From gaoguoying1981 at yahoo.com.cn  Tue May 16 15:09:45 2006
From: gaoguoying1981 at yahoo.com.cn (=?gb2312?q?=D3=A2=B9=FA=20=B8=DF?=)
Date: Tue, 16 May 2006 21:09:45 +0800 (CST)
Subject: [Pw_forum] some problems on the phonon
Message-ID: <20060516130945.59595.qmail@web15211.mail.cnb.yahoo.com>

Dear all,
  I'm learning on the PWSCF and meet with some problems when I calculate the phonon spectrum except Gramm.
  It is succesfull when calculating one qpoint but wrong many qpoints.Errors appear as follows 
  In nscf file
  task # 1
  from phq_init:error# 1
  wrong order of k points
or  
  from read_namelists:error # 225
  reading namelist cell
   
    The second question, when I calculate the phonon frequencies under pressure,it appears error in nscf file like this :
  from ptoinit :error
  starting and expected charges differ  
  
  I'm thirsty for your reply.Thanks.
Best regards

		
---------------------------------
????????-3.5G???20M??? 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: /pipermail/attachments/20060516/5d6cec0a/attachment.htm 

From wlyim at puccini.che.pitt.edu  Tue May 16 18:29:27 2006
From: wlyim at puccini.che.pitt.edu (wlyim at puccini.che.pitt.edu)
Date: Tue, 16 May 2006 12:29:27 -0400 (EDT)
Subject: [Pw_forum] NEC MathKeisan
Message-ID: <Pine.LNX.4.44.0605161220390.11882-100000@puccini.che.pitt.edu>

Dear all,

I am trying to test PWscf on NEC machine, using MathKeisan library. But in
the make.sys, which was generated by "configure.old sxcross", it stated
that the fft library of Mathkeisan (version <= 1.4, zzfft3d) is buggy.

I searched the forum for a while but there is not updated information 
about this issue. Does anyone know whether this fft bug has been fixed? 
Otherwise I would use jmfft.

Many thanks in advance!

Best regards,
William



From brsahu at physics.utexas.edu  Tue May 16 19:22:18 2006
From: brsahu at physics.utexas.edu (Bhagawan Sahu)
Date: Tue, 16 May 2006 12:22:18 -0500 (CDT)
Subject: [Pw_forum] SOC
In-Reply-To: <1147764733.4006.19.camel@dhpc-5-41.sissa.it>
Message-ID: <Pine.LNX.4.44.0605161219490.13417-100000@linux2.ph.utexas.edu>

Andrea and paolo,

thanx

I missed the point that spin-polarized pseudo's can not be generated in 
this scheme So the confusion about spin-up and down-states.

Is there a way to take in to account the semi-core states in the 
pseudo-valence set? 

Sahu

On Tue, 16 May 2006, Andrea Dal Corso wrote:

> On Mon, 2006-05-15 at 14:54 -0500, Bhagawan Sahu wrote:
> > Andrea,
> > 
> >  In the relativistic PP generation, How the the orbitals, to be pseudozied 
> > , are filled in in case of Pt?
> > 
> > I am trying to generate the Bi PP using the example of Pt given with the 
> > distribution.
> > 
> > for example, in pt.in 
> > 
> > 7
> > 5D  3  2  4.00  0.00  2.10  2.40  1.50
> > 5D  3  2  0.00 -0.20  2.10  2.40  1.50
> > 5D  3  2  4.00  0.00  2.10  2.40  2.50
> > 5D  3  2  0.00 -0.20  2.10  2.40  2.50
> > 6P  2  1  0.00 -0.00  3.30  3.30  0.50
> > 6P  2  1  0.00 -0.00  3.40  3.40  1.50
> > 6S  1  0  2.00  0.00  2.60  2.60  0.50
> > 
> > 6s is local so is placed at the end.
> > 
> > four 5d states---> for j=3/2, there are 4 states so max occupation in 
> > these states is 4.0 (for up-spin) so down-spin occupation is zero. 
> > These counts for first two 5d orbitals (one for up and one for down, 
> > j=3/2).
> 
> There are not up or down states. The states are eigenfunctions of J^2
> and J_z (total angular momentum). They are not eigenfunctions of the
> spin operator. When j=3/2, j_z can be -3/2, -1/2, 1/2, -3/2 so there are
> four states that can contain a maximum of 4 electrons. The other 4
> electrons go in j=5/2 states. 
> 
> > 
> > for j=5/2 there are 6 states, so maximum occupation in these states is 6.0 
> > but we have 4 remaining electrons so next two 5d states are for j=5/2 
> > (one up and one down j=5/2).
> > 
> > What I do not understand is the 6P states. for l=1 we have j=1/2 or 3/2 
> > 
> > for j=1/2, we have two states with maximum occupation 2.0 
> > for j=3/2, we have four states with maxmum occupation 4.0 
> > 
> > For Bi, there are 3 p-electrons  so 
> > the p config is
> > 
> > 6P  2  1  2.00 -0.00  3.30  3.30  0.50
> > 6P  2  1  0.00 -0.00  3.40  3.40  0.50
> > 6P  2  1  1.00 -0.00  3.30  3.30  1.50
> > 6P  2  1  0.00 -0.00  3.40  3.40  1.50
> > 
> 
> I think here the configuration is OK. However in US there are two states
> for each channel (these are not spin-up and spin-down but two different
> energies). Here you set all energies to 0.00 (that means: take the
> eigenvalue) and this is not correct. The second energy should be
> different from zero:
> 
> 6P  2  1  2.00 -0.00  3.30  3.30  0.50
> 6P  2  1  0.00 Energy  3.40  3.40  0.50
> 6P  2  1  1.00 -0.00  3.30  3.30  1.50
> 6P  2  1  0.00 Energy  3.40  3.40  1.50
> 
> Which energy I do not know. It depends on the atom and is a matter of
> cooking. Often an energy slightly higher than the eigenvalue works. In
> other cases you need to look at the logarithmic derivative to guess a
> reasonable value.
> 
> 
> > Is this true?
> > 
> > the first two are for j=1/2 and last two are for j=3/2
> > 
> > Also, since pseudotype =3 (ultrasoft), are the entries for rcut(meant for 
> > norm conserving PP's) in Pt.in meaningful? I guess they are ignored in the 
> > ultrasoft generation.
> > 
> 
> They are not ignored. The ultrasoft pseudization is done starting from
> the norm conserving wavefunctions.
> 
> 
> > Sahu
> > 
> > 
> > four 5d states--> why there are four of these?
> > 
> >  
> 
> two values of j and two energies for each j.
> 
> 
> > 
> > On Thu, 13 Apr 2006, Andrea Dal Corso wrote:
> > 
> > > Yes the SO term is only in the pseudo-potential but valence states
> > > interact with a fully relativistic pseudo-potential at each scf cycle. 
> > > We make an approximation because we neglect the spin-orbit coupling
> > > outside the core radius where however the effect is expected to be
> > > small.
> > > 
> > > Andrea
> > > 
> > > 
> > > On Wed, 2006-04-12 at 12:43 -0500, Bhagawan Sahu wrote:
> > > > Dear pwscf users,
> > > > 
> > > >   Is it true that
> > > > 
> > > > the  bulk spin-orbit coupling (SOC) calculation is done using a 
> > > > relativistic 
> > > > PP for the atom/ion (RRKJ scheme and Andrea Dal Carso's PRB paper) and the 
> > > > SOC effect is not included self-consistently on the valence states during 
> > > > the scf cycle?
> > > > 
> > > > Sahu 
> > > > 
> > > > _______________________________________________
> > > > Pw_forum mailing list
> > > > Pw_forum at pwscf.org
> > > > http://www.democritos.it/mailman/listinfo/pw_forum
> > > 
> > 
> > _______________________________________________
> > Pw_forum mailing list
> > Pw_forum at pwscf.org
> > http://www.democritos.it/mailman/listinfo/pw_forum
> 



From eyvaz_isaev at yahoo.com  Tue May 16 19:34:02 2006
From: eyvaz_isaev at yahoo.com (Eyvaz Isaev)
Date: Tue, 16 May 2006 10:34:02 -0700 (PDT)
Subject: [Pw_forum] some problems on the phonon
In-Reply-To: <20060516130945.59595.qmail@web15211.mail.cnb.yahoo.com>
Message-ID: <20060516173403.37115.qmail@web60318.mail.yahoo.com>

Hi,

Answers to these questions were done in the forum
before. Please search mails in the forum. 
Nevertheless I will try to give short answers again

>   In nscf file
>   task # 1
>   from phq_init:error# 1
>   wrong order of k points
In my experience it is caused by inproperly finished
nscf-calculations. Check your nscf.out file.

> or  
>   from read_namelists:error # 225
>   reading namelist cell
There is definitely an error in your input file. Check
carefully &cell section of your input file and compare
with keywords in INPUT_PW file in /Doc.

>     The second question, when I calculate the phonon
> frequencies under pressure,it appears error in nscf
> file like this :
>   from ptoinit :error
>   starting and expected charges differ  
So interesting. You did not finish any phonon
calculation and then started to study phonons under
pressure.

I remember there was such a kind question in the
forum, too. One possible error is that you try to
specify number of electrons, but number of output
electrons is not the same. More likely there is a
mistake in your input file. 
   
Bests,
Eyvaz.

>   I'm thirsty for your reply.Thanks.
> Best regards
> 
> 		
> ---------------------------------
> ????????????????-3.5G??????20M?????? 


__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 


From giannozz at nest.sns.it  Tue May 16 21:39:45 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Tue, 16 May 2006 21:39:45 +0200
Subject: [Pw_forum] some problems on the phonon
In-Reply-To: <20060516131043.22781.qmail@web15201.mail.cnb.yahoo.com>
References: <20060516131043.22781.qmail@web15201.mail.cnb.yahoo.com>
Message-ID: <200605162139.45728.giannozz@nest.sns.it>

On Tuesday 16 May 2006 15:10, ?? ? wrote:

> when I calculate the phonon frequencies under pressure,
> it appears error in nscf file like this : 
>   from potinit :error  
>   starting and expected charges differ

you are starting from a charge density file that is not 
consistent with the system you are trying to solve.
This has nothing to do with "phonons under pressure"

Paolo
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From giannozz at nest.sns.it  Tue May 16 21:43:47 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Tue, 16 May 2006 21:43:47 +0200
Subject: [Pw_forum] SOC
In-Reply-To: <Pine.LNX.4.44.0605161219490.13417-100000@linux2.ph.utexas.edu>
References: <Pine.LNX.4.44.0605161219490.13417-100000@linux2.ph.utexas.edu>
Message-ID: <200605162143.47073.giannozz@nest.sns.it>

On Tuesday 16 May 2006 19:22, Bhagawan Sahu wrote:

> Is there a way to take in to account the semi-core states in the
> pseudo-valence set?

this is an old problem. In principle you can; in practise it is hard
to get anything close to a decent pseudopotential. I am just 
trying with cobalt, with less than moderate success until now

Paolo
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From prasenjit.jnc at gmail.com  Wed May 17 09:34:27 2006
From: prasenjit.jnc at gmail.com (Prasenjit Ghosh)
Date: Wed, 17 May 2006 13:04:27 +0530
Subject: [Pw_forum] parallel restart error
Message-ID: <627e0ffa0605170034u56c7f5b9pfe1e070a5d65a03a@mail.gmail.com>

hello everyone.

when i am restarting a job in parallel, i am getting the foll. error msg :


     nbndx  =  1500  nbnd   =   750  natomwfc =   999  npwx   =    2227
     nelec  = 1221.00 nkb   =  1998  ngl    =   20482
forrtl: severe (47): write to READONLY file, unit 16, file /tmp/filename.igk
Image              PC        Routine            Line        Source
pw.x               0830CC30  Unknown               Unknown  Unknown
pw.x               0830C494  Unknown               Unknown  Unknown
pw.x               082D34B9  Unknown               Unknown  Unknown
pw.x               082979B8  Unknown               Unknown  Unknown


so far there has been no such problem.  i have checked the free space
and the read write permissions in the \tmp directory.  there is enough
space and proper permissions.  The machine details :Sun Fire V60x; OS-
RHE Linux 3.0: 20 processors (Xeon @2.8 GHz): 20 GB memory- i am
running on 14 procs with only PW parallelization.
can anyone help me?

thanks in anticipation!
best regards,
Prasenjit.

-- 
PRASENJIT GHOSH,
Ph. D STUDENT,
THEORETICAL SCIENCES UNIT,
JNCASR,
BANGALORE-560064,
INDIA.

PHONE:
OFFICE : +91-80-22082835/34
MOBILE:  +91-9880519401


From giannozz at nest.sns.it  Wed May 17 10:02:37 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Wed, 17 May 2006 10:02:37 +0200
Subject: [Pw_forum] parallel restart error
In-Reply-To: <627e0ffa0605170034u56c7f5b9pfe1e070a5d65a03a@mail.gmail.com>
References: <627e0ffa0605170034u56c7f5b9pfe1e070a5d65a03a@mail.gmail.com>
Message-ID: <200605171002.37637.giannozz@nest.sns.it>

On Wednesday 17 May 2006 09:34, Prasenjit Ghosh wrote:

> when i am restarting a job in parallel

it is quite tricky

> i am getting the foll. error msg : [...]
> forrtl: severe (47): write to READONLY file, unit 16, file /tmp/filename.igk

the message doesn't leave any space to doubt: you are trying to write 
to a file that is not writable. Where and why, it is hard to say.

Paolo
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From prasenjit at jncasr.ac.in  Wed May 17 11:05:17 2006
From: prasenjit at jncasr.ac.in (PRASENJIT GHOSH)
Date: Wed, 17 May 2006 14:35:17 +0530 (IST)
Subject: [Pw_forum] parallel restart error
Message-ID: <53274.202.41.111.151.1147856717.squirrel@202.41.111.151>

hello everyone.

when i am restarting a job in parallel, i am getting the foll. error msg :


    nbndx  =  1500  nbnd   =   750  natomwfc =   999  npwx   =    2227
    nelec  = 1221.00 nkb   =  1998  ngl    =   20482
forrtl: severe (47): write to READONLY file, unit 16, file /tmp/filename.igk
Image              PC        Routine            Line        Source
pw.x               0830CC30  Unknown               Unknown  Unknown
pw.x               0830C494  Unknown               Unknown  Unknown
pw.x               082D34B9  Unknown               Unknown  Unknown
pw.x               082979B8  Unknown               Unknown  Unknown


so far there has been no such problem.  i have checked the free space
and the read write permissions in the \tmp directory.  there is enough
space and proper permissions.  The machine details :Sun Fire V60x; OS-
RHE Linux 3.0: 20 processors (Xeon @2.8 GHz): 20 GB memory- i am
running on 14 procs with only PW parallelization.
can anyone help me?

thanks in anticipation!
best regards,
Prasenjit.


PRASENJIT GHOSH,
Ph. D STUDENT,
THEORETICAL SCIENCES UNIT,
JAWAHARLAL NEHRU CENTRE,
JAKKUR P. O.,
BANGALORE - 560064,
INDIA.

PHONE:+91-80-22082835 (OFFICE)
            +91 9880519401 (MOBILE)


From degironc at sissa.it  Wed May 17 13:17:08 2006
From: degironc at sissa.it (degironc)
Date: Wed, 17 May 2006 13:17:08 +0200
Subject: [Pw_forum] parallel restart error
In-Reply-To: <53274.202.41.111.151.1147856717.squirrel@202.41.111.151>
References: <53274.202.41.111.151.1147856717.squirrel@202.41.111.151>
Message-ID: <446B0634.3030907@sissa.it>

PRASENJIT GHOSH wrote:

>forrtl: severe (47): write to READONLY file, unit 16, file /tmp/filename.igk
>  
>
maybe file /tmp/filename.igk already exists and does not belong to you ... ?
usually is bettre to create your own temporarey directory in the scratch 
area.

stefano



From prasenjit.jnc at gmail.com  Wed May 17 12:24:24 2006
From: prasenjit.jnc at gmail.com (Prasenjit Ghosh)
Date: Wed, 17 May 2006 15:54:24 +0530
Subject: [Pw_forum] parallel restart error
In-Reply-To: <446B0634.3030907@sissa.it>
References: <53274.202.41.111.151.1147856717.squirrel@202.41.111.151>
	 <446B0634.3030907@sissa.it>
Message-ID: <627e0ffa0605170324i6c8e27cao9a9e3f2095213718@mail.gmail.com>

Dear  Paolo & Stefano,

Thanks a lot . Actually the ownership of the files were different & hence it
was unable to write.

I have changed the ownership & now its running fine.

Prasenjit.

On 5/17/06, degironc <degironc at sissa.it> wrote:
>
> PRASENJIT GHOSH wrote:
>
> >forrtl: severe (47): write to READONLY file, unit 16, file
> /tmp/filename.igk
> >
> >
> maybe file /tmp/filename.igk already exists and does not belong to you ...
> ?
> usually is bettre to create your own temporarey directory in the scratch
> area.
>
> stefano
>
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
>



-- 
PRASENJIT GHOSH,
Ph. D STUDENT,
THEORETICAL SCIENCES UNIT,
JNCASR,
BANGALORE-560064,
INDIA.

PHONE:
OFFICE : +91-80-22082835/34
MOBILE:  +91-9880519401
-------------- next part --------------
An HTML attachment was scrubbed...
URL: /pipermail/attachments/20060517/4cbe161c/attachment.htm 

From Conor.Hogan at roma2.infn.it  Wed May 17 18:22:27 2006
From: Conor.Hogan at roma2.infn.it (Conor Hogan)
Date: Wed, 17 May 2006 17:22:27 +0100 (GMT+0100)
Subject: [Pw_forum] Charge density overwrite?
In-Reply-To: <53274.202.41.111.151.1147856717.squirrel@202.41.111.151>
Message-ID: <Pine.OSF.3.95a.1060517170536.23045K-100000@axroma2.roma2.infn.it>

Hi folks,
I've seen some notes in the ChangeLog stating "do not overwrite the
charge density after a non-scf calculation". However, this is not my
experience with 3.0 (*.rho file) and even now with 3.1
(*.save/charge-density.xml file). An nscf run, as far as I can see,
overwrites the charge density which could of course give problems later
in life, if one is not careful... (at least a "diff" says the file has
changed)

In particular, I use the pw_export.x facility to calculate optical spectra
later. I guess I should be using the scf .rho, .pot. and the nscf .save
files? 

Just wondering if this has been overlooked, or is done on purpose and I'm
missing something.

Thanks, and all the best
Conor


----

Dr. Conor Hogan
Dipartimento di Fisica e CNR-INFM
Universita' di Roma "Tor Vergata"
Tel: +39 06 72594548
Fax: +39 06 2023507
http://www.fisica.uniroma2.it/~cmtheo-group/

The early bird gets the worm, but the second mouse gets the cheese. - S. Wright

If you go through a lot of hammers each month, I don't think it necessarily 
means you're a hard worker. It may just mean that you have a lot to learn 
about proper hammer maintenance - J. Handey



From giannozz at nest.sns.it  Wed May 17 18:48:28 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Wed, 17 May 2006 18:48:28 +0200
Subject: [Pw_forum] Charge density overwrite?
In-Reply-To: <Pine.OSF.3.95a.1060517170536.23045K-100000@axroma2.roma2.infn.it>
References: <Pine.OSF.3.95a.1060517170536.23045K-100000@axroma2.roma2.infn.it>
Message-ID: <200605171848.28827.giannozz@nest.sns.it>

On Wednesday 17 May 2006 18:22, Conor Hogan wrote:

> I've seen some notes in the ChangeLog stating "do not overwrite the
> charge density after a non-scf calculation". However, this is not my
> experience with 3.0 (*.rho file) and even now with 3.1
> (*.save/charge-density.xml file). An nscf run, as far as I can see,
> overwrites the charge density 

too bad: this shouldn't happen. Not that it is harmful in any way:
the charge density that is written to file should be the scf one 
that has just been read. If you end up with a strange charge
density, please comment line 56 in PW/punch.f90 :
  IF ( .NOT. lscf .AND. .NOT. lbands ) CALL sum_band()
which recalculates the charge density. I think this is completely
useless now.

> which could of course give problems later in life, if one is not 
> careful... (at least a "diff" says the file has changed)

a "diff" on a binary file is not especially significant

> In particular, I use the pw_export.x facility to calculate optical
> spectra later. I guess I should be using the scf .rho, .pot. and 
> the nscf .save files?

you should use the nscf .save files, which should contain all you need.
The potential is no longer written, since it can be recalculated from the
charge density

Paolo
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From brsahu at physics.utexas.edu  Wed May 17 18:59:50 2006
From: brsahu at physics.utexas.edu (Bhagawan Sahu)
Date: Wed, 17 May 2006 11:59:50 -0500 (CDT)
Subject: [Pw_forum] SOC
In-Reply-To: <1147764733.4006.19.camel@dhpc-5-41.sissa.it>
Message-ID: <Pine.LNX.4.44.0605171156330.22503-100000@linux2.ph.utexas.edu>

Andrea,

  Is the PAW feature in 'ld1 atomic' implemented with spin-orbit (ie 
rel=2). Is there a write-up explaining the implementation of PAW in RRKJ 
ultra-soft scheme? I guess there is not yet any publication. If any, pl. 
point it to me. 

Sahu

On Tue, 16 May 2006, Andrea Dal Corso wrote:

> On Mon, 2006-05-15 at 14:54 -0500, Bhagawan Sahu wrote:
> > Andrea,
> > 
> >  In the relativistic PP generation, How the the orbitals, to be pseudozied 
> > , are filled in in case of Pt?
> > 
> > I am trying to generate the Bi PP using the example of Pt given with the 
> > distribution.
> > 
> > for example, in pt.in 
> > 
> > 7
> > 5D  3  2  4.00  0.00  2.10  2.40  1.50
> > 5D  3  2  0.00 -0.20  2.10  2.40  1.50
> > 5D  3  2  4.00  0.00  2.10  2.40  2.50
> > 5D  3  2  0.00 -0.20  2.10  2.40  2.50
> > 6P  2  1  0.00 -0.00  3.30  3.30  0.50
> > 6P  2  1  0.00 -0.00  3.40  3.40  1.50
> > 6S  1  0  2.00  0.00  2.60  2.60  0.50
> > 
> > 6s is local so is placed at the end.
> > 
> > four 5d states---> for j=3/2, there are 4 states so max occupation in 
> > these states is 4.0 (for up-spin) so down-spin occupation is zero. 
> > These counts for first two 5d orbitals (one for up and one for down, 
> > j=3/2).
> 
> There are not up or down states. The states are eigenfunctions of J^2
> and J_z (total angular momentum). They are not eigenfunctions of the
> spin operator. When j=3/2, j_z can be -3/2, -1/2, 1/2, -3/2 so there are
> four states that can contain a maximum of 4 electrons. The other 4
> electrons go in j=5/2 states. 
> 
> > 
> > for j=5/2 there are 6 states, so maximum occupation in these states is 6.0 
> > but we have 4 remaining electrons so next two 5d states are for j=5/2 
> > (one up and one down j=5/2).
> > 
> > What I do not understand is the 6P states. for l=1 we have j=1/2 or 3/2 
> > 
> > for j=1/2, we have two states with maximum occupation 2.0 
> > for j=3/2, we have four states with maxmum occupation 4.0 
> > 
> > For Bi, there are 3 p-electrons  so 
> > the p config is
> > 
> > 6P  2  1  2.00 -0.00  3.30  3.30  0.50
> > 6P  2  1  0.00 -0.00  3.40  3.40  0.50
> > 6P  2  1  1.00 -0.00  3.30  3.30  1.50
> > 6P  2  1  0.00 -0.00  3.40  3.40  1.50
> > 
> 
> I think here the configuration is OK. However in US there are two states
> for each channel (these are not spin-up and spin-down but two different
> energies). Here you set all energies to 0.00 (that means: take the
> eigenvalue) and this is not correct. The second energy should be
> different from zero:
> 
> 6P  2  1  2.00 -0.00  3.30  3.30  0.50
> 6P  2  1  0.00 Energy  3.40  3.40  0.50
> 6P  2  1  1.00 -0.00  3.30  3.30  1.50
> 6P  2  1  0.00 Energy  3.40  3.40  1.50
> 
> Which energy I do not know. It depends on the atom and is a matter of
> cooking. Often an energy slightly higher than the eigenvalue works. In
> other cases you need to look at the logarithmic derivative to guess a
> reasonable value.
> 
> 
> > Is this true?
> > 
> > the first two are for j=1/2 and last two are for j=3/2
> > 
> > Also, since pseudotype =3 (ultrasoft), are the entries for rcut(meant for 
> > norm conserving PP's) in Pt.in meaningful? I guess they are ignored in the 
> > ultrasoft generation.
> > 
> 
> They are not ignored. The ultrasoft pseudization is done starting from
> the norm conserving wavefunctions.
> 
> 
> > Sahu
> > 
> > 
> > four 5d states--> why there are four of these?
> > 
> >  
> 
> two values of j and two energies for each j.
> 
> 
> > 
> > On Thu, 13 Apr 2006, Andrea Dal Corso wrote:
> > 
> > > Yes the SO term is only in the pseudo-potential but valence states
> > > interact with a fully relativistic pseudo-potential at each scf cycle. 
> > > We make an approximation because we neglect the spin-orbit coupling
> > > outside the core radius where however the effect is expected to be
> > > small.
> > > 
> > > Andrea
> > > 
> > > 
> > > On Wed, 2006-04-12 at 12:43 -0500, Bhagawan Sahu wrote:
> > > > Dear pwscf users,
> > > > 
> > > >   Is it true that
> > > > 
> > > > the  bulk spin-orbit coupling (SOC) calculation is done using a 
> > > > relativistic 
> > > > PP for the atom/ion (RRKJ scheme and Andrea Dal Carso's PRB paper) and the 
> > > > SOC effect is not included self-consistently on the valence states during 
> > > > the scf cycle?
> > > > 
> > > > Sahu 
> > > > 
> > > > _______________________________________________
> > > > Pw_forum mailing list
> > > > Pw_forum at pwscf.org
> > > > http://www.democritos.it/mailman/listinfo/pw_forum
> > > 
> > 
> > _______________________________________________
> > Pw_forum mailing list
> > Pw_forum at pwscf.org
> > http://www.democritos.it/mailman/listinfo/pw_forum
> 



From brsahu at physics.utexas.edu  Wed May 17 21:48:11 2006
From: brsahu at physics.utexas.edu (Bhagawan Sahu)
Date: Wed, 17 May 2006 14:48:11 -0500 (CDT)
Subject: [Pw_forum] compilation error
Message-ID: <Pine.LNX.4.44.0605171437460.23425-100000@linux2.ph.utexas.edu>

Dear Paolo and pwscf users,

I am trying to compile espresso-3.1 (release 17th May 2006) on 64-bit 
linux (EM64T) cluster( intel9.1 compiler, mkl8.1). ./configure step is 
successful and created the following make.sys file:

---
CC             = icc
MPICC          = mpicc
CFLAGS         = -O3 $(DFLAGS) $(IFLAGS)
CPP            = icc -E
CPPFLAGS       =  $(DFLAGS) $(IFLAGS)
F90            = mpif90
MPIF90         = mpif90
F90FLAGS       = $(FFLAGS) $(FDFLAGS) $(IFLAGS) $(MODFLAGS)
F90FLAGS_NOOPT = $\(FFLAGS_NOOPT) $(FDFLAGS) $(IFLAGS) $(MODFLAGS)
F77            = ifort
MPIF77         = mpif77
FFLAGS         = -O
FFLAGS_NOOPT   = -O0
LD             = mpif90
LDFLAGS        =
AR             = ar
ARFLAGS        = ruv
RANLIB         = echo
BLAS_LIBS      = -L/opt/intel/mkl8.1/lib/em64t -lmkl_em64t -lmkl_solver 
-lguide
LAPACK_LIBS    = -L/opt/intel/mkl8.1/lib/em64t -lmkl_lapack
FFT_LIBS       = -L/opt/apps/fftw/fftw-3.1.1/lib -lfftw3
MPI_LIBS       =
MASS_LIBS      =

# -----------------------------
# application-specific settings

# See include/defs.h.README for a list of precompilation options
# (possible arguments to -D or -U) and their meaning
DFLAGS         =  -D__LINUX -D__INTEL -D__FFTW -D__MPI -D__PARA
FDFLAGS        = $(DFLAGS)
IFLAGS         = -I../opt/apps/fftw/fftw-3.1.1/include -I../include
MODFLAGS       = -I. -I../Modules -I../PW -I../PH -I../iotk/src -I../CPV

LIBOBJS        = ../flib/ptools.a ../flib/flib.a ../clib/clib.a 
../iotk/src/libiotk.a
# LIBS must contain the location of all needed external libraries
LIBS           = $(LAPACK_LIBS) $(BLAS_LIBS) $(FFT_LIBS) $(MPI_LIBS) 
$(MASS_LIBS)
# MYLIB can be one of the following (depending on LIBS):
# blas       : compile the local copy of blas routines
# lapack     : compile the local copy of lapack routines
# blas_and_lapack : all of the above - use this for a quick test
#              or if you don't have an optimized blas/lapack library
# lapack_ibm : compile only lapack routines not present in IBM ESSL
#              use this together with IBM ESSL
# lapack_t3e : compile only lapack routines not present in T3E scilib
#              use this together with T3E scilib
# lapack_mkl : compile only lapack routines not present in Intel MKL
#              use this together with Intel MKL
MYLIB          = lapack_mkl
----

However, the 'make all' step gives the following compilation error:

mpif90 -O   -D__LINUX -D__INTEL -D__FFTW -D__MPI -D__PARA 
-I../opt/apps/fftw/fftw-3.1.1/include -I../include  -I. -I../Modules 
-I../PW -I../PH -I../iotk/src -I../CPV -c iotk_attr+COMPLEX1_0.F90 -o 
iotk_attr+COMPLEX1_0.o
fortcom: Error: iotk_attr.spp, line 144: Syntax error, found 
END-OF-STATEMENT when expecting one of: , )
    call iotk_strcat(string,trim(adjustl(tmpval))
-------------------------------------------------^
fortcom: Error: iotk_attr.spp, line 144: There is no matching specific 
subroutine for this generic subroutine call.   [IOTK_STRCAT]
    call iotk_strcat(string,trim(adjustl(tmpval))
---------^
compilation aborted for iotk_attr+COMPLEX1_0.F90 (code 1)
make[2]: *** [iotk_attr+COMPLEX1_0.o] Error 1
make[2]: Leaving directory `/home/utexas/ee/eefz340/espresso-3.1/iotk/src'
make[1]: *** [libiotk.a] Error 2
make[1]: Leaving directory `/home/utexas/ee/eefz340/espresso-3.1/iotk'
make: *** [libiotk] Error 2


Sahu



From giannozz at nest.sns.it  Thu May 18 09:22:45 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Thu, 18 May 2006 09:22:45 +0200
Subject: [Pw_forum] compilation error
In-Reply-To: <Pine.LNX.4.44.0605171437460.23425-100000@linux2.ph.utexas.edu>
References: <Pine.LNX.4.44.0605171437460.23425-100000@linux2.ph.utexas.edu>
Message-ID: <200605180922.45994.giannozz@nest.sns.it>

On Wednesday 17 May 2006 21:48, Bhagawan Sahu wrote:

> I am trying to compile espresso-3.1 (release 17th May 2006) 
> on 64-bit linux (EM64T) cluster( intel9.1 compiler, mkl8.1). 
> ./configure step is successful and created the following 
> make.sys file: 

which uses icc as external preprocessor, and this doesn't work
because it assumes that // is a C++ comment and cuts whatever
follows. You should use the fortran preprocessor instead. The 
first lines of the make.sys file should look like (beware the tabs)

.SUFFIXES :
.SUFFIXES : .o .c .f .f90
.f90.o:
        $(MPIF90) $(F90FLAGS) -c $<
.f.o:
        $(F77) $(FFLAGS) -c $<
.c.o:
        $(CC) $(CFLAGS) -c $<

and the lines

> F90FLAGS       = $(FFLAGS) $(FDFLAGS) $(IFLAGS) $(MODFLAGS)
> F90FLAGS_NOOPT = $(FFLAGS_NOOPT) $(FDFLAGS) $(IFLAGS) $(MODFLAGS)

should be

> F90FLAGS       = $(FFLAGS) -nomodule -fpp  $(FDFLAGS) $(IFLAGS) $(MODFLAGS)
> F90FLAGS_NOOPT = $(FFLAGS_NOOPT) -nomodule -fpp $(FDFLAGS) $(IFLAGS) 
$(MODFLAGS)

Paolo
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From roma at srmp.saclay.cea.fr  Thu May 18 09:45:20 2006
From: roma at srmp.saclay.cea.fr (Guido Roma)
Date: Thu, 18 May 2006 09:45:20 +0200
Subject: [Pw_forum] NEC MathKeisan
In-Reply-To: <Pine.LNX.4.44.0605161220390.11882-100000@puccini.che.pitt.edu>
References: <Pine.LNX.4.44.0605161220390.11882-100000@puccini.che.pitt.edu>
Message-ID: <1147938320.7733.17.camel@srmp19.saclay.cea.fr>

On Tue, 2006-05-16 at 12:29 -0400, wlyim at puccini.che.pitt.edu wrote:
> Dear all,
> 
> I am trying to test PWscf on NEC machine, using MathKeisan library. But in
> the make.sys, which was generated by "configure.old sxcross", it stated
> that the fft library of Mathkeisan (version <= 1.4, zzfft3d) is buggy.
> 

Hello, 

the problem was submitted long ago to the NEC people and, as far as I
know, it should have been corrected in the 1.5 version of MathKeisan.
The way to check it is easy: if example01 is ok (Cu and Ni having smooth
FFT grid), then it is ok.
Nevertheless I suggest that you use the ASL library for FFT, which is
faster. I have just downloaded the fresh 3.1 version of espresso and it
compiled out of the box with "configure.old sxcross;make pw".  
The first examples I run are fine.
It should also work in parallel for NEB images (I tested it on some
intermediate 3.1 cvs version) with "configure.old sxpara".
It should also work (but I did not test it) for parallelism on k-points;
in this case, however, performances are probably not so satisfactory,
due to unvectorized sections of v_of_rho.

Guido

-- 
Guido Roma <roma${at}cea.fr> -- CEA-Saclay - DEN/DMN/SRMP Bat.520/130
Phone: [+33]-1-69085738 -- Fax: ...6867 -- Mobile: [+33]-6-20069085




From giannozz at nest.sns.it  Thu May 18 09:46:00 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Thu, 18 May 2006 09:46:00 +0200
Subject: [Pw_forum] Charge density overwrite?
In-Reply-To: <200605171848.28827.giannozz@nest.sns.it>
References: <Pine.OSF.3.95a.1060517170536.23045K-100000@axroma2.roma2.infn.it> <200605171848.28827.giannozz@nest.sns.it>
Message-ID: <200605180946.00686.giannozz@nest.sns.it>

On Wednesday 17 May 2006 18:48, Paolo Giannozzi wrote:

> If you end up with a strange charge density, please comment
> line 56 in PW/punch.f90 : 
>   IF ( .NOT. lscf .AND. .NOT. lbands ) CALL sum_band()
> which recalculates the charge density. I think this is completely
> useless now.

well, not really: it recalculates state occupations, so it is definitely
needed in some cases

P.
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From dkorotin at gmail.com  Thu May 18 10:39:19 2006
From: dkorotin at gmail.com (Dmitry Korotin)
Date: Thu, 18 May 2006 14:39:19 +0600
Subject: [Pw_forum] only one l state?
Message-ID: <2fd252650605180139g61617465gca740cea6165332a@mail.gmail.com>

Dear all,
This is citation from  the "Notes of pseudopotential generation" by
Paolo Giannozzi:
"if you generate a single projector PP using a configuration with
semicore states in the valence, remember that for each l only the
state with lowest n can be used to generate the PP, and that the state
with same l and higher n MUST BE EMPTY".

In the case of Pu we have to include in the valence semicore 6s states
and 7s states too. Can it be done using more than one projector per
s-channel (and for which one: 7s or 6s or both)?

Thank you in advance,
Dmitry Korotin.


From giannozz at nest.sns.it  Thu May 18 12:38:47 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Thu, 18 May 2006 12:38:47 +0200
Subject: [Pw_forum] only one l state?
In-Reply-To: <2fd252650605180139g61617465gca740cea6165332a@mail.gmail.com>
References: <2fd252650605180139g61617465gca740cea6165332a@mail.gmail.com>
Message-ID: <200605181238.47958.giannozz@nest.sns.it>

On Thursday 18 May 2006 10:39, Dmitry Korotin wrote:

> "if you generate a single projector PP using a configuration with
> semicore states in the valence, remember that for each l only the
> state with lowest n can be used to generate the PP, and that the state
> with same l and higher n MUST BE EMPTY".

this is actually no longer true, as long as you specify the correct
configuration for unscreening (detailed instructions in file INPUT_LD1)

> In the case of Pu we have to include in the valence semicore 6s states
> and 7s states too. Can it be done using more than one projector per
> s-channel (and for which one: 7s or 6s or both)?

6s plus an unbound state with energy not too far from 6s. 
In principle one should use 6s and 7s but the code is presently
unable to deal with pseudo-states with nodes

Paolo
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From dkorotin at gmail.com  Thu May 18 13:08:17 2006
From: dkorotin at gmail.com (Dmitry Korotin)
Date: Thu, 18 May 2006 17:08:17 +0600
Subject: [Pw_forum] only one l state?
In-Reply-To: <200605181238.47958.giannozz@nest.sns.it>
References: <2fd252650605180139g61617465gca740cea6165332a@mail.gmail.com>
	 <200605181238.47958.giannozz@nest.sns.it>
Message-ID: <2fd252650605180408l63a132a4o99be66a76c7d8a25@mail.gmail.com>

2006/5/18, Paolo Giannozzi <giannozz at nest.sns.it>:
> > In the case of Pu we have to include in the valence semicore 6s states
> > and 7s states too. Can it be done using more than one projector per
> > s-channel (and for which one: 7s or 6s or both)?
>
> 6s plus an unbound state with energy not too far from 6s.
> In principle one should use 6s and 7s but the code is presently
> unable to deal with pseudo-states with nodes

You mean: 6s (occ=2)+unbound state + 7s(occ=2), or only: 6s
(occ=2)+unbound state (but I want to generate ground state (not
ionic))?

Dmitry Korotin.


From giannozz at nest.sns.it  Thu May 18 15:51:47 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Thu, 18 May 2006 15:51:47 +0200
Subject: [Pw_forum] only one l state?
In-Reply-To: <2fd252650605180408l63a132a4o99be66a76c7d8a25@mail.gmail.com>
References: <2fd252650605180139g61617465gca740cea6165332a@mail.gmail.com> <200605181238.47958.giannozz@nest.sns.it> <2fd252650605180408l63a132a4o99be66a76c7d8a25@mail.gmail.com>
Message-ID: <200605181551.48034.giannozz@nest.sns.it>

On Thursday 18 May 2006 13:08, Dmitry Korotin wrote:

> > 6s plus an unbound state with energy not too far from 6s.
> > In principle one should use 6s and 7s but the code is presently
> > unable to deal with pseudo-states with nodes
>
> You mean: 6s (occ=2)+unbound state + 7s(occ=2), or only: 
> 6s  (occ=2)+unbound state (but I want to generate ground state 
> (not ionic))?

don't confuse the reference configuration with the states you use
to generate the PP. You can start from the ground state (&input
namelist), use 6s + an unbound state to generate the PP for l=0
(&inputp namelist), unscreen with the ground state configuration
(&test namelist)

Paolo
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From brsahu at physics.utexas.edu  Thu May 18 17:47:50 2006
From: brsahu at physics.utexas.edu (Bhagawan Sahu)
Date: Thu, 18 May 2006 10:47:50 -0500 (CDT)
Subject: [Pw_forum] compilation error
In-Reply-To: <200605180922.45994.giannozz@nest.sns.it>
Message-ID: <Pine.LNX.4.44.0605181037450.30911-100000@linux2.ph.utexas.edu>

Paolo,

thanx. It could get past that. Now it gives errors in 'include' files of 
fftw. I used the machine installed fftw. The version is 2.1.5. 

In the make.sys,

DFLAGS         =  -D__LINUX -D__INTEL -D__FFTW -D__MPI -D__PARA
FDFLAGS        = $(DFLAGS)
IFLAGS         = -I../opt/apps/fftw/fftw-2.1.5/include -I../include

I have loaded the fftw-related environment variables. but it gives the 
error:

make[1]: Entering directory `/home/utexas/ee/eefz340/espresso-3.1/clib'
icc -O3 -D__LINUX -D__INTEL -D__FFTW -D__MPI -D__PARA 
-I../opt/apps/fftw/fftw-2.1.5/include -I../include  -c fft_stick.c
fft_stick.c(19): catastrophic error: could not open source file "fftw.h"
  #    include <fftw.h>
                       ^
 
compilation aborted for fft_stick.c (code 4)
make[1]: *** [fft_stick.o] Error 4
make[1]: Leaving directory `/home/utexas/ee/eefz340/espresso-3.1/clib'
make: *** [libs] Error 2

It is not able to find fftw.h from /opt/apps/fftw/fftw-2.1.5/include

I checked this path. fftw.h exist and not empty. Also in the clib 
directory of espresso-3.1, fftw.h exists.

We also have fftw version 3.1.1. It's include file is fftw2.h. In order to 
include this .h file, which files in clib or other directories need to be 
changed so that it can identify fftw3.h instead of fftw.h

Sahu 

On Thu, 18 May 2006, Paolo Giannozzi wrote:

> On Wednesday 17 May 2006 21:48, Bhagawan Sahu wrote:
> 
> > I am trying to compile espresso-3.1 (release 17th May 2006) 
> > on 64-bit linux (EM64T) cluster( intel9.1 compiler, mkl8.1). 
> > ./configure step is successful and created the following 
> > make.sys file: 
> 
> which uses icc as external preprocessor, and this doesn't work
> because it assumes that // is a C++ comment and cuts whatever
> follows. You should use the fortran preprocessor instead. The 
> first lines of the make.sys file should look like (beware the tabs)
> 
> .SUFFIXES :
> .SUFFIXES : .o .c .f .f90
> .f90.o:
>         $(MPIF90) $(F90FLAGS) -c $<
> .f.o:
>         $(F77) $(FFLAGS) -c $<
> .c.o:
>         $(CC) $(CFLAGS) -c $<
> 
> and the lines
> 
> > F90FLAGS       = $(FFLAGS) $(FDFLAGS) $(IFLAGS) $(MODFLAGS)
> > F90FLAGS_NOOPT = $(FFLAGS_NOOPT) $(FDFLAGS) $(IFLAGS) $(MODFLAGS)
> 
> should be
> 
> > F90FLAGS       = $(FFLAGS) -nomodule -fpp  $(FDFLAGS) $(IFLAGS) $(MODFLAGS)
> > F90FLAGS_NOOPT = $(FFLAGS_NOOPT) -nomodule -fpp $(FDFLAGS) $(IFLAGS) 
> $(MODFLAGS)
> 
> Paolo
> 



From giannozz at nest.sns.it  Thu May 18 18:37:38 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Thu, 18 May 2006 18:37:38 +0200
Subject: [Pw_forum] compilation error
In-Reply-To: <Pine.LNX.4.44.0605181037450.30911-100000@linux2.ph.utexas.edu>
References: <Pine.LNX.4.44.0605181037450.30911-100000@linux2.ph.utexas.edu>
Message-ID: <200605181837.38799.giannozz@nest.sns.it>

On Thursday 18 May 2006 17:47, Bhagawan Sahu wrote:

> make[1]: Entering directory `/home/utexas/ee/eefz340/espresso-3.1/clib'
> icc -O3 -D__LINUX -D__INTEL -D__FFTW -D__MPI -D__PARA
> -I../opt/apps/fftw/fftw-2.1.5/include -I../include  -c fft_stick.c
> fft_stick.c(19): catastrophic error: could not open source file "fftw.h"
>   #    include <fftw.h>
>
> It is not able to find fftw.h from /opt/apps/fftw/fftw-2.1.5/include

of course it is not: you are telling the compiler to look into
../opt/apps/fftw/fftw-2.1.5/include, not /opt/apps/fftw/fftw-2.1.5/include

> Also in the clib directory of espresso-3.1, fftw.h exists.

it is the include file for the internal version of fftw and shouldn't be
loaded when using an external fftw library

> We also have fftw version 3.1.1. It's include file is fftw2.h.

you do not need it with v.3 .

P.
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From brsahu at physics.utexas.edu  Thu May 18 19:28:58 2006
From: brsahu at physics.utexas.edu (Bhagawan Sahu)
Date: Thu, 18 May 2006 12:28:58 -0500 (CDT)
Subject: [Pw_forum] compilation error
In-Reply-To: <200605181837.38799.giannozz@nest.sns.it>
Message-ID: <Pine.LNX.4.44.0605181226310.31601-100000@linux2.ph.utexas.edu>

Paolo,

I used fftw.2.1.5 and correctly put the path now 

IFLAGS         = -I/opt/apps/fftw/fftw-2.1.5/include -I../include

It compiles past 'not getting fftw.h' but it now gives error 

make[1]: Entering directory `/home/utexas/ee/eefz340/espresso-3.1/CPV'
test -n "" && ( cd .. ; make -w  || exit 1) || :
mpif90 -O   -nomodule -fpp -D__LINUX -D__INTEL -D__FFTW -D__MPI -D__PARA 
-I/opt/apps/fftw/fftw-2.1.5/include -I../include  -I. -I../Modules -I../PW 
-I../PH -I../iotk/src -I../CPV -c fftdrv.f90
fortcom: Severe: **Internal compiler error: segmentation violation signal 
raised** Please report this error along with the circumstances in which it 
occurred in a Software Problem Report.  Note: File and line given may not 
be explicit cause of this error.
 
compilation aborted for fftdrv.f90 (code 3)
make[1]: *** [fftdrv.o] Error 3
make[1]: Leaving directory `/home/utexas/ee/eefz340/espresso-3.1/CPV'
make: *** [cp] Error 2


Is this somthing reported to the sysad here or have to do with 
espresso-3.1?

Sahu

On Thu, 18 May 2006, Paolo Giannozzi wrote:

> On Thursday 18 May 2006 17:47, Bhagawan Sahu wrote:
> 
> > make[1]: Entering directory `/home/utexas/ee/eefz340/espresso-3.1/clib'
> > icc -O3 -D__LINUX -D__INTEL -D__FFTW -D__MPI -D__PARA
> > -I../opt/apps/fftw/fftw-2.1.5/include -I../include  -c fft_stick.c
> > fft_stick.c(19): catastrophic error: could not open source file "fftw.h"
> >   #    include <fftw.h>
> >
> > It is not able to find fftw.h from /opt/apps/fftw/fftw-2.1.5/include
> 
> of course it is not: you are telling the compiler to look into
> ../opt/apps/fftw/fftw-2.1.5/include, not /opt/apps/fftw/fftw-2.1.5/include
> 
> > Also in the clib directory of espresso-3.1, fftw.h exists.
> 
> it is the include file for the internal version of fftw and shouldn't be
> loaded when using an external fftw library
> 
> > We also have fftw version 3.1.1. It's include file is fftw2.h.
> 
> you do not need it with v.3 .
> 
> P.
> 



From proffess at yandex.ru  Thu May 18 19:59:18 2006
From: proffess at yandex.ru (Sergey Lisenkov)
Date: Thu, 18 May 2006 21:59:18 +0400 (MSD)
Subject: [Pw_forum] compilation error
In-Reply-To: <Pine.LNX.4.44.0605181226310.31601-100000@linux2.ph.utexas.edu>
References: <Pine.LNX.4.44.0605181226310.31601-100000@linux2.ph.utexas.edu>
Message-ID: <446CB5F6.000007.02738@mfront7.yandex.ru>

Hi, Sahu

>-I../PH -I../iotk/src -I../CPV -c fftdrv.f90
>fortcom: Severe: **Internal compiler error: segmentation violation signal 
>raised** Please report this error along with the circumstances in which it 
>occurred in a Software Problem Report.  Note: File and line given may not 
>be explicit cause of this error.
> 

 I found the same error using ifort9.1 on our altix box. The compiler has some troubles with compiling esspresso (for example,  addusden.f90 file, it took 1 hour and I stopped it). so, I back to ifort9.0 and it works.

 S.


From giannozz at nest.sns.it  Thu May 18 20:53:17 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Thu, 18 May 2006 20:53:17 +0200
Subject: [Pw_forum] compilation error
In-Reply-To: <446CB5F6.000007.02738@mfront7.yandex.ru>
References: <Pine.LNX.4.44.0605181226310.31601-100000@linux2.ph.utexas.edu> <446CB5F6.000007.02738@mfront7.yandex.ru>
Message-ID: <200605182053.17175.giannozz@nest.sns.it>

On Thursday 18 May 2006 19:59, Sergey Lisenkov wrote:
> Hi, Sahu
>
> >-I../PH -I../iotk/src -I../CPV -c fftdrv.f90
> >fortcom: Severe: **Internal compiler error: segmentation violation signal
> >raised** Please report this error along with the circumstances in which it
> >occurred in a Software Problem Report.  Note: File and line given may not
> >be explicit cause of this error.
>
>  I found the same error using ifort9.1 on our altix box. The compiler has
> some troubles with compiling espresso (for example,  addusden.f90 file,
> it took 1 hour and I stopped it). so, I back to ifort9.0 and it works.

good to know: I'll update the infinite list of problems with intel
compilers ... Paolo

-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From wenpengan at gmail.com  Fri May 19 04:32:25 2006
From: wenpengan at gmail.com (an peng)
Date: Fri, 19 May 2006 10:32:25 +0800
Subject: [Pw_forum] how ti obtain the 2-D charge density ?
Message-ID: <95f06eac0605181932x7cc45606u1358b6fe246b1db9@mail.gmail.com>

Hi PW-SCF users:

   I am interesting to obtain the 2-D chanrge density,and I have the
charge. Would  someone tell me how to obtain the chdens.out or give me
format about  the input.

   Thanks
-------------- next part --------------
An HTML attachment was scrubbed...
URL: /pipermail/attachments/20060519/d845216a/attachment.htm 

From dkorotin at gmail.com  Fri May 19 08:01:20 2006
From: dkorotin at gmail.com (Dmitry Korotin)
Date: Fri, 19 May 2006 12:01:20 +0600
Subject: [Pw_forum] only one l state?
In-Reply-To: <200605181551.48034.giannozz@nest.sns.it>
References: <2fd252650605180139g61617465gca740cea6165332a@mail.gmail.com>
	 <200605181238.47958.giannozz@nest.sns.it>
	 <2fd252650605180408l63a132a4o99be66a76c7d8a25@mail.gmail.com>
	 <200605181551.48034.giannozz@nest.sns.it>
Message-ID: <2fd252650605182301r6422b762haf292b511c7083c1@mail.gmail.com>

> don't confuse the reference configuration with the states you use
> to generate the PP. You can start from the ground state (&input
> namelist), use 6s + an unbound state to generate the PP for l=0
> (&inputp namelist), unscreen with the ground state configuration
> (&test namelist)

Well, now I have following input file:
&input
	title = 'Pu',
	zed = 94.0,
	rel = 2,
	iswitch = 3,
	xmin=-7
	beta=0.1
	dx=0.01
	rmax=250
	dft='PBE'
	prefix = 'Pu'
	config = '[Rn] 5f6 6d0 7s2'
	nld = 4
	rlderiv = 2.5
	eminld = -3.0
	emaxld = 2.0
	deld = 0.01
/
&test
	nconf = 1
	configts(1) = '5f6 6d0 7s2'
/

&inputp
	pseudotype = 2
	file_pseudopw = 'Pu.rel.upf'
	lloc = 3
	rcloc = 1.6
	tm = .true.
	nlcc = .true.
/
10
6S  1  0  2.00  0.00  1.30  1.30  0.0
6S  1  0  0.00 -3.40  1.30  1.30  0.0
6P  2  1  2.00  0.00  1.60  1.60  0.5
6P  2  1  4.00  0.00  1.60  1.60  1.5
6D  3  2  0.00 -0.09  2.30  2.30  1.5
6D  3  2  0.00 -0.09  2.30  2.30  2.5
5F  4  3  6.00  0.00  1.05  1.05  2.5
5F  4  3  0.00  0.00  1.05  1.05  3.5
5F  4  3  0.00 -0.11  1.15  1.15  2.5
5F  4  3  0.00 -0.11  1.15  1.15  3.5

and it gives me some pseudo. But when I'm trying to test it following
error appears:
 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
     from ascheqps : error #         3
     too many attempts
 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

     stopping ...
2

What is the reason?

Input file for test calc:
&input
	title = 'Pu',
	zed = 94.0,
	rel = 2,
	iswitch = 2,
	xmin=-7
	beta=0.1
	dx=0.01
	rmax=250
	dft='PBE'
	prefix = 'Pu'
	config = '[Rn] 5f6 6d0 7s2'
/
&test
	nconf = 1
	file_pseudo = 'Pu.rel.upf'
	configts(1) = '5f6 6d0 7s2'
/


Thank you in advance,
Dmitry Korotin.


From giannozz at nest.sns.it  Fri May 19 09:05:59 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Fri, 19 May 2006 09:05:59 +0200
Subject: [Pw_forum] only one l state?
In-Reply-To: <2fd252650605182301r6422b762haf292b511c7083c1@mail.gmail.com>
References: <2fd252650605180139g61617465gca740cea6165332a@mail.gmail.com> <200605181551.48034.giannozz@nest.sns.it> <2fd252650605182301r6422b762haf292b511c7083c1@mail.gmail.com>
Message-ID: <200605190905.59829.giannozz@nest.sns.it>

On Friday 19 May 2006 08:01, Dmitry Korotin wrote:

> Well, now I have following input file:
> &input
> ...
> /
> &test
> ...
> /
> &inputp
> ...
> /

after v.3 the namelist test is the last. 

> and it gives me some pseudo. But when I'm trying to test it following
> error appears:

>      from ascheqps : error #         3
>      too many attempts

it means that your PP is bad: it has ghosts

P.
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From giannozz at nest.sns.it  Fri May 19 09:07:04 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Fri, 19 May 2006 09:07:04 +0200
Subject: [Pw_forum] how ti obtain the 2-D charge density ?
In-Reply-To: <95f06eac0605181932x7cc45606u1358b6fe246b1db9@mail.gmail.com>
References: <95f06eac0605181932x7cc45606u1358b6fe246b1db9@mail.gmail.com>
Message-ID: <200605190907.04065.giannozz@nest.sns.it>

On Friday 19 May 2006 04:32, an peng wrote:

>    I am interesting to obtain the 2-D chanrge density,and I have the
> charge. Would  someone tell me how to obtain the chdens.out or 
> give me format about  the input.

this is done in example 5
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From roma at srmp.saclay.cea.fr  Fri May 19 10:46:59 2006
From: roma at srmp.saclay.cea.fr (Guido Roma)
Date: Fri, 19 May 2006 10:46:59 +0200
Subject: [Pw_forum] constraints
In-Reply-To: <Pine.LNX.4.44.0605161220390.11882-100000@puccini.che.pitt.edu>
References: <Pine.LNX.4.44.0605161220390.11882-100000@puccini.che.pitt.edu>
Message-ID: <1148028419.7733.46.camel@srmp19.saclay.cea.fr>

Dear All,

I have been using contrained relaxations with older version of PWSCF,
I've seen that in version 3.1 many things are changed in that respect.
The module coping with constraints seems to allow only certain types of
constraints (distances, angles), not center of mass type constraints for
examples, and is not any more so easy to write one's own constrain.f90
subroutine. But, further more, I don't understand if the allowed
constraints as specified in the CONSTRAINTS input card are working or
not. Looking at the code I have the impression that they should work if
relaxing by damped dynamics (ion_dynamics='damp'), but not with bfgs.
Can someone confirm this or tell me if and where I'm wrong? 
Any don't see any hint in the user-guide.

Guido
-- 
Guido Roma <roma${at}cea.fr> -- CEA-Saclay - DEN/DMN/SRMP Bat.520/130
Phone: [+33]-1-69085738 -- Fax: ...6867 -- Mobile: [+33]-6-20069085




From giannozz at nest.sns.it  Fri May 19 11:52:35 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Fri, 19 May 2006 11:52:35 +0200
Subject: [Pw_forum] constraints
In-Reply-To: <1148028419.7733.46.camel@srmp19.saclay.cea.fr>
References: <Pine.LNX.4.44.0605161220390.11882-100000@puccini.che.pitt.edu> <1148028419.7733.46.camel@srmp19.saclay.cea.fr>
Message-ID: <200605191152.35071.giannozz@nest.sns.it>

On Friday 19 May 2006 10:46, Guido Roma wrote:

> I have been using contrained relaxations with older version of PWSCF,
> I've seen that in version 3.1 many things are changed in that respect.
> The module coping with constraints seems to allow only certain types of
> constraints (distances, angles), not center of mass type constraints for
> examples, and is not any more so easy to write one's own constrain.f90
> subroutine. But, further more, I don't understand if the allowed
> constraints as specified in the CONSTRAINTS input card are working or
> not. Looking at the code I have the impression that they should work if
> relaxing by damped dynamics (ion_dynamics='damp'), but not with bfgs.
> Can someone confirm this or tell me if and where I'm wrong?

someone can (Carlo S.)  but he is absent until the end of the month.
Meanwhile: what is written in INPUT_PW should correspond to what 
happens in reality. If it doesn't, please complain!

Paolo
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From wenpengan at gmail.com  Fri May 19 15:01:06 2006
From: wenpengan at gmail.com (an peng)
Date: Fri, 19 May 2006 21:01:06 +0800
Subject: [Pw_forum] how to obtain the 2-D charge density ?
Message-ID: <95f06eac0605190601l255f4049l1555a565c6ca738@mail.gmail.com>

yes, I know I should  read example05, but  I cannot plot the 2-D charge
density with pp_rho.out.



cat > naalh4.chdens.in << EOF
 &input
    nfile = 1,
    filepp(1) = 'naalh4charge',
    weight(1) =1,
    iflag = 2,
    plot_out = 1,
    output_format =1,
    fileout = naalh4.rho.dat',
    e1(1) =1.0, e1(2)=1.0, e1(3) = 0.0,
    e2(1) =0.0, e2(2)=0.0, e2(3) = 1.0,
        nx=56, ny=40

  could anyone tell me what is wrong with the shell.

 thanks!
-------------- next part --------------
An HTML attachment was scrubbed...
URL: /pipermail/attachments/20060519/7ba7c65f/attachment.htm 

From giannozz at nest.sns.it  Fri May 19 18:17:08 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Fri, 19 May 2006 18:17:08 +0200
Subject: [Pw_forum] how to obtain the 2-D charge density ?
In-Reply-To: <95f06eac0605190601l255f4049l1555a565c6ca738@mail.gmail.com>
References: <95f06eac0605190601l255f4049l1555a565c6ca738@mail.gmail.com>
Message-ID: <200605191817.08821.giannozz@nest.sns.it>

On Friday 19 May 2006 15:01, an peng wrote:

> yes, I know I should  read example05

and the input documentation in Doc/

> cat > naalh4.chdens.in << EOF
>  &input
> [...]
>   could anyone tell me what is wrong with the shell.

for sure the first line (the namelist is "&inputpp") and the last
(the namelist it is not terminated)
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From mpierce at cs.indiana.edu  Fri May 19 19:23:22 2006
From: mpierce at cs.indiana.edu (Marlon Pierce)
Date: Fri, 19 May 2006 13:23:22 -0400
Subject: [Pw_forum] Problems installing Espresso 2.1 and 3.0
Message-ID: <446DFF0A.8010909@cs.indiana.edu>

I'm trying to install the espresso package on the machine described 
here: http://www.tacc.utexas.edu/services/userguides/lonestar2/.   Both 
2.1.2 and 3.0 fail at the "make all" stage.

Some notes:
1. I manually change my $ARCH to "linux64". 
2. ./configure does not complain.
3. The "make all" error is below.

-------------------
mpif90 -O2 -assume byterecl -nomodule -fpp -D__LINUX64 -D__INTEL 
-D__FFTW -D__USE_INTERNAL_FFTW -D__MPI -D__PARA -I../include  -I. 
-I../Modules -I../PW -I../PH -I../iotk/src -c path_reparametrisation.f90
mpif90 -O2 -assume byterecl -nomodule -fpp -D__LINUX64 -D__INTEL 
-D__FFTW -D__USE_INTERNAL_FFTW -D__MPI -D__PARA -I../include  -I. 
-I../Modules -I../PW -I../PH -I../iotk/src -c path_base.f90
fortcom: Severe: **Internal compiler error: segmentation violation 
signal raised** Please report this error along with the circumstances in 
which it occurred in a Software Problem Report.  Note: File and line 
given may not be explicit cause of this error.

compilation aborted for path_base.f90 (code 3)
make[1]: *** [path_base.o] Error 3
make[1]: Leaving directory `/home/teragrid/tg457032/espresso-3.0/Modules'
make: *** [mods] Error 2
-------------------


I'm happy to send additional information (config logs, entire make 
command output, etc).

Thanks--

Marlon Pierce
Community Grids Laboratory
Indiana University


-------------- next part --------------
An HTML attachment was scrubbed...
URL: /pipermail/attachments/20060519/18a13bc0/attachment.htm 

From akohlmey at cmm.upenn.edu  Fri May 19 20:06:01 2006
From: akohlmey at cmm.upenn.edu (Axel Kohlmeyer)
Date: Fri, 19 May 2006 14:06:01 -0400
Subject: [Pw_forum] Problems installing Espresso 2.1 and 3.0
In-Reply-To: <446DFF0A.8010909@cs.indiana.edu>
References: <446DFF0A.8010909@cs.indiana.edu>
Message-ID: <7b6913e90605191106yc53127dved5c0e9eaf78642b@mail.gmail.com>

On 5/19/06, Marlon Pierce <mpierce at cs.indiana.edu> wrote:
>

marlon,

> I'm trying to install the espresso package on the machine described here:
> http://www.tacc.utexas.edu/services/userguides/lonestar2/.
>  Both 2.1.2 and 3.0 fail at the "make all" stage.

please note, that version 3.0 has been superseded by
version 3.1 recently.


>  Some notes:
>  1. I manually change my $ARCH to "linux64".
>  2. ./configure does not complain.
>  3. The "make all" error is below.
>
>  -------------------
>  mpif90 -O2 -assume byterecl -nomodule -fpp -D__LINUX64 -D__INTEL -D__FFTW
> -D__USE_INTERNAL_FFTW -D__MPI -D__PARA -I../include  -I. -I../Modules
> -I../PW -I../PH -I../iotk/src -c path_reparametrisation.f90
>  mpif90 -O2 -assume byterecl -nomodule -fpp -D__LINUX64 -D__INTEL -D__FFTW
> -D__USE_INTERNAL_FFTW -D__MPI -D__PARA -I../include  -I. -I../Modules
> -I../PW -I../PH -I../iotk/src -c path_base.f90
>  fortcom: Severe: **Internal compiler error: segmentation violation signal
> raised** Please report this error along with the circumstances in which it
> occurred in a Software Problem Report.  Note: File and line given may not be
> explicit cause of this error.

and 'internal compiler error' is _always_ a bug in the compiler. so you should
contact the compiler vendor for an update or file a bug report there.
sometimes, you can work around it, by reducing the optimization level for
this one object.


>  compilation aborted for path_base.f90 (code 3)
>  make[1]: *** [path_base.o] Error 3
>  make[1]: Leaving directory
> `/home/teragrid/tg457032/espresso-3.0/Modules'
>  make: *** [mods] Error 2
>  -------------------

regards,
     axel.

>  I'm happy to send additional information (config logs, entire make command
> output, etc).


>
>  Thanks--
>
>  Marlon Pierce
>  Community Grids Laboratory
>  Indiana University
>
>
>
>
>


-- 
=======================================================================
Axel Kohlmeyer   akohlmey at cmm.chem.upenn.edu   http://www.cmm.upenn.edu
  Center for Molecular Modeling   --   University of Pennsylvania
Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323
tel: 1-215-898-1582,  fax: 1-215-573-6233,  office-tel: 1-215-898-5425
=======================================================================
If you make something idiot-proof, the universe creates a better idiot.


From lanhaiping at gmail.com  Mon May 22 11:07:53 2006
From: lanhaiping at gmail.com (lan haiping)
Date: Mon, 22 May 2006 17:07:53 +0800
Subject: [Pw_forum] How to Fix certain atoms in relax calculation?
Message-ID: <c92002fa0605220207x73c30ef1m209a8e148f5ef620@mail.gmail.com>

Hi ,

   i wanna do some  relaxation calculations with some atoms fixed, but i
donot know how to in PWSCF.
I have tried to read the Doc/INPUT_PW, and was confused that "if_pos(i)"
with 1 and CONSTRAINT CARDS might realize my goal.   Would you please give
me some hints about these two differences ?

 Regards,

Hai-Ping
-------------- next part --------------
An HTML attachment was scrubbed...
URL: /pipermail/attachments/20060522/45e196e5/attachment.htm 

From lanhaiping at gmail.com  Mon May 22 11:18:58 2006
From: lanhaiping at gmail.com (lan haiping)
Date: Mon, 22 May 2006 17:18:58 +0800
Subject: [Pw_forum] INPUT file for average.x
Message-ID: <c92002fa0605220218y46ad4f98xdfef08c77e1ceaa7@mail.gmail.com>

Dear developers,

        someone else mentioned  that macroscopic average technique has been
realized in average.x.
I  tried to read  the code average.f90 , but i still donot  know the INPUT
file's format. Would you please explain more about the input variables ?
thank you!

      Regards,

       Hai-Ping
-------------- next part --------------
An HTML attachment was scrubbed...
URL: /pipermail/attachments/20060522/0b321a72/attachment.htm 

From Giovanni.Cantele at na.infn.it  Mon May 22 11:43:28 2006
From: Giovanni.Cantele at na.infn.it (Giovanni Cantele)
Date: Mon, 22 May 2006 11:43:28 +0200
Subject: [Pw_forum] INPUT file for average.x
In-Reply-To: <c92002fa0605220218y46ad4f98xdfef08c77e1ceaa7@mail.gmail.com>
References: <c92002fa0605220218y46ad4f98xdfef08c77e1ceaa7@mail.gmail.com>
Message-ID: <447187C0.1070100@na.infn.it>

lan haiping wrote:

> Dear developers,
>  
>         someone else mentioned  that macroscopic average technique has 
> been realized in average.x.
> I  tried to read  the code average.f90 , but i still donot  know the 
> INPUT file's format. Would you please explain more about the input 
> variables ? thank you!
>  
>       Regards,
>  
>        Hai-Ping

A typical input should look like this:

1
filename
1.0
3000
3
2.59

The explanation is in the header of the file PP/average.f90:

  !      nfile        ! the number of 3D-FFT files
  ! for each file:
  !      filename     ! the name of the 3D-FFT file
  !      weight       ! the weight of the quantity in this file
  !      .
  !      .
  ! end
  !      npt          ! the number of points of the thick mesh
  !      idir         ! 1,2 or 3. It is the fixed index which defines
  !                   ! the planes of the planar average
  !      awin         ! the size of the window for macroscopic averages.
  !

With the example above you'll get a 3-column output file for the 
quantity to average
V(u):
- column 1: coordinate u, defined over a mesh of npt=3000 points along 
the direction
idir=3
- column 2: averaged quantity V
- column 3: macroscopically averaged quantity V', over a windows with size
awin=2.50 au

If you want to average a weighted sum of different quantities V1,...,VN
defined over the same FFT mesh, just set nfile to N and specify N file names
with the relative weight.

Regards,
Giovanni

-- 



Dr. Giovanni Cantele
Coherentia CNR-INFM and Dipartimento di Scienze Fisiche
Universita' di Napoli "Federico II"
Complesso Universitario di Monte S. Angelo - Ed. G
Via Cintia, I-80126, Napoli, Italy
Phone: +39 081 676910
Fax:   +39 081 676346
E-mail: Giovanni.Cantele at na.infn.it
Web: http://people.na.infn.it/~cantele



From ezadshojaee at hotmail.com  Mon May 22 11:48:04 2006
From: ezadshojaee at hotmail.com (Ezad Shojaee)
Date: Mon, 22 May 2006 09:48:04 +0000
Subject: [Pw_forum] about band gap
Message-ID: <BAY19-F20FC1C50A65C4606BD3738AA9A0@phx.gbl>

Hi everyone
i am trying to obtain the band gap of TiO2(Anatase) and i did not have 
better than about 9ev ( instead of  3.2 ! )  i have tried many paths in the 
BZ including high-symmetry points & also i tried { automatic } k_points  but 
it did not help.
any suggestion will be appreciated
this in my input file for band structure

&control
    calculation='nscf'
    prefix='TiO2',
    pseudo_dir = '/root/pseudo/',
    outdir='/root/tmp/'
/
&system
    ibrav= 7,  celldm(1) =7.155,   celldm(3) =2.5136,
    nat=  6, ntyp= 2,
    ecutwfc =50.0,
    nbnd = 30,
/
&electrons
/
ATOMIC_SPECIES
Ti  47.867     Ti.vdb.UPF
O   15.9994    O.vdb.UPF
ATOMIC_POSITIONS
Ti  0.12500000 -0.12500000 -0.25000000
Ti -0.12500000  0.12500000  0.25000000
O   0.08159975  0.33159975 -0.16319950
O  -0.08159975 -0.33159975  0.16319950
O   0.33159975  0.08159975  0.33680050
O  -0.33159975 -0.08159975 -0.33680050
K_POINTS { automatic }
9  9  9   0 0 0

_________________________________________________________________
FREE pop-up blocking with the new MSN Toolbar - get it now! 
http://toolbar.msn.click-url.com/go/onm00200415ave/direct/01/



From Giovanni.Cantele at na.infn.it  Mon May 22 12:27:04 2006
From: Giovanni.Cantele at na.infn.it (Giovanni Cantele)
Date: Mon, 22 May 2006 12:27:04 +0200
Subject: [Pw_forum] about band gap
In-Reply-To: <BAY19-F20FC1C50A65C4606BD3738AA9A0@phx.gbl>
References: <BAY19-F20FC1C50A65C4606BD3738AA9A0@phx.gbl>
Message-ID: <447191F8.1000009@na.infn.it>

Ezad Shojaee wrote:

> Hi everyone
> i am trying to obtain the band gap of TiO2(Anatase) and i did not have 
> better than about 9ev ( instead of  3.2 ! )  i have tried many paths 
> in the BZ including high-symmetry points & also i tried { automatic } 
> k_points  but it did not help.
> any suggestion will be appreciated
> this in my input file for band structure
>
> &control
>    calculation='nscf'
>    prefix='TiO2',
>    pseudo_dir = '/root/pseudo/',
>    outdir='/root/tmp/'
> /
> &system
>    ibrav= 7,  celldm(1) =7.155,   celldm(3) =2.5136,
>    nat=  6, ntyp= 2,
>    ecutwfc =50.0,
>    nbnd = 30,
> /
> &electrons
> /
> ATOMIC_SPECIES
> Ti  47.867     Ti.vdb.UPF
> O   15.9994    O.vdb.UPF
> ATOMIC_POSITIONS
> Ti  0.12500000 -0.12500000 -0.25000000
> Ti -0.12500000  0.12500000  0.25000000
> O   0.08159975  0.33159975 -0.16319950
> O  -0.08159975 -0.33159975  0.16319950
> O   0.33159975  0.08159975  0.33680050
> O  -0.33159975 -0.08159975 -0.33680050
> K_POINTS { automatic }
> 9  9  9   0 0 0
>
If you try to visualize your structure, it does not look really as an
anatase TiO2 crystal.
I think that you have two problems with you input file.
The anatase structure should be as follows:

Primitive vectors
a(1) =   3.78500000  0.00000000  0.00000000
a(2) =   0.00000000  3.78500000  0.00000000
a(3) =   1.89250000  1.89250000  4.75700000


Basis Vectors:
Atom    Lattice Coordinates                Cartesian Coordinates

Ti  0.12500000 -0.12500000 -0.25000000     0.00000000 -0.94625000 -1.18925000
Ti -0.12500000  0.12500000  0.25000000     0.00000000  0.94625000  1.18925000
O   0.08159975  0.33159975 -0.16319950     0.00000000  0.94625000 -0.77634000
O  -0.08159975 -0.33159975  0.16319950     0.00000000 -0.94625000  0.77634000
O   0.33159975  0.08159975  0.33680050     1.89250000  0.94625000  1.60216000
O  -0.33159975 -0.08159975 -0.33680050    -1.89250000 -0.94625000 -1.60216000


- 1st problem, from Doc/INPUT_PW:
ATOMIC_POSITIONS { alat | bohr | crystal | angstrom }

   alat    : atomic positions are in units of alat (default)
   bohr    : atomic positions are in a.u.
   crystal : atomic positions are in crystal coordinates (see below)
   angstrom: atomic positions are in A

It seems that you specified your positions in lattice (crystal) coordinates
without specifying {crystal} after ATOMIC_POSITIONS
The code will understand that those positions are in units of alat
(celldm(1))

- 2d problem, from Doc/INPUT_PW:

 body centered tetragonal (i)
  ================================
     a1 = (a/2)(1,-1,c/a),  a2 = (a/2)(1,1,c/a),  a3 = (a/2)(-1,-1,c/a).

This is the choice that PWSCF makes for the lattice vectors (you can 
check it in
your output,  they are explicitly written). It could be that if you 
specified the crystal
coordinates referred to a different choice of the crystal axes, the 
atoms are
not placed where they should.
So, check your axes and try to change the crystal coordinates 
accordingly or,
more simply, try to use Cartesian coordinates (listed above, in 
Angstrom). I tried
this last choice and it looks much better. If everything works well, you can
easily revert back to crystal coordinates reading them from the output


Hope this helps,

Giovanni


-- 



Dr. Giovanni Cantele
Coherentia CNR-INFM and Dipartimento di Scienze Fisiche
Universita' di Napoli "Federico II"
Complesso Universitario di Monte S. Angelo - Ed. G
Via Cintia, I-80126, Napoli, Italy
Phone: +39 081 676910
Fax:   +39 081 676346
E-mail: Giovanni.Cantele at na.infn.it
Web: http://people.na.infn.it/~cantele



From ezadshojaee at hotmail.com  Mon May 22 12:50:35 2006
From: ezadshojaee at hotmail.com (Ezad Shojaee)
Date: Mon, 22 May 2006 10:50:35 +0000
Subject: [Pw_forum] about band gap
Message-ID: <BAY19-F997ED0C0CF52478AB77B0AA9A0@phx.gbl>

thank you Giovanni
don't you think that if i want to use the cartesian coordinates as you said 
, i have to convert them to BOHR? because the celldm(1) must be speciefied 
in Bohr  & so on.
thanks

_________________________________________________________________
FREE pop-up blocking with the new MSN Toolbar - get it now! 
http://toolbar.msn.click-url.com/go/onm00200415ave/direct/01/



From lanhaiping at gmail.com  Mon May 22 13:01:37 2006
From: lanhaiping at gmail.com (lan haiping)
Date: Mon, 22 May 2006 19:01:37 +0800
Subject: [Pw_forum] INPUT file for average.x
In-Reply-To: <447187C0.1070100@na.infn.it>
References: <c92002fa0605220218y46ad4f98xdfef08c77e1ceaa7@mail.gmail.com>
	 <447187C0.1070100@na.infn.it>
Message-ID: <c92002fa0605220401t2c9d0305pf9e45a5dd949ae3@mail.gmail.com>

Thank you, Giovanni.

I would like to ask another naive question.
what is the 3-D FFT file? is it saved in tmp  directory ,or else ?

Does the setting of npt have relation to  the thickness of slab?

 Regards,

Hai-Ping

On 5/22/06, Giovanni Cantele <Giovanni.Cantele at na.infn.it> wrote:
>
> lan haiping wrote:
>
> > Dear developers,
> >
> >         someone else mentioned  that macroscopic average technique has
> > been realized in average.x.
> > I  tried to read  the code average.f90 , but i still donot  know the
> > INPUT file's format. Would you please explain more about the input
> > variables ? thank you!
> >
> >       Regards,
> >
> >        Hai-Ping
>
> A typical input should look like this:
>
> 1
> filename
> 1.0
> 3000
> 3
> 2.59
>
> The explanation is in the header of the file PP/average.f90:
>
> !      nfile        ! the number of 3D-FFT files
> ! for each file:
> !      filename     ! the name of the 3D-FFT file
> !      weight       ! the weight of the quantity in this file
> !      .
> !      .
> ! end
> !      npt          ! the number of points of the thick mesh
> !      idir         ! 1,2 or 3. It is the fixed index which defines
> !                   ! the planes of the planar average
> !      awin         ! the size of the window for macroscopic averages.
> !
>
> With the example above you'll get a 3-column output file for the
> quantity to average
> V(u):
> - column 1: coordinate u, defined over a mesh of npt=3000 points along
> the direction
> idir=3
> - column 2: averaged quantity V
> - column 3: macroscopically averaged quantity V', over a windows with size
> awin=2.50 au
>
> If you want to average a weighted sum of different quantities V1,...,VN
> defined over the same FFT mesh, just set nfile to N and specify N file
> names
> with the relative weight.
>
> Regards,
> Giovanni
>
> --
>
>
>
> Dr. Giovanni Cantele
> Coherentia CNR-INFM and Dipartimento di Scienze Fisiche
> Universita' di Napoli "Federico II"
> Complesso Universitario di Monte S. Angelo - Ed. G
> Via Cintia, I-80126, Napoli, Italy
> Phone: +39 081 676910
> Fax:   +39 081 676346
> E-mail: Giovanni.Cantele at na.infn.it
> Web: http://people.na.infn.it/~cantele
>
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: /pipermail/attachments/20060522/091bc900/attachment.htm 

From Giovanni.Cantele at na.infn.it  Mon May 22 13:01:47 2006
From: Giovanni.Cantele at na.infn.it (Giovanni Cantele)
Date: Mon, 22 May 2006 13:01:47 +0200
Subject: [Pw_forum] about band gap
In-Reply-To: <BAY19-F997ED0C0CF52478AB77B0AA9A0@phx.gbl>
References: <BAY19-F997ED0C0CF52478AB77B0AA9A0@phx.gbl>
Message-ID: <44719A1B.4000605@na.infn.it>

Ezad Shojaee wrote:

> thank you Giovanni
> don't you think that if i want to use the cartesian coordinates as you 
> said , i have to convert them to BOHR? because the celldm(1) must be 
> speciefied in Bohr  & so on.
> thanks

If you use
ATOMIC_POSITIONS {Angstrom}

then Cartesian coordinates are read in Angstrom. This is independent of 
units of celldm(1), that instead
is always in bohr.

Giovanni


-- 



Dr. Giovanni Cantele
Coherentia CNR-INFM and Dipartimento di Scienze Fisiche
Universita' di Napoli "Federico II"
Complesso Universitario di Monte S. Angelo - Ed. G
Via Cintia, I-80126, Napoli, Italy
Phone: +39 081 676910
Fax:   +39 081 676346
E-mail: Giovanni.Cantele at na.infn.it
Web: http://people.na.infn.it/~cantele



From Giovanni.Cantele at na.infn.it  Mon May 22 13:22:34 2006
From: Giovanni.Cantele at na.infn.it (Giovanni Cantele)
Date: Mon, 22 May 2006 13:22:34 +0200
Subject: [Pw_forum] INPUT file for average.x
In-Reply-To: <c92002fa0605220401t2c9d0305pf9e45a5dd949ae3@mail.gmail.com>
References: <c92002fa0605220218y46ad4f98xdfef08c77e1ceaa7@mail.gmail.com>	 <447187C0.1070100@na.infn.it> <c92002fa0605220401t2c9d0305pf9e45a5dd949ae3@mail.gmail.com>
Message-ID: <44719EFA.4040804@na.infn.it>

lan haiping wrote:

> Thank you, Giovanni.
>  
> I would like to ask another naive question.
> what is the 3-D FFT file? is it saved in tmp  directory ,or else ?

You have to generate this file according to what you want to average
and/or plot. The executable is pp.x, the input variables are described
in Doc/INPUT_PP

For example if you need the charge density you run pp.x with this input:

&inputPP
    prefix='...'
    outdir='...'
    filplot='filename'
    plot_num=0
/

This will store the charge density calculated on the 3D FFT grid in the
file filename (the one you'll specify in the input of average.x)
prefix and outdir should be the same as in the scf/relax/... run you did
with pw.x

>  
> Does the setting of npt have relation to  the thickness of slab?

No, it is the mesh of points for calculating the average. The spacing 
between
the points will be related with the dimension L of the unit cell along 
the idir
direction (L/npt).

Giovanni

-- 



Dr. Giovanni Cantele
Coherentia CNR-INFM and Dipartimento di Scienze Fisiche
Universita' di Napoli "Federico II"
Complesso Universitario di Monte S. Angelo - Ed. G
Via Cintia, I-80126, Napoli, Italy
Phone: +39 081 676910
Fax:   +39 081 676346
E-mail: Giovanni.Cantele at na.infn.it
Web: http://people.na.infn.it/~cantele



From dkorotin at gmail.com  Mon May 22 13:38:35 2006
From: dkorotin at gmail.com (Dmitry Korotin)
Date: Mon, 22 May 2006 17:38:35 +0600
Subject: [Pw_forum] PAW pseudos
Message-ID: <2fd252650605220438r7f19be5nf73f216a5ffbff51@mail.gmail.com>

Dear all,
Is it possible to convert PAW pseudopotential (from the VASP library)
to UPF format (or use it with ESPRESSO in another way)?

Thank you in advance,
Dmitry Korotin.


From Andrea.Marini at roma2.infn.it  Mon May 22 14:04:08 2006
From: Andrea.Marini at roma2.infn.it (Andrea Marini)
Date: Mon, 22 May 2006 14:04:08 +0200 (CEST)
Subject: [Pw_forum] Symmetrized el-ph coupling
Message-ID: <Pine.LNX.4.64.0605221350010.1824@anomaly.roma2.infn.it>

Dear forum users,

  I am trying to use PW to calculate the electron-phonon matrix elements 
(EPME) to perform external  self-energy based calculations.

I have studied the elphon.f90 routine and I have understood the way PH 
builds the EPME and how it sums them in the lambda/gamma factors.

My problem is the call to the symdynph_gq routine to symmetrize the 
el_ph_sum array (that contains the sum of the square EPME).

As I need the EPME and not their square I do not understand how to perform 
the symmetrization on the el_ph_mat array (and not on the summed array 
el_ph_sum) in such a way to get exactly the same results for the physical 
lambda/gamma factors.

Thanks

Andrea Marini
--------------------------------------------------------------------------
                                Andrea MARINI

        Physics Department, University of Rome "Tor Vergata" (Italy)
               - phone: +39-0672594894 - fax: +39-062023507 -

-- andrea.marini at roma2.infn.it - http://www.fisica.uniroma2.it/~marini/ --


From eyvaz_isaev at yahoo.com  Mon May 22 16:27:44 2006
From: eyvaz_isaev at yahoo.com (Eyvaz Isaev)
Date: Mon, 22 May 2006 07:27:44 -0700 (PDT)
Subject: [Pw_forum] PAW pseudos
In-Reply-To: <2fd252650605220438r7f19be5nf73f216a5ffbff51@mail.gmail.com>
Message-ID: <20060522142744.71202.qmail@web60313.mail.yahoo.com>

Hi Dmitry,

Did you ask George Kresse about this one? More
seriously, you are not allowed to do it, because it is
a commercial product.

Bests,
Eyvaz.

--- Dmitry Korotin <dkorotin at gmail.com> wrote:

> Dear all,
> Is it possible to convert PAW pseudopotential (from
> the VASP library)
> to UPF format (or use it with ESPRESSO in another
> way)?
> 
> Thank you in advance,
> Dmitry Korotin.
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
> 


__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 


From akohlmey at cmm.upenn.edu  Mon May 22 16:28:34 2006
From: akohlmey at cmm.upenn.edu (Axel Kohlmeyer)
Date: Mon, 22 May 2006 10:28:34 -0400
Subject: [Pw_forum] How to Fix certain atoms in relax calculation?
In-Reply-To: <c92002fa0605220207x73c30ef1m209a8e148f5ef620@mail.gmail.com>
References: <c92002fa0605220207x73c30ef1m209a8e148f5ef620@mail.gmail.com>
Message-ID: <7b6913e90605220728k732e312ia446e269500d3d06@mail.gmail.com>

On 5/22/06, lan haiping <lanhaiping at gmail.com> wrote:
>
> Hi ,
>
>    i wanna do some  relaxation calculations with some atoms fixed, but i
> donot know how to in PWSCF.
> I have tried to read the Doc/INPUT_PW, and was confused that "if_pos(i)"
> with 1 and CONSTRAINT CARDS might realize my goal.   Would you please give
> me some hints about these two differences ?

if you just want to fix positions in space, if_pos(i) is the method to use.
just after the coordinates you can put 3 numbers or x-, y-, and z-component.
in each md or relax step, the corresponding force component for that coordinate
is multiplied with the factor you give. i.e. an atom can freely move
with a  1 1 1
and will be frozen with 0 0 0. using, e.g.,  1 1 0 will only allow
movement in xy-plane
0 0 1 only in z-direction and so on.

CONSTRAINT_CARDS is for e.g. distance constraint and similar stuff.

regards,
     axel.
>  Regards,
>
> Hai-Ping
>
>


-- 
=======================================================================
Axel Kohlmeyer   akohlmey at cmm.chem.upenn.edu   http://www.cmm.upenn.edu
  Center for Molecular Modeling   --   University of Pennsylvania
Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323
tel: 1-215-898-1582,  fax: 1-215-573-6233,  office-tel: 1-215-898-5425
=======================================================================
If you make something idiot-proof, the universe creates a better idiot.


From akohlmey at cmm.upenn.edu  Mon May 22 16:32:46 2006
From: akohlmey at cmm.upenn.edu (Axel Kohlmeyer)
Date: Mon, 22 May 2006 10:32:46 -0400
Subject: [Pw_forum] PAW pseudos
In-Reply-To: <20060522142744.71202.qmail@web60313.mail.yahoo.com>
References: <2fd252650605220438r7f19be5nf73f216a5ffbff51@mail.gmail.com>
	 <20060522142744.71202.qmail@web60313.mail.yahoo.com>
Message-ID: <7b6913e90605220732g354f5a63lfa41a51f102ef55c@mail.gmail.com>

On 5/22/06, Eyvaz Isaev <eyvaz_isaev at yahoo.com> wrote:
> Hi Dmitry,
>
> Did you ask George Kresse about this one? More
> seriously, you are not allowed to do it, because it is
> a commercial product.

regardless of that. please see the vanderbilt uspp homepage:
http://www.physics.rutgers.edu/%7Edhv/uspp/#USER

Another well-known proprietary code that uses ultrasoft potentials is
the  VASP package. However, VASP uses ultrasoft potentials generated
using a somewhat different scheme from the one implemented here. Thus,
pseudopotentials taken from the library distributed as part of VASP
cannot be imported into the above code packages, and vice versa.

regards,
    axel.
> Bests,
> Eyvaz.
>
> --- Dmitry Korotin <dkorotin at gmail.com> wrote:
>
> > Dear all,
> > Is it possible to convert PAW pseudopotential (from
> > the VASP library)
> > to UPF format (or use it with ESPRESSO in another
> > way)?
> >
> > Thank you in advance,
> > Dmitry Korotin.
> > _______________________________________________
> > Pw_forum mailing list
> > Pw_forum at pwscf.org
> > http://www.democritos.it/mailman/listinfo/pw_forum
> >
>
>
> __________________________________________________
> Do You Yahoo!?
> Tired of spam?  Yahoo! Mail has the best spam protection around
> http://mail.yahoo.com
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
>
>


-- 
=======================================================================
Axel Kohlmeyer   akohlmey at cmm.chem.upenn.edu   http://www.cmm.upenn.edu
  Center for Molecular Modeling   --   University of Pennsylvania
Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323
tel: 1-215-898-1582,  fax: 1-215-573-6233,  office-tel: 1-215-898-5425
=======================================================================
If you make something idiot-proof, the universe creates a better idiot.


From eyvaz_isaev at yahoo.com  Mon May 22 17:33:33 2006
From: eyvaz_isaev at yahoo.com (Eyvaz Isaev)
Date: Mon, 22 May 2006 08:33:33 -0700 (PDT)
Subject: [Pw_forum] PAW pseudos
In-Reply-To: <7b6913e90605220732g354f5a63lfa41a51f102ef55c@mail.gmail.com>
Message-ID: <20060522153333.75489.qmail@web60323.mail.yahoo.com>

Hi Axel,

--- Axel Kohlmeyer <akohlmey at cmm.upenn.edu> wrote:

> On 5/22/06, Eyvaz Isaev <eyvaz_isaev at yahoo.com>
> wrote:
> > Hi Dmitry,
> >
> > Did you ask George Kresse about this one? More
> > seriously, you are not allowed to do it, because
> it is
> > a commercial product.
> 
> regardless of that. please see the vanderbilt uspp
> homepage:
> http://www.physics.rutgers.edu/%7Edhv/uspp/#USER
> 
> Another well-known proprietary code that uses
> ultrasoft potentials is
> the  VASP package. However, VASP uses ultrasoft
> potentials generated
> using a somewhat different scheme from the one
> implemented here. Thus,
> pseudopotentials taken from the library distributed
> as part of VASP
> cannot be imported into the above code packages, and
> vice versa.

That is exactly I mean. The question about Kresse was
just a jock. 

Bests,
Eyvaz.


> regards,
>     axel.
> > Bests,
> > Eyvaz.
> >
> > --- Dmitry Korotin <dkorotin at gmail.com> wrote:
> >
> > > Dear all,
> > > Is it possible to convert PAW pseudopotential
> (from
> > > the VASP library)
> > > to UPF format (or use it with ESPRESSO in
> another
> > > way)?
> > >
> > > Thank you in advance,
> > > Dmitry Korotin.
> > > _______________________________________________
> > > Pw_forum mailing list
> > > Pw_forum at pwscf.org
> > >
> http://www.democritos.it/mailman/listinfo/pw_forum
> > >
> >
> >
> > __________________________________________________
> > Do You Yahoo!?
> > Tired of spam?  Yahoo! Mail has the best spam
> protection around
> > http://mail.yahoo.com
> > _______________________________________________
> > Pw_forum mailing list
> > Pw_forum at pwscf.org
> > http://www.democritos.it/mailman/listinfo/pw_forum
> >
> >
> 
> 
> -- 
>
=======================================================================
> Axel Kohlmeyer   akohlmey at cmm.chem.upenn.edu  
> http://www.cmm.upenn.edu
>   Center for Molecular Modeling   --   University of
> Pennsylvania
> Department of Chemistry, 231 S.34th Street,
> Philadelphia, PA 19104-6323
> tel: 1-215-898-1582,  fax: 1-215-573-6233, 
> office-tel: 1-215-898-5425
>
=======================================================================
> If you make something idiot-proof, the universe
> creates a better idiot.
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
> 


__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 


From virginie.quequet at polytechnique.edu  Mon May 22 18:15:33 2006
From: virginie.quequet at polytechnique.edu (=?ISO-8859-1?Q?Virginie_Trinit=E9?=)
Date: Mon, 22 May 2006 18:15:33 +0200
Subject: [Pw_forum] question on random_matrix
Message-ID: <4471E3A5.2060500@polytechnique.edu>

Dear all

I have a question about the subroutine random_matrix in PH. here a 
simplify version of the code:

**************************
wdyn (:, :, :, : )= (0d0, 0d0)
do na = 1, nat
    wdyn (ipol, ipol, na, na) = random
    do ipol = 1, 3
    wdyn (ipol, jpol, na, na) = random
    wdyn (jpol, ipol, na, na) = CONJG(wdyn (ipol, jpol, na, na) )
    enddo
    do nb = na + 1, nat
       do isymq = 1, nsymq
          irot = irgq (isymq)
          ira = irt (irot, na)
           if ( (nb == ira)) then
               do jpol = 1, 3
                  wdyn (ipol, jpol, na, nb) = random
                  wdyn(jpol, ipol, nb, na) = CONJG(wdyn(ipol, jpol, na, 
nb))
               enddo
                goto 10
             endif
          enddo
10         continue
***************************

I didn't understand why if nb is never the transform of na by some 
symmetry then the dynamical matrix is zero?
Best regards

Virginie



From hsd22 at hermes.cam.ac.uk  Mon May 22 20:15:53 2006
From: hsd22 at hermes.cam.ac.uk (H.S.Domingos)
Date: Mon, 22 May 2006 19:15:53 +0100 (BST)
Subject: [Pw_forum] Quetion
In-Reply-To: <4471E3A5.2060500@polytechnique.edu>
References: <4471E3A5.2060500@polytechnique.edu>
Message-ID: <Pine.LNX.4.64.0605221912030.827@hermes-1.csi.cam.ac.uk>

Dear all,

at the risk of sounding very stupid I would like to ask if you could 
please tell me how the program defines total magnetization and absolute 
magnetization in colinear and non-colinear spin dependent calculations ?

I am confused, in particular, about the meaning of absolute magnetization.

Thank you very much in advance.

Best regards,

Helder

=======================================================================
|  Dr. Helder S. Domingos                                             |
|                                                                     |
| INESC Microsyst & Nanotechnol, Lisbon, P-1000 Portugal              |
| and                                                                 |
| R&D unit for  Molecular Chemical Physics                            |
| Chemistry Department, University of Coimbra                         |
=======================================================================




From akohlmey at cmm.upenn.edu  Tue May 23 04:14:39 2006
From: akohlmey at cmm.upenn.edu (Axel Kohlmeyer)
Date: Mon, 22 May 2006 22:14:39 -0400
Subject: [Pw_forum] Quetion
In-Reply-To: <Pine.LNX.4.64.0605221912030.827@hermes-1.csi.cam.ac.uk>
References: <4471E3A5.2060500@polytechnique.edu>
	 <Pine.LNX.4.64.0605221912030.827@hermes-1.csi.cam.ac.uk>
Message-ID: <7b6913e90605221914v6eb5446dibe72918436ecded4@mail.gmail.com>

On 5/22/06, H.S.Domingos <hsd22 at hermes.cam.ac.uk> wrote:
> Dear all,
>
> at the risk of sounding very stupid I would like to ask if you could
> please tell me how the program defines total magnetization and absolute
> magnetization in colinear and non-colinear spin dependent calculations ?
>
> I am confused, in particular, about the meaning of absolute magnetization.

helder,

total magnetization is the integral over the difference between alpha
and beta spin electron densities, whereas the absolute magnetization is
the corresponding integral over the _absolute_ of the difference.

consider two hydrogen atoms quite far apart: now in the singlet-like
configuration your total magnetization would be 0 but the absolute magnetization
would be  2, in the triplet-like configuration both would be 2.0.

in case of a hydrogen molecule, however, both would be 0.0.

best regards,
    axel.
>
> Thank you very much in advance.
>
> Best regards,
>
> Helder
>
> =======================================================================
> |  Dr. Helder S. Domingos                                             |
> |                                                                     |
> | INESC Microsyst & Nanotechnol, Lisbon, P-1000 Portugal              |
> | and                                                                 |
> | R&D unit for  Molecular Chemical Physics                            |
> | Chemistry Department, University of Coimbra                         |
> =======================================================================
>
>
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
>
>


-- 
=======================================================================
Axel Kohlmeyer   akohlmey at cmm.chem.upenn.edu   http://www.cmm.upenn.edu
  Center for Molecular Modeling   --   University of Pennsylvania
Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323
tel: 1-215-898-1582,  fax: 1-215-573-6233,  office-tel: 1-215-898-5425
=======================================================================
If you make something idiot-proof, the universe creates a better idiot.


From mass at uniovi.es  Tue May 23 18:25:53 2006
From: mass at uniovi.es (=?iso-8859-1?Q?Miguel_A._Salvad=F3?=)
Date: Tue, 23 May 2006 18:25:53 +0200
Subject: [Pw_forum] symmetry operations
Message-ID: <01M2RROA8ORK009R47@fresno.net.uniovi.es>

 

Dear PWusers:

 

I do not understand how PWSCF generate the symmetry operations from ibrav
and crystal coordinates.

A simple example: C diamond (Fd-3m) with C at 8a (0,0,0) (Cell choice 1).

If I run a calculation with ibrav 2 (cubic F-centered) and two C atoms at
0,0,0 and 3/4,1/4,3/4 the program found 16 symmetry operations.

But if I change the second position to the equivalent one 1/4 1/4 1/4, then
the program found successfully 48 symmetry operations.

I have found this problem with more complex structures and it is not easy to
check all the possibilities of equivalent positions combinations.

Any kind of help will be welcome,

Best wishes,

 

Miguel A. Salvad?

 

 

Miguel A. Salvad?

Dpto. Quimica Fisica y Analitica

Universidad de Oviedo

Spain

E-mail: mass at uniovi.es

 

 

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: /pipermail/attachments/20060523/1c4788e2/attachment.htm 

From jcconesa at icp.csic.es  Tue May 23 18:35:22 2006
From: jcconesa at icp.csic.es (=?ISO-8859-1?Q?Jos=E9_C._Conesa?=)
Date: Tue, 23 May 2006 18:35:22 +0200
Subject: [Pw_forum] compilation in SGI
Message-ID: <447355EA.19131.411E56A6@localhost>

Hi,
I try to compile espresso 3.1 in a SGI Origin 200 system . When 
running configure there appear in the screen some error messages 
within the part of the output related to checking dependencies. This 
part is:

(.....)
directory upftools : ok
Cannot open ../iotk/src/iotk_attr+COMP: No such file or directory
pw_export.o : @iotk_module@
WARNING: dependencies not found in directory PP
Cannot open ../iotk/src/iotk_at: No such file or directory
directory PWCOND : ok
Cannot open ../iotk/src/iotk: No such file or directory
directory Gamma : ok
(.....)

otherwise, at the end it says "configure: success". But I'm afraid 
something is not runnig well; in particular it seems as if there was 
string truncation in some places, so before going into make all I wish to 
ask the experts whether something should be modified in configure or 
in some of the shell scripts launched by it.

Thanks in advance,




Jose C. Conesa
Instituto de Catalisis y Petroleoquimica, CSIC
Campus de Cantoblanco
28049 Madrid  -  Spain
Phone Nr. 34-91-5854766
Fax Nr. 34-91-5854760



From degironc at sissa.it  Tue May 23 19:44:22 2006
From: degironc at sissa.it (degironc)
Date: Tue, 23 May 2006 19:44:22 +0200
Subject: [Pw_forum] symmetry operations
In-Reply-To: <01M2RROA8ORK009R47@fresno.net.uniovi.es>
References: <01M2RROA8ORK009R47@fresno.net.uniovi.es>
Message-ID: <447349F6.2080004@sissa.it>

Are you using CRYSTAL coordinates?
in this case 3/4 1/4 3/4 and 1/4 1/4 1/4 are NOT equivalent.
the former is wrong and corrseponds to the following cartesian coordinates
(from the PWscf output)
         2           Si  tau(  2) = (  -0.7500000   0.5000000   0.5000000  )
which are obviously wrong.
stefano


Miguel A. Salvad? wrote:

>  
>
> Dear PWusers:
>
>  
>
> I do not understand how PWSCF generate the symmetry operations from 
> ibrav and crystal coordinates.
>
> A simple example: C diamond (Fd-3m) with C at 8a (0,0,0) (Cell choice 1).
>
> If I run a calculation with ibrav 2 (cubic F-centered) and two C atoms 
> at 0,0,0 and 3/4,1/4,3/4 the program found 16 symmetry operations.
>
> But if I change the second position to the equivalent one 1/4 1/4 1/4, 
> then the program found successfully 48 symmetry operations.
>
> I have found this problem with more complex structures and it is not 
> easy to check all the possibilities of equivalent positions combinations.
>
> Any kind of help will be welcome,
>
> Best wishes,
>
>  
>
> Miguel A. Salvad?
>
>  
>
>  
>
> Miguel A. Salvad?
>
> Dpto. Quimica Fisica y Analitica
>
> Universidad de Oviedo
>
> Spain
>
> E-mail: mass at uniovi.es
>
>  
>
>  
>
>  
>



From mass at uniovi.es  Tue May 23 19:27:27 2006
From: mass at uniovi.es (=?iso-8859-1?Q?Miguel_A._Salvad=F3?=)
Date: Tue, 23 May 2006 19:27:27 +0200
Subject: [Pw_forum] symmetry operations
In-Reply-To: <447349F6.2080004@sissa.it>
Message-ID: <01M2RTTM3XN4009RQ3@fresno.net.uniovi.es>

Yes, I am using CRYSTAL coordinates. 3/4 1/4 3/4 and 1/4 1/4 1/4 differs in
a 1/2 0 1/2 translation and then they are equivalent in a F-centered cell.
In fact 0 0 0 and 3/4 1/4 3/4 are the coordinates for the position 8a in
International Tables of Crystallography!!

Best wishes,
Miguel

-----Mensaje original-----
De: pw_forum-admin at pwscf.org [mailto:pw_forum-admin at pwscf.org] En nombre de
degironc
Enviado el: martes, 23 de mayo de 2006 19:44
Para: pw_forum at pwscf.org
Asunto: Re: [Pw_forum] symmetry operations

Are you using CRYSTAL coordinates?
in this case 3/4 1/4 3/4 and 1/4 1/4 1/4 are NOT equivalent.
the former is wrong and corrseponds to the following cartesian coordinates
(from the PWscf output)
         2           Si  tau(  2) = (  -0.7500000   0.5000000   0.5000000  )
which are obviously wrong.
stefano


Miguel A. Salvad? wrote:

>  
>
> Dear PWusers:
>
>  
>
> I do not understand how PWSCF generate the symmetry operations from 
> ibrav and crystal coordinates.
>
> A simple example: C diamond (Fd-3m) with C at 8a (0,0,0) (Cell choice 1).
>
> If I run a calculation with ibrav 2 (cubic F-centered) and two C atoms 
> at 0,0,0 and 3/4,1/4,3/4 the program found 16 symmetry operations.
>
> But if I change the second position to the equivalent one 1/4 1/4 1/4, 
> then the program found successfully 48 symmetry operations.
>
> I have found this problem with more complex structures and it is not 
> easy to check all the possibilities of equivalent positions combinations.
>
> Any kind of help will be welcome,
>
> Best wishes,
>
>  
>
> Miguel A. Salvad?
>
>  
>
>  
>
> Miguel A. Salvad?
>
> Dpto. Quimica Fisica y Analitica
>
> Universidad de Oviedo
>
> Spain
>
> E-mail: mass at uniovi.es
>
>  
>
>  
>
>  
>

_______________________________________________
Pw_forum mailing list
Pw_forum at pwscf.org
http://www.democritos.it/mailman/listinfo/pw_forum



From emenendez at macul.ciencias.uchile.cl  Tue May 23 23:14:31 2006
From: emenendez at macul.ciencias.uchile.cl (Eduardo Ariel Menendez P)
Date: Tue, 23 May 2006 17:14:31 -0400 (CLT)
Subject: [Pw_forum] phonon segmentation fault
Message-ID: <Pine.LNX.4.58.0605231638470.30205@macul.ciencias.uchile.cl>


Hello,
I m trynig to do a phonon calculation with ph.x version 3.1. The code
aborts like this
**********************************
     Representation   124      2 modes - To be done

     Representation   125      2 modes - To be done
     PHONON       :    15h13m CPU time

/home/eduardo/opt/mpich/1.2.7/bin/mpirun.ch_p4: line 243: 14349
Segmentation fault
"/home/eduardo/Compilaciones/Espresso/3.1/bin/ph.x" -p4pg
"/home/eduardo/MyPWSCF/CdTe/Defect/PI14169" -p4wd
"/home/eduardo/MyPWSCF/CdTe/Defect"

The example 02 seems to work ok.

Should I find the cause in MPICH compilation or in Espresso?
The system has 63 atoms? Is it too big for ph.x? Memory limited?

Thanks
Eduardo


From konstantin_kudin at yahoo.com  Wed May 24 05:24:43 2006
From: konstantin_kudin at yahoo.com (Konstantin Kudin)
Date: Tue, 23 May 2006 20:24:43 -0700 (PDT)
Subject: [Pw_forum] stress in CP and FFT grids
Message-ID: <20060524032443.41418.qmail@web52009.mail.yahoo.com>

 Hi all,

 It appears that the stress that comes out of the CP code is rather
sensitive to the FFT grids used.

 For example, I ran a constant pressure simulation ('vc-cp') started
from a certain unit cell. Naturally, the code kept all the FFT grids
unchanged throughout (generated via the initial cell dimensions), and
optimized the cell vectors to certain values.

 Then, when I started a fresh simulation with the optimized cell
dimensions, the stress was quite a bit different, on average about 0.6 
GPa higher than before. Looking at the settings, it appears that the
FFT grids have changed because of the fresh start, and that made the
stress different.

 Any suggestions on how to deal gracefully with such situations,
especially when the system is "soft" along certain cell coordinates? It
hardly comforting to have about 5% variations in the cell parameters
due to the grids ...

 Thanks!
 Kostya
 
 

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 


From degironc at sissa.it  Wed May 24 10:14:16 2006
From: degironc at sissa.it (degironc)
Date: Wed, 24 May 2006 10:14:16 +0200
Subject: [Pw_forum] symmetry operations
In-Reply-To: <01M2RTTM3XN4009RQ3@fresno.net.uniovi.es>
References: <01M2RTTM3XN4009RQ3@fresno.net.uniovi.es>
Message-ID: <447415D8.9060401@sissa.it>

quoting from "Wyckoff" about Diamond
BEGIN{QUOTE}
II,i1:
The unit cubes of crystals with the {\it diamond structure} have eight atoms
in the position of $O_h ^7$ ($Fd3m$):
 (8a) 000;0 1/2 1/2; 1/2 0 1/2; 1/2 1/2 0;
         1/4 1/4 1/4; 1/4 3/4 3/4; 3/4 1/4 3/4; 3/4 3/4 1/4
or more briefly
  (8a) 000; 1/4 1/4 1/4; F.C.
END{QUOTE}

This is fine ! BUT this refers to the cubic  (SIMPLE CUBIC) structure 
with 8 atoms per cell (the "F.C."  label tells you to replicate the 
previous 2-atom coordinates by  shifting them by
0 1/2 1/2; 1/2 0 1/2; 1/2 1/2 0.

If you chose the FACE CENTERED CUBIC cell you have only two atoms per 
unit cell
and different fundamental lattice vectors, hence 1/4 1/4 1/4 and 3/4 1/4 
3/4 in crystal coordinates (w.r.t  to these FCC fundamental lattice 
vectors , NOT to the simple cubic ones) are not equivalent.
If you chose CARTESIAN coordinates in unit of alat(the default for 
PWscf) then the two positions are obviously equivalent due to 
translational symmetry.

best regards,

stefano


Miguel A. Salvad? wrote:

>Yes, I am using CRYSTAL coordinates. 3/4 1/4 3/4 and 1/4 1/4 1/4 differs in
>a 1/2 0 1/2 translation and then they are equivalent in a F-centered cell.
>In fact 0 0 0 and 3/4 1/4 3/4 are the coordinates for the position 8a in
>International Tables of Crystallography!!
>
>Best wishes,
>Miguel
>
>-----Mensaje original-----
>De: pw_forum-admin at pwscf.org [mailto:pw_forum-admin at pwscf.org] En nombre de
>degironc
>Enviado el: martes, 23 de mayo de 2006 19:44
>Para: pw_forum at pwscf.org
>Asunto: Re: [Pw_forum] symmetry operations
>
>Are you using CRYSTAL coordinates?
>in this case 3/4 1/4 3/4 and 1/4 1/4 1/4 are NOT equivalent.
>the former is wrong and corrseponds to the following cartesian coordinates
>(from the PWscf output)
>         2           Si  tau(  2) = (  -0.7500000   0.5000000   0.5000000  )
>which are obviously wrong.
>stefano
>
>
>Miguel A. Salvad? wrote:
>
>  
>
>> 
>>
>>Dear PWusers:
>>
>> 
>>
>>I do not understand how PWSCF generate the symmetry operations from 
>>ibrav and crystal coordinates.
>>
>>A simple example: C diamond (Fd-3m) with C at 8a (0,0,0) (Cell choice 1).
>>
>>If I run a calculation with ibrav 2 (cubic F-centered) and two C atoms 
>>at 0,0,0 and 3/4,1/4,3/4 the program found 16 symmetry operations.
>>
>>But if I change the second position to the equivalent one 1/4 1/4 1/4, 
>>then the program found successfully 48 symmetry operations.
>>
>>I have found this problem with more complex structures and it is not 
>>easy to check all the possibilities of equivalent positions combinations.
>>
>>Any kind of help will be welcome,
>>
>>Best wishes,
>>
>> 
>>
>>Miguel A. Salvad?
>>
>> 
>>
>> 
>>
>>Miguel A. Salvad?
>>
>>Dpto. Quimica Fisica y Analitica
>>
>>Universidad de Oviedo
>>
>>Spain
>>
>>E-mail: mass at uniovi.es
>>
>> 
>>
>> 
>>
>> 
>>
>>    
>>
>
>_______________________________________________
>Pw_forum mailing list
>Pw_forum at pwscf.org
>http://www.democritos.it/mailman/listinfo/pw_forum
>
>_______________________________________________
>Pw_forum mailing list
>Pw_forum at pwscf.org
>http://www.democritos.it/mailman/listinfo/pw_forum
>  
>



From mass at uniovi.es  Wed May 24 13:20:31 2006
From: mass at uniovi.es (=?iso-8859-1?Q?Miguel_A._Salvad=F3?=)
Date: Wed, 24 May 2006 13:20:31 +0200
Subject: [Pw_forum] symmetry operations
In-Reply-To: <447415D8.9060401@sissa.it>
Message-ID: <01M2SVB258OK009Z64@fresno.net.uniovi.es>

Well, from a practical point of view: If I have a complex structure whose
space group (let say cubic), cell parameters and crystal coordinates of the
asymmetric unit I know, How is the simplest form to input these atomic
coordinates in order the program recognize correctly the structure and all
the symmetry operations present?

Thanks and best regards,

Miguel Angel

-----Mensaje original-----
De: pw_forum-admin at pwscf.org [mailto:pw_forum-admin at pwscf.org] En nombre de
degironc
Enviado el: mi?rcoles, 24 de mayo de 2006 10:14
Para: pw_forum at pwscf.org
Asunto: Re: [Pw_forum] symmetry operations

quoting from "Wyckoff" about Diamond
BEGIN{QUOTE}
II,i1:
The unit cubes of crystals with the {\it diamond structure} have eight atoms
in the position of $O_h ^7$ ($Fd3m$):
 (8a) 000;0 1/2 1/2; 1/2 0 1/2; 1/2 1/2 0;
         1/4 1/4 1/4; 1/4 3/4 3/4; 3/4 1/4 3/4; 3/4 3/4 1/4
or more briefly
  (8a) 000; 1/4 1/4 1/4; F.C.
END{QUOTE}

This is fine ! BUT this refers to the cubic  (SIMPLE CUBIC) structure 
with 8 atoms per cell (the "F.C."  label tells you to replicate the 
previous 2-atom coordinates by  shifting them by
0 1/2 1/2; 1/2 0 1/2; 1/2 1/2 0.

If you chose the FACE CENTERED CUBIC cell you have only two atoms per 
unit cell
and different fundamental lattice vectors, hence 1/4 1/4 1/4 and 3/4 1/4 
3/4 in crystal coordinates (w.r.t  to these FCC fundamental lattice 
vectors , NOT to the simple cubic ones) are not equivalent.
If you chose CARTESIAN coordinates in unit of alat(the default for 
PWscf) then the two positions are obviously equivalent due to 
translational symmetry.

best regards,

stefano


Miguel A. Salvad? wrote:

>Yes, I am using CRYSTAL coordinates. 3/4 1/4 3/4 and 1/4 1/4 1/4 differs in
>a 1/2 0 1/2 translation and then they are equivalent in a F-centered cell.
>In fact 0 0 0 and 3/4 1/4 3/4 are the coordinates for the position 8a in
>International Tables of Crystallography!!
>
>Best wishes,
>Miguel
>
>-----Mensaje original-----
>De: pw_forum-admin at pwscf.org [mailto:pw_forum-admin at pwscf.org] En nombre de
>degironc
>Enviado el: martes, 23 de mayo de 2006 19:44
>Para: pw_forum at pwscf.org
>Asunto: Re: [Pw_forum] symmetry operations
>
>Are you using CRYSTAL coordinates?
>in this case 3/4 1/4 3/4 and 1/4 1/4 1/4 are NOT equivalent.
>the former is wrong and corrseponds to the following cartesian coordinates
>(from the PWscf output)
>         2           Si  tau(  2) = (  -0.7500000   0.5000000   0.5000000
)
>which are obviously wrong.
>stefano
>
>
>Miguel A. Salvad? wrote:
>
>  
>
>> 
>>
>>Dear PWusers:
>>
>> 
>>
>>I do not understand how PWSCF generate the symmetry operations from 
>>ibrav and crystal coordinates.
>>
>>A simple example: C diamond (Fd-3m) with C at 8a (0,0,0) (Cell choice 1).
>>
>>If I run a calculation with ibrav 2 (cubic F-centered) and two C atoms 
>>at 0,0,0 and 3/4,1/4,3/4 the program found 16 symmetry operations.
>>
>>But if I change the second position to the equivalent one 1/4 1/4 1/4, 
>>then the program found successfully 48 symmetry operations.
>>
>>I have found this problem with more complex structures and it is not 
>>easy to check all the possibilities of equivalent positions combinations.
>>
>>Any kind of help will be welcome,
>>
>>Best wishes,
>>
>> 
>>
>>Miguel A. Salvad?
>>
>> 
>>
>> 
>>
>>Miguel A. Salvad?
>>
>>Dpto. Quimica Fisica y Analitica
>>
>>Universidad de Oviedo
>>
>>Spain
>>
>>E-mail: mass at uniovi.es
>>
>> 
>>
>> 
>>
>> 
>>
>>    
>>
>
>_______________________________________________
>Pw_forum mailing list
>Pw_forum at pwscf.org
>http://www.democritos.it/mailman/listinfo/pw_forum
>
>_______________________________________________
>Pw_forum mailing list
>Pw_forum at pwscf.org
>http://www.democritos.it/mailman/listinfo/pw_forum
>  
>

_______________________________________________
Pw_forum mailing list
Pw_forum at pwscf.org
http://www.democritos.it/mailman/listinfo/pw_forum



From ferretti.andrea at unimore.it  Wed May 24 13:57:38 2006
From: ferretti.andrea at unimore.it (Andrea Ferretti)
Date: Wed, 24 May 2006 13:57:38 +0200 (CEST)
Subject: [Pw_forum] compilation in SGI
In-Reply-To: <447355EA.19131.411E56A6@localhost>
References: <447355EA.19131.411E56A6@localhost>
Message-ID: <Pine.LNX.4.58.0605231841020.25402@bamboo.fisica.unimo.it>


Hi Jose,

I think I found a similar problem on an old SGI O2 machine...
as it seems in your case, my trouble was related to the truncation of some 
strings in the moduldep.sh script (in particular the variables $sources 
and $sources_all in the lines with egrep)

I tried with some stupid fix (eg, doing the egrep file by file inside a 
loop) , but found nothing interesting.... some more serious hacking is 
needed :-), ie, those variables must be made actually shorter... 

anyway,
if you only need to compile PW and PP it is much easier (the usual 
"quick and dirty" way) to edit PP/make.depend  and substitute

pw_export.o : @iotk_module@  

by

pw_export.o : ../iotk/src/iotk_module.o

it works fine for me...

 andrea


On Tue, 23 May 2006, Jos? C. Conesa wrote:

> Hi,
> I try to compile espresso 3.1 in a SGI Origin 200 system . When 
> running configure there appear in the screen some error messages 
> within the part of the output related to checking dependencies. This 
> part is:
> 
> (.....)
> directory upftools : ok
> Cannot open ../iotk/src/iotk_attr+COMP: No such file or directory
> pw_export.o : @iotk_module@
> WARNING: dependencies not found in directory PP
> Cannot open ../iotk/src/iotk_at: No such file or directory
> directory PWCOND : ok
> Cannot open ../iotk/src/iotk: No such file or directory
> directory Gamma : ok
> (.....)
> 
> otherwise, at the end it says "configure: success". But I'm afraid 
> something is not runnig well; in particular it seems as if there was 
> string truncation in some places, so before going into make all I wish to 
> ask the experts whether something should be modified in configure or 
> in some of the shell scripts launched by it.
> 
> Thanks in advance,
> 
> 
> 
> 
> Jose C. Conesa
> Instituto de Catalisis y Petroleoquimica, CSIC
> Campus de Cantoblanco
> 28049 Madrid  -  Spain
> Phone Nr. 34-91-5854766
> Fax Nr. 34-91-5854760
> 
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
> 


From akohlmey at vitae.cmm.upenn.edu  Wed May 24 17:42:06 2006
From: akohlmey at vitae.cmm.upenn.edu (Axel Kohlmeyer)
Date: Wed, 24 May 2006 11:42:06 -0400 (EDT)
Subject: [Pw_forum] stress in CP and FFT grids
In-Reply-To: <20060524032443.41418.qmail@web52009.mail.yahoo.com>
Message-ID: <Pine.LNX.4.44.0605241114510.12083-100000@localhost.localdomain>

On Tue, 23 May 2006, Konstantin Kudin wrote:

KK>  Hi all,

hi kostya,

KK>  It appears that the stress that comes out of the CP code is rather
KK> sensitive to the FFT grids used.

what kind of cutoffs are you using? 
does your system expand or shrink? and how much?

of course the results will always depend on the grids 
(which are defined implicitely by plane wave and density 
cutoff). due to the implementation, you cannot easily 
change grids during the simulation. so you _always_ have 
to restart or use a higher cutoff to start with, so you 
won't have too much troubles with the effects you are 
seeing (and also to reduce pulay stress).

best regards,
   axel.

KK>  For example, I ran a constant pressure simulation ('vc-cp') started
KK> from a certain unit cell. Naturally, the code kept all the FFT grids
KK> unchanged throughout (generated via the initial cell dimensions), and
KK> optimized the cell vectors to certain values.
KK> 
KK>  Then, when I started a fresh simulation with the optimized cell
KK> dimensions, the stress was quite a bit different, on average about 0.6 
KK> GPa higher than before. Looking at the settings, it appears that the
KK> FFT grids have changed because of the fresh start, and that made the
KK> stress different.
KK> 
KK>  Any suggestions on how to deal gracefully with such situations,
KK> especially when the system is "soft" along certain cell coordinates? It
KK> hardly comforting to have about 5% variations in the cell parameters
KK> due to the grids ...
KK> 
KK>  Thanks!
KK>  Kostya
KK>  
KK>  
KK> 
KK> __________________________________________________
KK> Do You Yahoo!?
KK> Tired of spam?  Yahoo! Mail has the best spam protection around 
KK> http://mail.yahoo.com 
KK> _______________________________________________
KK> Pw_forum mailing list
KK> Pw_forum at pwscf.org
KK> http://www.democritos.it/mailman/listinfo/pw_forum
KK> 

-- 
=======================================================================
Axel Kohlmeyer   akohlmey at cmm.chem.upenn.edu   http://www.cmm.upenn.edu
   Center for Molecular Modeling   --   University of Pennsylvania
Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323
tel: 1-215-898-1582,  fax: 1-215-573-6233,  office-tel: 1-215-898-5425
=======================================================================
If you make something idiot-proof, the universe creates a better idiot.




From burde1jb at cmich.edu  Wed May 24 18:20:04 2006
From: burde1jb at cmich.edu (John B Burdette)
Date: Wed, 24 May 2006 12:20:04 -0400 (EDT)
Subject: [Pw_forum] Negative PDOS and Graphing of
 Energy Bands
Message-ID: <20060524122004.BKK39174@student-b.csv.cmich.edu>

Hello everyone,

I am currently working with CaF2 and I recently used PWSCF to 
obtain the energy bands and density of states.  Here I have 
two questions for anyone willing to help me.

1. When observing the graph for the density of states, 
everything appeared okay and was positive.  However, using 
the program projwfc.x I obtained the projected density of 
states and when I looked at the graph for the total PDOS, I 
saw areas of negative values before each spike.  Can anyone 
explain how a negative value for the PDOS could occur and 
also if this is actually a problem I should worry about?

2. I also looked at the Energy Band Structure for CaF2 and I 
am unaware of what the independent variable used for plotting 
the energies from the program bands.x and plotband.x is.  Can 
someone please help me here?

Best,
John


From konstantin_kudin at yahoo.com  Wed May 24 20:41:23 2006
From: konstantin_kudin at yahoo.com (Konstantin Kudin)
Date: Wed, 24 May 2006 11:41:23 -0700 (PDT)
Subject: [Pw_forum] Re: stress in CP and FFT grids
In-Reply-To: <Pine.LNX.4.44.0605241114510.12083-100000@localhost.localdomain>
Message-ID: <20060524184123.78257.qmail@web52006.mail.yahoo.com>

 Hi Axel!

> what kind of cutoffs are you using? 
> does your system expand or shrink? and how much?

 Ordinarily, I would use the cutoff of 25 Ryd, however, for the
constant pressure I was using 30 Ryd with the modified kinetic energy
functional 
[ecutwfc = 30.0, ecfixed = 25.0, qcutz = 25.0, q2sigma = 3.0]

 I started with Z=22 A, the grid was:
   Array leading dimensions ( nr1x, nr2x, nr3x )   =     90    90   192
   Array leading dimensions ( nr1x, nr2x, nr3x )   =     72    72   150
   Array leading dimensions ( nr1x, nr2x, nr3x )   =     24    24    24

 It settled around Z=23 A (I optimized only Z), and if I restart a
fresh calculation with all the same parameters, i get:

   Array leading dimensions ( nr1x, nr2x, nr3x )   =     90    90   200
   Array leading dimensions ( nr1x, nr2x, nr3x )   =     72    72   160
   Array leading dimensions ( nr1x, nr2x, nr3x )   =     24    24    24
 
 So the change in the grid is pretty much unaccounted for by the
constant pressure simulation. I guess I could force the Z=23A run to
use the same grid as before, however, that is not so clean.

 Thanks!
 Kostya

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 


From akohlmey at cmm.upenn.edu  Wed May 24 22:20:40 2006
From: akohlmey at cmm.upenn.edu (Axel Kohlmeyer)
Date: Wed, 24 May 2006 16:20:40 -0400
Subject: [Pw_forum] Re: stress in CP and FFT grids
In-Reply-To: <20060524184123.78257.qmail@web52006.mail.yahoo.com>
References: <Pine.LNX.4.44.0605241114510.12083-100000@localhost.localdomain>
	 <20060524184123.78257.qmail@web52006.mail.yahoo.com>
Message-ID: <7b6913e90605241320t2dd08ecbjdbee4e456ea1a02d@mail.gmail.com>

On 5/24/06, Konstantin Kudin <konstantin_kudin at yahoo.com> wrote:
>  Hi Axel!
>
> > what kind of cutoffs are you using?
> > does your system expand or shrink? and how much?
>
>  Ordinarily, I would use the cutoff of 25 Ryd, however, for the
> constant pressure I was using 30 Ryd with the modified kinetic energy
> functional
> [ecutwfc = 30.0, ecfixed = 25.0, qcutz = 25.0, q2sigma = 3.0]

ok, i have very little experience how well this works with ultra-soft
pseudopotentials, but i suggest you try with a higher plane wave
cutoff, e.g. 35/40 nevertheless and a too large system. this way
your basis set can only get 'better'  when the cell shrinks and not
the other way around. at least this is how it worked for me with
norm-conserving pseudopotentials (and a different code, too).

of course the 'softer' the potential curve with respect to the
box size, the better your stress tensor has to be converged
with respect to the basis set and the density grid.

>  I started with Z=22 A, the grid was:
>    Array leading dimensions ( nr1x, nr2x, nr3x )   =     90    90   192
>    Array leading dimensions ( nr1x, nr2x, nr3x )   =     72    72   150
>    Array leading dimensions ( nr1x, nr2x, nr3x )   =     24    24    24
>
>  It settled around Z=23 A (I optimized only Z), and if I restart a
> fresh calculation with all the same parameters, i get:
>
>    Array leading dimensions ( nr1x, nr2x, nr3x )   =     90    90   200
>    Array leading dimensions ( nr1x, nr2x, nr3x )   =     72    72   160
>    Array leading dimensions ( nr1x, nr2x, nr3x )   =     24    24    24
>
>  So the change in the grid is pretty much unaccounted for by the
> constant pressure simulation. I guess I could force the Z=23A run to
> use the same grid as before, however, that is not so clean.

no you want it the other way around. you cannot change the grid, so
you want the code to use the finest grid that you might need during the
run. i.e. something corresponding to the maximally expected cell size.

best regards,
     axel.

>
>  Thanks!
>  Kostya
>
> __________________________________________________
> Do You Yahoo!?
> Tired of spam?  Yahoo! Mail has the best spam protection around
> http://mail.yahoo.com
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
>
>


-- 
=======================================================================
Axel Kohlmeyer   akohlmey at cmm.chem.upenn.edu   http://www.cmm.upenn.edu
  Center for Molecular Modeling   --   University of Pennsylvania
Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323
tel: 1-215-898-1582,  fax: 1-215-573-6233,  office-tel: 1-215-898-5425
=======================================================================
If you make something idiot-proof, the universe creates a better idiot.


From giannozz at nest.sns.it  Thu May 25 09:54:19 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Thu, 25 May 2006 09:54:19 +0200
Subject: [Pw_forum] Negative PDOS and Graphing of Energy Bands
In-Reply-To: <20060524122004.BKK39174@student-b.csv.cmich.edu>
References: <20060524122004.BKK39174@student-b.csv.cmich.edu>
Message-ID: <200605250954.19131.giannozz@nest.sns.it>

On Wednesday 24 May 2006 18:20, John B Burdette wrote:

> [...] Can anyone explain how a negative value for the PDOS 
> could occur

the PDOS is calculated using smearing. Are you by any chance
using Methfessel-Paxton broadening (ngauss=1) ? it may produce
negative weights. Try simple gaussian, or Marzari-Vanderbilt 
(cold) smearing: ngauss=0 and -1 respectively. Of course it would
be nicer to use tetrahedra, but it has to be implemented (I heard 
that somebody had actually done this, but I have never see any
code)

> and also if this is actually a problem I should  worry about? 

if the negative values are small, you shouldn't, but of cource it
doesn't look nice, especially in a figure for publication!

> 2. I also looked at the Energy Band Structure for CaF2 and I
> am unaware of what the independent variable used for plotting
> the energies from the program bands.x and plotband.x is.

you mean: the x-axis of the plot? it is k, along the lines that are
supplied in input, as in any respectable band structure plot

Paolo
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From giannozz at nest.sns.it  Thu May 25 10:36:26 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Thu, 25 May 2006 10:36:26 +0200
Subject: [Pw_forum] symmetry operations
In-Reply-To: <01M2SVB258OK009Z64@fresno.net.uniovi.es>
References: <01M2SVB258OK009Z64@fresno.net.uniovi.es>
Message-ID: <200605251036.26583.giannozz@nest.sns.it>

On Wednesday 24 May 2006 13:20, Miguel A. Salvad? wrote:

> Well, from a practical point of view: If I have a complex structure
> whose space group (let say cubic), cell parameters and crystal 
> coordinates of the asymmetric unit I know, How is the simplest form 
> to input these atomic coordinates in order the program recognize 
> correctly the structure and all the symmetry operations present? 

you can specify the cell parameters in three different ways:
- as a bravais lattice (as defined in variable "ibrav") and
  cell parameters (as defined in variable "celldm")
- using traditional crystallographic parameters :
  a, b, c, cos(ab), cos(ac), cos(bc)
- providing your own set of crystal axis
and atomic positions in
- cartesian coordinates, either A or a.u., or
- crystal coordinates, with respect to the axis as given in input
  or calculated by the code
The symmetry is then automatically calculated. All of the above
is accurately documented in detail. If the symmetry you think
is there is not found,  it is because it is not there, and why it
is not there is explained in detail in the manual (for instance:
http://www.pwscf.org/guide/3.0/html-node/node59.html
http://www.pwscf.org/guide/3.0/html-node/node60.html )

Paolo
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From giannozz at nest.sns.it  Thu May 25 10:40:27 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Thu, 25 May 2006 10:40:27 +0200
Subject: [Pw_forum] symmetry operations
In-Reply-To: <200605251036.26583.giannozz@nest.sns.it>
References: <01M2SVB258OK009Z64@fresno.net.uniovi.es> <200605251036.26583.giannozz@nest.sns.it>
Message-ID: <200605251040.27971.giannozz@nest.sns.it>

On Thursday 25 May 2006 10:36, Paolo Giannozzi wrote:

> and atomic positions in
> - cartesian coordinates, either A or a.u., 

and also in units of the lattice parameter a=celldm(1)

-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From giannozz at nest.sns.it  Thu May 25 11:20:52 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Thu, 25 May 2006 11:20:52 +0200
Subject: [Pw_forum] phonon segmentation fault
In-Reply-To: <Pine.LNX.4.58.0605231638470.30205@macul.ciencias.uchile.cl>
References: <Pine.LNX.4.58.0605231638470.30205@macul.ciencias.uchile.cl>
Message-ID: <200605251120.52623.giannozz@nest.sns.it>

On Tuesday 23 May 2006 23:14, Eduardo Ariel Menendez P wrote:

> I m trying to do a phonon calculation with ph.x version 3.1. The code
> aborts like this
> **********************************
>      Representation   124      2 modes - To be done
>
>      Representation   125      2 modes - To be done
>      PHONON       :    15h13m CPU time
>
> /home/eduardo/opt/mpich/1.2.7/bin/mpirun.ch_p4: line 243: 14349
> Segmentation fault [...]
> Should I find the cause in MPICH compilation or in Espresso?
> The system has 63 atoms? Is it too big for ph.x? Memory limited?

hard to say, but it is very likely that memory is the reason.
The phonon code, especially the ultrasoft case, is not well 
optimized for memory usage in large calculations. A few
places where excessive memory was required have been
fixed, but more work is needed

Paolo
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From mass at uniovi.es  Thu May 25 11:31:55 2006
From: mass at uniovi.es (=?iso-8859-1?Q?Miguel_A._Salvad=F3?=)
Date: Thu, 25 May 2006 11:31:55 +0200
Subject: [Pw_forum] symmetry operations
In-Reply-To: <200605251040.27971.giannozz@nest.sns.it>
Message-ID: <01M2U5SS0VH800AA3W@fresno.net.uniovi.es>

Thanks Stefano and Paolo. I think now I am able to run my calculations OK.

Best wishes

Miguel Angel

-----Mensaje original-----
De: pw_forum-admin at pwscf.org [mailto:pw_forum-admin at pwscf.org] En nombre de
Paolo Giannozzi
Enviado el: jueves, 25 de mayo de 2006 10:40
Para: pw_forum at pwscf.org
Asunto: Re: [Pw_forum] symmetry operations

On Thursday 25 May 2006 10:36, Paolo Giannozzi wrote:

> and atomic positions in
> - cartesian coordinates, either A or a.u., 

and also in units of the lattice parameter a=celldm(1)

-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy
_______________________________________________
Pw_forum mailing list
Pw_forum at pwscf.org
http://www.democritos.it/mailman/listinfo/pw_forum



From burde1jb at cmich.edu  Thu May 25 16:14:01 2006
From: burde1jb at cmich.edu (John B Burdette)
Date: Thu, 25 May 2006 10:14:01 -0400 (EDT)
Subject: [Pw_forum] Re: Negative PDOS and Graphing
 of Energy Bands
Message-ID: <20060525101401.BKL63011@student-b.csv.cmich.edu>

Hi Paolo,

Regarding the smearing, yes indeed I was using the Methfessel-
Paxton broadening.  When I used the simple Gaussian 
broadening, I received only positive weights.  Thank you very 
much Paolo and Andrea.

Also thank you for clearing up the independent variable 
question.  I believed they were the k-points, but I could not 
find this information in the code so I felt inclined to ask.

Thank you very much,

John


From yan at im.imag.kiev.ua  Thu May 25 18:17:02 2006
From: yan at im.imag.kiev.ua (B. Yanchitsky)
Date: Thu, 25 May 2006 19:17:02 +0300
Subject: [Pw_forum] wrong para?
Message-ID: <4475D87E.3080308@im.imag.kiev.ua>

Dear users,

There is a strange thing in parallel version of pw.x.
Parallel version running on 2 processors (cores) (dual-core pentium-d, mpich, npool=2)
and the same pw.x code running on single processor (core) produce different results.
As example
First iteration for 2 processors
      total energy              =  -154.98769308 ryd
First iteration for 1 processor
      total energy              =  -154.98786918 ryd


2 processors
!    total energy              =  -155.02553035 ryd
1 processor
!    total energy              =  -155.02553039 ryd

Forces are also different
2 processors
      atom  61 type  1   force =    -0.00042175   -0.00024350    0.00000000
1 processor
      atom  61 type  1   force =    -0.00042242   -0.00024388    0.00000000


My configuration is ifort 9.0 + mkl 8.0.2.
Could you give any ideas/propositions?
Any additional info will be sent on your request from my side.

Regards,
Bogdan Yanchitsky

Institute of Magnetism
Vernadsky Blvd., 36-b
03142  Kiev
UKRAINE

Tel. (+380-44) 444 34 20
Fax. (+380-44) 444 10 20


From jess.kondor at gmail.com  Thu May 25 20:59:39 2006
From: jess.kondor at gmail.com (Jess Kondor)
Date: Thu, 25 May 2006 13:59:39 -0500
Subject: [Pw_forum] Ecutrho's choice
Message-ID: <9d7293e10605251159t2c1bde16hfdc233f6bd8a8b48@mail.gmail.com>

 Hi all,

  I have some doubts with choice of ecutrho. For my test material, I use
ultrasoft pseudopotentials, LDA. Using high ecutrho=400, I found that
Ecutwfc=40 Ry, 4x4x4 k-point mesh is enough in order to achieve convergence.
Then, I fixed these parameters and changed ecutrho. I found, that properties
of my test system [rhombohedral unit cell] (total energy, lattice constant,
angle between cell vectors, unit-cell volume) are almost identical to one
obtained by using ecutrho=160 Ry. :

===================================================================================================
 EcutRho (Ry)    Etot (Ry)                 alat (Ang)
Angle                           volume (ang**3)
===================================================================================================
   160       -983.62434066       5.47150468539004      60.1841848177694
116.30927040909165860688
   200       -983.62458218       5.47150604640841      60.1841524430312
116.30927040909165860688
   240       -983.62458762       5.47150640562892      60.1841431281284
116.30927040909165860688
   280       -983.62458015       5.47150677576742      60.1841338128593
116.30927040909165860688
   320       -983.62458383       5.47150716228288      60.1841244970402
116.30927040909165860688
   360       -983.62456511       5.47150753788046      60.1841151815896
116.30927040909165860688
   400       -983.62456736       5.47150792439603      60.1841058657727
116.30927040909165860688
   440       -983.62456667       5.47150829999372      60.1840965503242
116.30927040909165860688
   480       -983.62456522       5.47150868104283      60.1840871687159
116.30927040909165860688
   520       -983.62456799       5.47150905663303      60.1840777872924
116.30927040909165860688

I am a little bit confused, because in ultrasoft case ecutrho should be >
4*ecutwfc  . My system contains 3 atomic species, one of them is iron. I use
Fe.pz-sp-van_ak.UPF pseudopotential. I read the info about these
pseudopotentials (unfortunately, for GGA only). I saw that recommended value
of ecutho should be around 10-12*ecutwfc. What is the criteria for that
choice?

  Best regards,
    JK
-------------- next part --------------
An HTML attachment was scrubbed...
URL: /pipermail/attachments/20060525/b1456748/attachment.htm 

From proffess at yandex.ru  Fri May 26 00:56:30 2006
From: proffess at yandex.ru (Sergey Lisenkov)
Date: Fri, 26 May 2006 02:56:30 +0400 (MSD)
Subject: [Pw_forum] stability of fftw3 in qe-3.1
Message-ID: <4476361E.000001.26510@tide.yandex.ru>

 Dear developers,

  I tried to run a simple calculation of SrRuO3 using pwscf 3.1 version linked with fftw-3.1.1 library on intel xeon em64t cluster. Code crashed before the first scf iteration (I see that integrated charge is a very huge value). But when I used the same pwscf version but linked with fftw2 library, calculation is going smoothly. It seems that it is fftw issue. Is it safe way to use fftw3 library ?

  Thanks,
    Sergey


From dkorotin at gmail.com  Fri May 26 07:11:33 2006
From: dkorotin at gmail.com (Dmitry Korotin)
Date: Fri, 26 May 2006 11:11:33 +0600
Subject: [Pw_forum] ld1.x and test
Message-ID: <2fd252650605252211k1980075ai7907adff155840eb@mail.gmail.com>

Dear ESPRESSO users,
I'm trying to reproduce Np pseudopotential from the article PRB 66, 235112

after pseudo-generation run I have following output:
     n l  j  nl             e AE (Ryd)       e PS (Ryd)   De AE-PS (Ryd)
     2 1 0.5 6P   1( 2.00)       -3.27104       -3.27129        0.00025
     2 1 1.5 6P   1( 4.00)       -2.55754       -2.55786        0.00032
     3 2 1.5 6D   1( 1.00)       -1.09034       -1.09045        0.00011
     3 2 2.5 6D   1( 0.00)       -1.03768       -1.03782        0.00014
     4 3 2.5 5F   1( 4.00)       -1.30480       -1.30299       -0.00181
     4 3 3.5 5F   1( 0.00)       -1.23331       -1.23252       -0.00079
     1 0 0.5 6S   1( 2.00)       -4.62634       -4.62636        0.00003
(De AE-PS (Ryd) seems ok)

but after pseudo-test run with the same electrons configuration I have
following:
     n l  j  nl             e AE (Ryd)       e PS (Ryd)   De AE-PS (Ryd)
     4 3 2.5 5F   1( 4.00)       -1.30480       -9.85980        8.55500
     4 3 3.5 5F   1( 0.00)       -1.23331       -9.75683        8.52352
     3 2 1.5 6D   1( 1.00)       -1.09034       -8.00574        6.91540
     3 2 2.5 6D   1( 0.00)       -1.03768       -7.85645        6.81877
     1 0 0.5 7S   1( 0.00)       -1.12548      -12.93386       11.80838

I think that results should be exactly the same and I confused.
Do you have any suggestions?


Thank you in advance,
Dmitry Korotin.


From dkorotin at gmail.com  Fri May 26 07:18:49 2006
From: dkorotin at gmail.com (Dmitry Korotin)
Date: Fri, 26 May 2006 11:18:49 +0600
Subject: [Pw_forum] Re: ld1.x and test
In-Reply-To: <2fd252650605252211k1980075ai7907adff155840eb@mail.gmail.com>
References: <2fd252650605252211k1980075ai7907adff155840eb@mail.gmail.com>
Message-ID: <2fd252650605252218t50640911ncec0cfa106048476@mail.gmail.com>

Sorry, I forgot to attach input and test files in previous mail.

Dmitry Korotin.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: pg.in
Type: application/octet-stream
Size: 477 bytes
Desc: not available
Url : /pipermail/attachments/20060526/097253f2/attachment.obj 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: pt.in
Type: application/octet-stream
Size: 211 bytes
Desc: not available
Url : /pipermail/attachments/20060526/097253f2/attachment-0001.obj 

From konstantin_kudin at yahoo.com  Fri May 26 08:29:23 2006
From: konstantin_kudin at yahoo.com (Konstantin Kudin)
Date: Thu, 25 May 2006 23:29:23 -0700 (PDT)
Subject: [Pw_forum] wrong para?
In-Reply-To: <4475D87E.3080308@im.imag.kiev.ua>
Message-ID: <20060526062923.69787.qmail@web52006.mail.yahoo.com>

 Dear Bogdan,

 Different results actually imply something that differs by say, ~ 0.1
Ryd at convergence, not 4e-8 ... Your results are perfectly fine! You
should expect differences when you change number of cpus, or the
computer architecture.

 It'd be great if the codes could always produce the same numerical
answer with respect to the double precision accuracy, but I am afraid
that'll never happen.

 Cheers,
 Kostya
 

--- "B. Yanchitsky" <yan at im.imag.kiev.ua> wrote:

> Dear users,
> 
> There is a strange thing in parallel version of pw.x.
> Parallel version running on 2 processors (cores) (dual-core
> pentium-d, mpich, npool=2)
> and the same pw.x code running on single processor (core) produce
> different results.
> As example
> First iteration for 2 processors
>       total energy              =  -154.98769308 ryd
> First iteration for 1 processor
>       total energy              =  -154.98786918 ryd
> 
> 
> 2 processors
> !    total energy              =  -155.02553035 ryd
> 1 processor
> !    total energy              =  -155.02553039 ryd
> 
> Forces are also different
> 2 processors
>       atom  61 type  1   force =    -0.00042175   -0.00024350   
> 0.00000000
> 1 processor
>       atom  61 type  1   force =    -0.00042242   -0.00024388   
> 0.00000000
> 
> 
> My configuration is ifort 9.0 + mkl 8.0.2.
> Could you give any ideas/propositions?
> Any additional info will be sent on your request from my side.
> 
> Regards,
> Bogdan Yanchitsky
> 
> Institute of Magnetism
> Vernadsky Blvd., 36-b
> 03142  Kiev
> UKRAINE
> 
> Tel. (+380-44) 444 34 20
> Fax. (+380-44) 444 10 20
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
> 


__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 


From ding at sissa.it  Fri May 26 08:40:37 2006
From: ding at sissa.it (Xunlei Ding)
Date: Fri, 26 May 2006 08:40:37 +0200
Subject: [Pw_forum] Is it possible to convert the file format between 32 and 64 system?
Message-ID: <4476A2E5.7000508@sissa.it>

Dear All,

I do calculations on SP5 and XD1. The wfc and rho files can not be used
when I change for one to the other,
because they have 64 and 32 bit file system.
Are there any tools to convert the files from one format to the other?

Thank you!

Yours,
Ding




From giannozz at nest.sns.it  Fri May 26 10:00:26 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Fri, 26 May 2006 10:00:26 +0200
Subject: [Pw_forum] stability of fftw3 in qe-3.1
In-Reply-To: <4476361E.000001.26510@tide.yandex.ru>
References: <4476361E.000001.26510@tide.yandex.ru>
Message-ID: <200605261000.26619.giannozz@nest.sns.it>

On Friday 26 May 2006 00:56, Sergey Lisenkov wrote:

>   I tried to run a simple calculation of SrRuO3 using pwscf 3.1 version
> linked with fftw-3.1.1 library on intel xeon em64t cluster. Code crashed
> [...] But when I used the same pwscf version but linked with fftw2
> library, calculation is going smoothly. It seems that it is fftw issue. Is
> it safe way to use fftw3 library ?

maybe, or maybe not.

Drivers for the fftw3 library were kindly provided by Pascal Thibaudeau 
and Stephane Lefranc, ported to the new version by myself and tested only
on a 32-bit PC in a few cases. They sort of worked, i.e. they produced correct
results in most cases, but crashed for no apparent good reason in other 
cases. Anybody who is willing to investigate this issue is welcome.

Paolo
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From dalcorso at sissa.it  Fri May 26 10:17:58 2006
From: dalcorso at sissa.it (Andrea Dal Corso)
Date: Fri, 26 May 2006 10:17:58 +0200
Subject: [Pw_forum] ld1.x and test
In-Reply-To: <2fd252650605252211k1980075ai7907adff155840eb@mail.gmail.com>
References: <2fd252650605252211k1980075ai7907adff155840eb@mail.gmail.com>
Message-ID: <1148631479.4007.4.camel@dhpc-5-58.sissa.it>

Your test configuration is not the same as in the pseudopotential
generation. You have 5 electrons in the test and 13 electrons in the
pseudo generation.

Try with:

        configts(1) = '6s2 6p2 6p4 5f4 6d1 7s0'

Best Regards,

Andrea


On Fri, 2006-05-26 at 11:11 +0600, Dmitry Korotin wrote:
> Dear ESPRESSO users,
> I'm trying to reproduce Np pseudopotential from the article PRB 66, 235112
> 
> after pseudo-generation run I have following output:
>      n l  j  nl             e AE (Ryd)       e PS (Ryd)   De AE-PS (Ryd)
>      2 1 0.5 6P   1( 2.00)       -3.27104       -3.27129        0.00025
>      2 1 1.5 6P   1( 4.00)       -2.55754       -2.55786        0.00032
>      3 2 1.5 6D   1( 1.00)       -1.09034       -1.09045        0.00011
>      3 2 2.5 6D   1( 0.00)       -1.03768       -1.03782        0.00014
>      4 3 2.5 5F   1( 4.00)       -1.30480       -1.30299       -0.00181
>      4 3 3.5 5F   1( 0.00)       -1.23331       -1.23252       -0.00079
>      1 0 0.5 6S   1( 2.00)       -4.62634       -4.62636        0.00003
> (De AE-PS (Ryd) seems ok)
> 
> but after pseudo-test run with the same electrons configuration I have
> following:
>      n l  j  nl             e AE (Ryd)       e PS (Ryd)   De AE-PS (Ryd)
>      4 3 2.5 5F   1( 4.00)       -1.30480       -9.85980        8.55500
>      4 3 3.5 5F   1( 0.00)       -1.23331       -9.75683        8.52352
>      3 2 1.5 6D   1( 1.00)       -1.09034       -8.00574        6.91540
>      3 2 2.5 6D   1( 0.00)       -1.03768       -7.85645        6.81877
>      1 0 0.5 7S   1( 0.00)       -1.12548      -12.93386       11.80838
> 
> I think that results should be exactly the same and I confused.
> Do you have any suggestions?
> 
> 
> Thank you in advance,
> Dmitry Korotin.
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
-- 
Andrea Dal Corso                    Tel. 0039-040-3787428
SISSA, Via Beirut 2/4               Fax. 0039-040-3787528
34014 Trieste (Italy)               e-mail: dalcorso at sissa.it



From giannozz at nest.sns.it  Fri May 26 10:34:49 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Fri, 26 May 2006 10:34:49 +0200
Subject: [Pw_forum] Ecutrho's choice
In-Reply-To: <9d7293e10605251159t2c1bde16hfdc233f6bd8a8b48@mail.gmail.com>
References: <9d7293e10605251159t2c1bde16hfdc233f6bd8a8b48@mail.gmail.com>
Message-ID: <200605261034.49980.giannozz@nest.sns.it>

On Thursday 25 May 2006 20:59, Jess Kondor wrote:

> I found, that properties of my test system [rhombohedral unit cell]
> (total energy, lattice constant, angle between cell vectors, unit-cell
> volume) are almost identical to one obtained by using ecutrho=160 Ry.
> [...] I am a little bit confused, because in ultrasoft case ecutrho should
> be > 4*ecutwfc  .

it may be, but not necessarily so. If it isn't, good for you.
Consider that 40 Ry is a relatively high cutoff for Ultrasoft PP's. 
One tries to perform calculations with ecutwfc = 25 Ry or so.

> My system contains 3 atomic species, one of them is iron. I 
> use Fe.pz-sp-van_ak.UPF pseudopotential. I read the info about 
> these pseudopotentials (unfortunately, for GGA only). I saw that
> recommended value of ecutho should be around 10-12*ecutwfc.

GGA PP's are typically nastier than LDA ones.

Paolo
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From ferretti.andrea at unimore.it  Fri May 26 11:40:01 2006
From: ferretti.andrea at unimore.it (Andrea Ferretti)
Date: Fri, 26 May 2006 11:40:01 +0200 (CEST)
Subject: [Pw_forum] Is it possible to convert the file format between 32
 and 64 system?
In-Reply-To: <4476A2E5.7000508@sissa.it>
References: <4476A2E5.7000508@sissa.it>
Message-ID: <Pine.LNX.4.58.0605261110290.25402@bamboo.fisica.unimo.it>



if you are using espresso 3.1 it should be possible.
basically, the default internal format is now based on the "iotk" library, 
which provides some tools for converting iotk binary formatted files to 
iotk textual files (XML compliant)...

so, in principle (but also in practise), you can convert your 64-bit 
binary files to txt, gzip them, move to a different machine and reconvert 
them to a different binary fmt (this finally ensures a better IO 
performance)...   

during the conversion from bin to txt, you need some disk workspace 
(data txt files are roughly 3 times larger than their related binaries)

the tool to convert files is  
$ESPRESSO_HOME/iotk/src/iotk_copy.x

just type   make iotk_copy.x (or make all)   in $ESPRESSO_HOME/iotk/
to compile it
then run the iotk_copy.x utility and follow the instructions (give the 
name fo the file to be converted, the new file name, the format of the 
output file (textual in your case), -1 (to convert all the file and not 
to skip large data tags) )

hope it helps
andrea



On Fri, 26 May 2006, Xunlei Ding wrote:

> Dear All,
> 
> I do calculations on SP5 and XD1. The wfc and rho files can not be used
> when I change for one to the other,
> because they have 64 and 32 bit file system.
> Are there any tools to convert the files from one format to the other?
> 
> Thank you!
> 
> Yours,
> Ding
> 
> 
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
> 


From dkorotin at gmail.com  Fri May 26 12:06:32 2006
From: dkorotin at gmail.com (Dmitry Korotin)
Date: Fri, 26 May 2006 16:06:32 +0600
Subject: [Pw_forum] ld1.x and test
In-Reply-To: <1148631479.4007.4.camel@dhpc-5-58.sissa.it>
References: <2fd252650605252211k1980075ai7907adff155840eb@mail.gmail.com>
	 <1148631479.4007.4.camel@dhpc-5-58.sissa.it>
Message-ID: <2fd252650605260306p3c15cf9ck1921baba005c86ac@mail.gmail.com>

Thank you

2006/5/26, Andrea Dal Corso <dalcorso at sissa.it>:
> Your test configuration is not the same as in the pseudopotential
> generation. You have 5 electrons in the test and 13 electrons in the
> pseudo generation.
>
> Try with:
>
>         configts(1) = '6s2 6p2 6p4 5f4 6d1 7s0'
>
> Best Regards,
>
> Andrea
>
>
> On Fri, 2006-05-26 at 11:11 +0600, Dmitry Korotin wrote:
> > Dear ESPRESSO users,
> > I'm trying to reproduce Np pseudopotential from the article PRB 66, 235112
> >
> > after pseudo-generation run I have following output:
> >      n l  j  nl             e AE (Ryd)       e PS (Ryd)   De AE-PS (Ryd)
> >      2 1 0.5 6P   1( 2.00)       -3.27104       -3.27129        0.00025
> >      2 1 1.5 6P   1( 4.00)       -2.55754       -2.55786        0.00032
> >      3 2 1.5 6D   1( 1.00)       -1.09034       -1.09045        0.00011
> >      3 2 2.5 6D   1( 0.00)       -1.03768       -1.03782        0.00014
> >      4 3 2.5 5F   1( 4.00)       -1.30480       -1.30299       -0.00181
> >      4 3 3.5 5F   1( 0.00)       -1.23331       -1.23252       -0.00079
> >      1 0 0.5 6S   1( 2.00)       -4.62634       -4.62636        0.00003
> > (De AE-PS (Ryd) seems ok)
> >
> > but after pseudo-test run with the same electrons configuration I have
> > following:
> >      n l  j  nl             e AE (Ryd)       e PS (Ryd)   De AE-PS (Ryd)
> >      4 3 2.5 5F   1( 4.00)       -1.30480       -9.85980        8.55500
> >      4 3 3.5 5F   1( 0.00)       -1.23331       -9.75683        8.52352
> >      3 2 1.5 6D   1( 1.00)       -1.09034       -8.00574        6.91540
> >      3 2 2.5 6D   1( 0.00)       -1.03768       -7.85645        6.81877
> >      1 0 0.5 7S   1( 0.00)       -1.12548      -12.93386       11.80838
> >
> > I think that results should be exactly the same and I confused.
> > Do you have any suggestions?
> >
> >
> > Thank you in advance,
> > Dmitry Korotin.
> > _______________________________________________
> > Pw_forum mailing list
> > Pw_forum at pwscf.org
> > http://www.democritos.it/mailman/listinfo/pw_forum
> --
> Andrea Dal Corso                    Tel. 0039-040-3787428
> SISSA, Via Beirut 2/4               Fax. 0039-040-3787528
> 34014 Trieste (Italy)               e-mail: dalcorso at sissa.it
>
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
>


From proffess at yandex.ru  Fri May 26 16:04:58 2006
From: proffess at yandex.ru (Sergey Lisenkov)
Date: Fri, 26 May 2006 18:04:58 +0400 (MSD)
Subject: [Pw_forum] stability of fftw3 in qe-3.1
In-Reply-To: <200605261000.26619.giannozz@nest.sns.it>
References: <4476361E.000001.26510@tide.yandex.ru> <200605261000.26619.giannozz@nest.sns.it>
Message-ID: <44770B0A.000003.09457@tide.yandex.ru>

>
>on a 32-bit PC in a few cases. They sort of worked, i.e. they produced correct
>results in most cases, but crashed for no apparent good reason in other 
>cases. Anybody who is willing to investigate this issue is welcome.

 Yes, sometimes they worked. But not always.

 Sergey


From ding at sissa.it  Fri May 26 17:43:38 2006
From: ding at sissa.it (Xunlei Ding)
Date: Fri, 26 May 2006 17:43:38 +0200
Subject: [Pw_forum] Re: about electric field potential in pp.x
Message-ID: <4477222A.5010602@sissa.it>

Dear All,

I am now using v3.0. This error is still there.
No matter whether tefield = .true. and dipfield = .true. are set in the pw.x calculation,
/the electric field potential //in output file of pp.x is zero./
How to use /plot_num =12 to plot the e_field?

Yours,
Ding

/On Sat, 2005-03-26 at 13:58 +0800, xywu wrote:
>/ Dear all,
/>/ 
/>/  I add a sawlike electric potential in SCF and want to get the electric 
/>/ potential from pp.x.
/>/ But when i use the pp.x with plot_num=12, the electric field potential 
/>/ in output file is zero.
/
Thank you for noticing this. You are right, there is a bug with
plot_num=12. This will be fixed on the cvs version.

>/ 
/>/ But if i get the electric potential from pp.x without adding electric 
/>/ field in SCF, the error occurs as follows;
/>/ 
/>/  %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
/>/      from punch_plot : error #        -1
/>/      e_field is not calculated
/>/  %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
/>/  
/>/ what's the wrong with it.
/>/ 
/plot_num=12 is supposed to work only if tefield=.true. in the pw.x
calculation, otherwise pp.x will not plot any field and stops with an
error message.


>/ Please give me some advice.
/>/ 
/>/ ! plot_num    selects what is saved in filplot:
/>/   !                0=charge
/>/   !                1=total potential V_bare+V_H + V_xc
/>/   !                2=local ionic potential
/>/   !                3=local density of states at e_fermi
/>/   !                4=local density of electronic entropy
/>/   !                5=STM images
/>/   !                6=spin polarization (rho(up)-rho(down))
/>/   !                7=|psi|^2
/>/   !                8=electron localization function (ELF)
/>/   !                9=planar average of all |psi|^2
/>/   !               10=integrated local density of states (ILDOS) from
/>/   !                  emin to emax (emin, emax in eV)
/>/   !                  if emax is not specified, emax=E_fermi
/>/   !               11=the V_bare + V_H potential
/>/   !               12=the electric field potential
/>/   !               13=the noncolinear magnetization.
/>/ 
/>/ BTW, still another question about add_efield.
/>/ 
/>/ In the file add_efield.f90
/>/ 
/>/ " !--------------------------------------------------------------------------
/>/   subroutine add_efield(rho,vpoten,etotefield)
/>/ !--------------------------------------------------------------------------
/>/ !
/>/ !   This routine adds an electric field to the local potential. The
/>/ !   field is made artificially periodic by introducing a saw-tooth
/>/ !   potential. The field is parallel to a reciprocal lattice vector bg, 
/>/ !   according to the index edir.
/>/ !
/>/ !   if dipfield is false the electric field correction is added to the
/>/ !   potential given as input (the bare local potential) only
/>/ !   at the first call to this routine. In the following calls
/>/ !   the routine exit.
/>/ !
/>/ !   if dipfield is true the dipole moment per unit surface is calculated
/>/ !   and used to cancel the electric field due to periodic boundary
/>/ !   conditions. This potential is added to the Hartree and xc potential
/>/ !   in v_of_rho. NB: in this case the electric field contribution to the 
/>/ !   band energy is subtracted by deband.
/>/ "
/>/ "  if (first) then
/>/      WRITE( stdout,*)
/>/      WRITE( stdout,'(5x,"Adding an external electric field")')
/>/      WRITE( stdout,'(5x,"Intensity [a.u.]: ",f15.8)') eamp
/>/   endif
/>/   if (dipfield) WRITE( stdout,'(5x,"Dipole field [a.u.]: ", f15.8)') dip
/>/   if (first) then
/>/      WRITE( stdout,'(5x,"Potential amplitude [Ry]: ", f15.8)') vamp
/>/      WRITE( stdout,'(5x,"Total length [points]: ", i5)') npoints
/>/      WRITE( stdout,'(5x,"Total length [bohr rad]: ", f15.8)') length
/>/      WRITE( stdout,'(5x,"Field is reversed between points: ",2i6)')nmax, nmax+ndesc
/>/   endif
/>/ "
/>/ 
/>/ And in a typical output file of PW.x with efield:
/>/ "
/>/      Adding an external electric field
/>/      Intensity [a.u.]:      0.01000000
/>/      Potential amplitude [Ry]:      0.24480000
/>/      Total length [points]:   200
/>/      Total length [bohr rad]:     61.20000000
/>/      Field is reversed between points:      1   161
/>/ 
/>/ "
/>/ >From the output file and the code in add_efield.f90, the "first" in the code is TRUE and the "dipfield" is  FALSE, which means that the subroutine is only running only at the first call and the dipole correction in periodic boundary conditions is never used! Is it the truth?
/>/ 
/>/ 
/Yes this is true. However, from your output, the region where the field
is reversed seems too large. This region is supposed to be a narrow
region within the vacuum. 


>/ ------------------------------------------------------------
/>/ Best regards
/>/ 
/>/ Xiyong Wu(X.Y.Wu) 
/>/ E_mail:xywu at imr.ac.cn <mailto:xywu at imr.ac.cn>
/



From giannozz at nest.sns.it  Fri May 26 18:16:12 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Fri, 26 May 2006 18:16:12 +0200
Subject: [Pw_forum] Re: about electric field potential in pp.x
In-Reply-To: <4477222A.5010602@sissa.it>
References: <4477222A.5010602@sissa.it>
Message-ID: <200605261816.12223.giannozz@nest.sns.it>

On Friday 26 May 2006 17:43, Xunlei Ding wrote:

> I am now using v3.0. This error is still there.

> > Thank you for noticing this. You are right, there is a bug with
> > plot_num=12. This will be fixed on the cvs version.

the cvs version has meanwhile become v3.1, not v3.0
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From etabel at hotmail.com  Fri May 26 18:26:23 2006
From: etabel at hotmail.com (Eric Abel)
Date: Fri, 26 May 2006 12:26:23 -0400
Subject: [Pw_forum] imaginary eigenvectors
Message-ID: <BAY105-F15A8702254AC1BBB2F68D2D49E0@phx.gbl>

Hello,

I had a quick phonon question...is there a physical interpretation to a mode 
which is imaginary with a positive frequency?

Eric




From degironc at sissa.it  Fri May 26 21:55:39 2006
From: degironc at sissa.it (Stefano de Gironcoli)
Date: Fri, 26 May 2006 21:55:39 +0200 (CEST)
Subject: [Pw_forum] imaginary eigenvectors
In-Reply-To: <BAY105-F15A8702254AC1BBB2F68D2D49E0@phx.gbl>
References: <BAY105-F15A8702254AC1BBB2F68D2D49E0@phx.gbl>
Message-ID: <Pine.LNX.4.64.0605262151170.3042@shannon-01.sissa.it>

Eigenvectors are defined up to a phase factor which is picked 
somehow randomly by the diagonalization routine.
nothing special in an immaginary eigenvector.

stefano

On Fri, 26 May 2006, Eric Abel wrote:
> Hello,
>
> I had a quick phonon question...is there a physical interpretation to a mode 
> which is imaginary with a positive frequency?
>
> Eric
>
>
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
>


From hwlalves at ufsj.edu.br  Fri May 26 23:34:09 2006
From: hwlalves at ufsj.edu.br (Horacio Wagner Leite Alves)
Date: Fri, 26 May 2006 18:34:09 -0300
Subject: [Pw_forum] imaginary eigenvectors
Message-ID: <20060526213409.D03EE149087@ns.ufsj.edu.br>

Hello,

only to complete Stefano's observation:
to obtain an imaginary phonon wavevector, the q-vector must not be at the
high symmetric points of the Brillouin zone.

good continuation

Horacio W. LEITE ALVES
Departamento de Ci?ncias Naturais
Universidade Federal de S?o Jo?o del Rei
Caixa Postal: 110  CEP: 36.300-000
S?o Jo?o del Rei - MG
Fone: (55)(32)33.79.24.89
Fax: (55)(32)33.79.24.83


--------- Mensagem Original --------
De: pw_forum at pwscf.org
Para: pw_forum at pwscf.org <pw_forum at pwscf.org>
Assunto: [Pw_forum] imaginary eigenvectors
Data: 26/05/06 16:19

>
> Hello,
>
> I had a quick phonon question...is there a physical interpretation to a
mode
> which is imaginary with a positive frequency?
>
> Eric
>
>
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
>
>
>
>
>
>

________________________________________________
Message sent using UebiMiau 2.7.2



From ezadshojaee at hotmail.com  Sat May 27 10:56:01 2006
From: ezadshojaee at hotmail.com (Ezad Shojaee)
Date: Sat, 27 May 2006 08:56:01 +0000
Subject: [Pw_forum] problem with gradient-corrected PP's
Message-ID: <BAY19-F64A898AC70ED96DE4FE8AAA9F0@phx.gbl>

hi
i am trying to use gradient-corrected pseudo potentials for O in TiO2. 
according to the INPUT_PW file , the XC functional is read from PP's file 
.so why do i have this error among an scf calculation :

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
     from readpp : error #         2
     inconsistent DFT read
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

any suggestion will be appreciated

_________________________________________________________________
Don't just search. Find. Check out the new MSN Search! 
http://search.msn.com/



From eyvaz_isaev at yahoo.com  Sat May 27 11:41:02 2006
From: eyvaz_isaev at yahoo.com (Eyvaz Isaev)
Date: Sat, 27 May 2006 02:41:02 -0700 (PDT)
Subject: [Pw_forum] problem with gradient-corrected PP's
In-Reply-To: <BAY19-F64A898AC70ED96DE4FE8AAA9F0@phx.gbl>
Message-ID: <20060527094102.83762.qmail@web60322.mail.yahoo.com>

Hi,
 
> i am trying to use gradient-corrected pseudo
> potentials for O in TiO2. 
> according to the INPUT_PW file , the XC functional
> is read from PP's file 
> .so why do i have this error among an scf
> calculation :
> 
> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
>      from readpp : error #         2
>      inconsistent DFT read
> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
It is very famous error message (please have a look at
the forum archive) and means that you try to use
pseudopotentials with different XC-functionals. 
As example for Ti you use PBE XC-functional, but for
O- LDA based PsP. Use PsPs with the same
XC-functional.

Bests,
Eyvaz.

 

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 


From ding at sissa.it  Sun May 28 10:46:22 2006
From: ding at sissa.it (Xunlei Ding)
Date: Sun, 28 May 2006 10:46:22 +0200
Subject: [Pw_forum] Re: about electric field potential in pp.x
In-Reply-To: <20060526214301.12197.83726.Mailman@democritos.sissa.it>
References: <20060526214301.12197.83726.Mailman@democritos.sissa.it>
Message-ID: <4479635E.1060800@sissa.it>

But, it seems that v3.1 gives the same error, i.e. 
in the out file of pp.x: "     e_field is not calculated" , 
and in the data file the electric field potential data are all zero.

Now I add these lines in the  add_efield.f90 file.
    tefield = .true.
    dipfield = .true.
    edir=3
    eamp=0
    emaxpos=0.5
    eopreg=0.1
Then it can give non-zero data for plot_num=12. 
But I don't know whether it is correct.

Another question is, how to check whether the dipole correction is added correctly for the dipole molecular calculation such as H2O?
I add the above lines to the input file for pw.x calculation for H2O molecule, but I don't how to check whether the correction is correct.

Message: 10
From: Paolo Giannozzi <giannozz at nest.sns.it>
Subject: Re: [Pw_forum] Re: about electric field potential in pp.x


the cvs version has meanwhile become v3.1, not v3.0





From ding at sissa.it  Sun May 28 11:42:06 2006
From: ding at sissa.it (Xunlei Ding)
Date: Sun, 28 May 2006 11:42:06 +0200
Subject: [Pw_forum] [Re] Is it possible to convert the file format between 32
Message-ID: <4479706E.8000803@sissa.it>

Dear andrea,

Thank you for your answer. But I don't know how to use iotk_copy.x.
I did as the instruction, but get some errors below.
Can you help me?

 $ iotk_copy.x
 Name of input file:
test.in
 Name of output file:
test.out
 Textual or Binary:
T
 Maximum data size (-1 for unlimited size)
-1
########################################################################################################################
# UNRECOVERABLE ERROR (ierr=1)
# ERROR IN: iotk_getline (iotk_scan.spp:785)
# CVS Revision: 1.2
iostat=-1
# ERROR IN: iotk_scan_tag (iotk_scan.spp:470)
# CVS Revision: 1.2
# ERROR IN: iotk_open_read (iotk_files.spp:493)
# CVS Revision: 1.2

Best regards!

Yours,
Ding


From giannozz at nest.sns.it  Sun May 28 20:46:58 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Sun, 28 May 2006 20:46:58 +0200
Subject: [Pw_forum] Re: about electric field potential in pp.x
In-Reply-To: <4479635E.1060800@sissa.it>
References: <20060526214301.12197.83726.Mailman@democritos.sissa.it> <4479635E.1060800@sissa.it>
Message-ID: <200605282046.58873.giannozz@nest.sns.it>

On Sunday 28 May 2006 10:46, Xunlei Ding wrote:

> Another question is, how to check whether the dipole correction is added
> correctly for the dipole molecular calculation such as H2O? I add the above
> lines to the input file for pw.x calculation for H2O molecule, but I don't
> how to check whether the correction is correct.

read the paper by Bengtsson (PRB 59, 12301 (1999)) and make the
same test calculation he made
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From gbussi at ethz.ch  Mon May 29 09:33:18 2006
From: gbussi at ethz.ch (Giovanni Bussi)
Date: Mon, 29 May 2006 09:33:18 +0200 (CEST)
Subject: [Pw_forum] [Re] Is it possible to convert the file format between
 32
In-Reply-To: <4479706E.8000803@sissa.it>
References: <4479706E.8000803@sissa.it>
Message-ID: <Pine.LNX.4.61.0605290924430.12660@salorino.usilu.net>

Dear Xunlei,

the error message is a (cryptic) log of what happened
inside the library.

Apparently there is an end of file somewhere (iostat<0),
so that it is possible that:
(1) file test.in does not exist in the current directory
(2) the format of the file is wrong

In the case (2), it is possible that you are using the wrong
version of iotk_copy.x . Remember that to read files produced on the SP5
you need to use the version compiled on that machine.

Ciao!

Giovanni

On Sun, 28 May 2006, Xunlei Ding wrote:

> Dear andrea,
> 
> Thank you for your answer. But I don't know how to use iotk_copy.x.
> I did as the instruction, but get some errors below.
> Can you help me?
> 
> $ iotk_copy.x
> Name of input file:
> test.in
> Name of output file:
> test.out
> Textual or Binary:
> T
> Maximum data size (-1 for unlimited size)
> -1
> ########################################################################################################################
> #  UNRECOVERABLE ERROR (ierr=1)
> #  ERROR IN: iotk_getline (iotk_scan.spp:785)
> #  CVS Revision: 1.2
> iostat=-1
> #  ERROR IN: iotk_scan_tag (iotk_scan.spp:470)
> #  CVS Revision: 1.2
> #  ERROR IN: iotk_open_read (iotk_files.spp:493)
> #  CVS Revision: 1.2
> 
> Best regards!
> 
> Yours,
> Ding
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
> 

-- 
Giovanni Bussi
Computational Science
Department of Chemistry and Applied Biosciences, ETH Zurich
USI-Campus, Via Giuseppe Buffi, 13
CH-6900 Lugano, Switzerland
phone : +41 58 6664809
fax   : +41 58 6664817
email : gbussi at ethz.ch
        giovanni.bussi at gmail.com


From ding at sissa.it  Mon May 29 10:16:00 2006
From: ding at sissa.it (Xunlei Ding)
Date: Mon, 29 May 2006 10:16:00 +0200
Subject: [Pw_forum] [Re] Is it possible to convert the file format between
 32
In-Reply-To: <Pine.LNX.4.61.0605290924430.12660@salorino.usilu.net>
References: <4479706E.8000803@sissa.it> <Pine.LNX.4.61.0605290924430.12660@salorino.usilu.net>
Message-ID: <447AADC0.4000704@sissa.it>

Thank you!
I use iotk_copy.x compiled on sp5, and the test.in file is just a TXT file.
I think there must be something else wrong.
The file is on sp5 at 
/scratch/dingcne0sid5/espresso-3.1/iotk/src/iotk_copy.x
Can anybody help me to test it?

Yours,
Ding

Giovanni Bussi wrote:

>Dear Xunlei,
>
>the error message is a (cryptic) log of what happened
>inside the library.
>
>Apparently there is an end of file somewhere (iostat<0),
>so that it is possible that:
>(1) file test.in does not exist in the current directory
>(2) the format of the file is wrong
>
>In the case (2), it is possible that you are using the wrong
>version of iotk_copy.x . Remember that to read files produced on the SP5
>you need to use the version compiled on that machine.
>
>Ciao!
>
>Giovanni
>
>On Sun, 28 May 2006, Xunlei Ding wrote:
>
>  
>
>>Dear andrea,
>>
>>Thank you for your answer. But I don't know how to use iotk_copy.x.
>>I did as the instruction, but get some errors below.
>>Can you help me?
>>
>>$ iotk_copy.x
>>Name of input file:
>>test.in
>>Name of output file:
>>test.out
>>Textual or Binary:
>>T
>>Maximum data size (-1 for unlimited size)
>>-1
>>########################################################################################################################
>>#  UNRECOVERABLE ERROR (ierr=1)
>>#  ERROR IN: iotk_getline (iotk_scan.spp:785)
>>#  CVS Revision: 1.2
>>iostat=-1
>>#  ERROR IN: iotk_scan_tag (iotk_scan.spp:470)
>>#  CVS Revision: 1.2
>>#  ERROR IN: iotk_open_read (iotk_files.spp:493)
>>#  CVS Revision: 1.2
>>
>>Best regards!
>>
>>Yours,
>>Ding
>>_______________________________________________
>>Pw_forum mailing list
>>Pw_forum at pwscf.org
>>http://www.democritos.it/mailman/listinfo/pw_forum
>>
>>    
>>
>
>  
>


From giannozz at nest.sns.it  Mon May 29 10:51:23 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Mon, 29 May 2006 10:51:23 +0200
Subject: [Pw_forum] [Re] Is it possible to convert the file format between 32
In-Reply-To: <447AADC0.4000704@sissa.it>
References: <4479706E.8000803@sissa.it> <Pine.LNX.4.61.0605290924430.12660@salorino.usilu.net> <447AADC0.4000704@sissa.it>
Message-ID: <200605291051.23645.giannozz@nest.sns.it>

On Monday 29 May 2006 10:16, Xunlei Ding wrote:

> I use iotk_copy.x compiled on sp5, and the test.in file is just a TXT file.

iotk_copy.x translates from/to binary files produced by iotk, not
any file. So you have to translate each binary file separately:
charge-density.xml, gvectors.dat, and all files contained in 
subdirectories K*/ , in the data file directory ("outdir"/"prefix".save).
Note that you will need input option "wf_collect=.true.", or else
the wavefunctions will be stored in a nonportable format
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From takahiro at xyz.mbox.media.kyoto-u.ac.jp  Mon May 29 10:54:06 2006
From: takahiro at xyz.mbox.media.kyoto-u.ac.jp (Takahiro Shimada)
Date: Mon, 29 May 2006 17:54:06 +0900
Subject: [Pw_forum] USPP of Zr (LDA)
Message-ID: <447AB6AE.8020703@xyz.mbox.media.kyoto-u.ac.jp>

Dear users,

I'm looking for a Vanderbilt-type USPP of Zr for LDA.
(valence state: 4s, 4p, 4d, 5s)

It's for PbZrO3.
In the official web page, Vanderbilt-USPPs (LDA) for Pb and O
are supplied, however I could not find that of Zr.

If you have already constructed USPP of Zr,
could you upload or send it by e-mail?
Or, do you have any suggestions?
Thank you.


With best regards,
    Shimada


From Andrea.Marini at roma2.infn.it  Mon May 29 12:32:08 2006
From: Andrea.Marini at roma2.infn.it (Andrea Marini)
Date: Mon, 29 May 2006 12:32:08 +0200 (CEST)
Subject: [Pw_forum] Different Gamma factors for Q-points in the same star
Message-ID: <Pine.LNX.4.64.0605291229230.23347@anomaly.roma2.infn.it>

Dear users,

 	I am a newbie in using pw/ph to get electron-phonon coupling 
matrix elements (ME).

Now I have the following problem. To understand the ME symmetry properties 
I have done two calculations of the Eliashberg function for Al using a 
Monkhorst-Pack set of Q-points in the whole BZ and only in the irreducible 
wedge and I have found tiny (but not negligible) differences in the two 
calculations.

I have traced back the problem to different Gamma factors (calculated  in 
elphonon.f90) for two Q-points in the same star. This confuses me as I was 
expecting possible differences only  for degenerate modes.

Am I wrong ? Am I doing sone stupid mistake ?

THANKS!

Andrea

  **************************************************************************

      q = (    0.250000000   0.750000000  -0.250000000 )

      omega( 1) =       2.619122 [THz] =      87.365086 [cm-1]
      omega( 2) =       3.738457 [THz] =     124.702317 [cm-1]
      omega( 3) =       4.666853 [THz] =     155.670490 [cm-1]

      Gaussian Broadening:   0.100 Ry, ngauss=   1
      DOS =  4.376674 states/spin/Ry/Unit Cell at Ef=  8.063192 eV
      double delta at Ef = 33.847218
      lambda( 1)=  1.2622   gamma=   36.19 GHz
      lambda( 2)=  0.9349   gamma=   54.61 GHz
      lambda( 3)=  0.3583   gamma=   32.61 GHz

  **************************************************************************

      q = (    0.250000000  -0.250000000   0.750000000 )

      omega( 1) =       2.619570 [THz] =      87.380025 [cm-1]
      omega( 2) =       3.738907 [THz] =     124.717352 [cm-1]
      omega( 3) =       4.666851 [THz] =     155.670416 [cm-1]

      Gaussian Broadening:   0.100 Ry, ngauss=   1
      DOS =  4.376674 states/spin/Ry/Unit Cell at Ef=  8.063192 eV
      double delta at Ef = 33.847218
      lambda( 1)=  1.7620   gamma=   50.53 GHz
      lambda( 2)=  0.9916   gamma=   57.94 GHz
      lambda( 3)=  0.1641   gamma=   14.93 GHz

To obtain these results I have used

pw.x< Al.scf.in
pw.x< Al-1.nscf.in
ph.x< Al-1.ph_elph.in

for q =  0.250000000   0.750000000  -0.250000000

and

pw.x< Al.scf.in
pw.x< Al-2.nscf.in
ph.x< Al-2.ph_elph.in

for q =  0.250000000  -0.250000000   0.750000000

  = Al.scf.in =============================================================
  &control
     calculation = 'scf',prefix='Al',
     pseudo_dir = '/home/marini/Sources/espresso/espresso-3.0/pseudo/',
     wf_collect=.true.,
  /
  &system
     ibrav=  2, celldm(1) =7.50, nat=  1, ntyp= 1,ecutwfc = 16, nbnd = 5,
     occupations='smearing', smearing='mp', degauss=0.022
  /
  &electrons
     diagonalization='david'
     conv_thr = 1.0d-10
  /
ATOMIC_SPECIES
  Al  26.98  Al.vbc.UPF
ATOMIC_POSITIONS
  Al 0.00 0.00 0.00
K_POINTS {automatic}
   4 4 4 0 0 0

  = Al-1.nscf.in ==========================================================
  &control
     calculation='phonon',prefix  = 'Al'
     pseudo_dir = '/home/marini/Sources/espresso/espresso-3.0/pseudo/',
  /
  &system
     ibrav=  2, celldm(1) =7.50, nat=  1, ntyp= 1,ecutwfc = 16, nbnd = 5,
     occupations='smearing', smearing='mp', degauss=0.022
  /
  &electrons
   diago_thr_init =  1.0d-6
  /
  &phonon
     xqq(1)= 0.25  , xqq(2)= 0.75 , xqq(3)= -0.25
  /
ATOMIC_SPECIES
  Al  26.98  Al.vbc.UPF
ATOMIC_POSITIONS
  Al 0.00 0.00 0.00
K_POINTS {automatic}
   4 4 4 0 0 0

  = Al-1.ph_elph.in =======================================================

phonons of Al
  &inputph
   tr2_ph=1.0d-14,
   prefix='Al',
   amass(1)=26.98,
   fildyn='Al.dyn',
   fildvscf='Al.dvscf',
   filelph='Al-1.elph',
   iverbosity = 1,
   elph=.true.,
   trans=.true.,
  /
  0.25  0.75 -0.25

  = Al-2.nscf.in ==========================================================

  &control
     calculation='phonon',prefix  = 'Al'
     pseudo_dir = '/home/marini/Sources/espresso/espresso-3.0/pseudo/',
  /
  &system
     ibrav=  2, celldm(1) =7.50, nat=  1, ntyp= 1,ecutwfc = 16, nbnd = 5,
     occupations='smearing', smearing='mp', degauss=0.022
  /
  &electrons
   diago_thr_init =  1.0d-6
  /
  &phonon
     xqq(1)= 0.25  , xqq(2)= -0.25 , xqq(3)= 0.75
  /
ATOMIC_SPECIES
  Al  26.98  Al.vbc.UPF
ATOMIC_POSITIONS
  Al 0.00 0.00 0.00
K_POINTS {automatic}
   4 4 4 0 0 0

  = Al-2.ph_elph.in =======================================================

phonons of Al
  &inputph
   tr2_ph=1.0d-14,
   prefix='Al',
   amass(1)=26.98,
   fildyn='Al.dyn',
   fildvscf='Al.dvscf',
   filelph='Al-2.elph',
   iverbosity = 1,
   elph=.true.,
   trans=.true.,
  /
  0.25  -0.25 0.75


--------------------------------------------------------------------------
                                Andrea MARINI

        Physics Department, University of Rome "Tor Vergata" (Italy)
               - phone: +39-0672594894 - fax: +39-062023507 -

-- andrea.marini at roma2.infn.it - http://www.fisica.uniroma2.it/~marini/ --


From eyvaz_isaev at yahoo.com  Mon May 29 12:59:58 2006
From: eyvaz_isaev at yahoo.com (Eyvaz Isaev)
Date: Mon, 29 May 2006 03:59:58 -0700 (PDT)
Subject: [Pw_forum] Different Gamma factors for Q-points in the same star
In-Reply-To: <Pine.LNX.4.64.0605291229230.23347@anomaly.roma2.infn.it>
Message-ID: <20060529105958.60693.qmail@web60318.mail.yahoo.com>

Hi Andrea,

At least your k-points set looks very-very small for
electron-phonon coupling constants calculations. Try
increase number of k-points.
It is worth increasing ecutwfc, too (say, up to 20-25
Ry).

Bests,
Eyvaz.

--- Andrea Marini <Andrea.Marini at roma2.infn.it> wrote:

> Dear users,
> 
>  	I am a newbie in using pw/ph to get
> electron-phonon coupling 
> matrix elements (ME).
> 
> Now I have the following problem. To understand the
> ME symmetry properties 
> I have done two calculations of the Eliashberg
> function for Al using a 
> Monkhorst-Pack set of Q-points in the whole BZ and
> only in the irreducible 
> wedge and I have found tiny (but not negligible)
> differences in the two 
> calculations.
> 
> I have traced back the problem to different Gamma
> factors (calculated  in 
> elphonon.f90) for two Q-points in the same star.
> This confuses me as I was 
> expecting possible differences only  for degenerate
> modes.
> 
> Am I wrong ? Am I doing sone stupid mistake ?
> 
> THANKS!
> 
> Andrea
> 
>  
>
**************************************************************************
> 
>       q = (    0.250000000   0.750000000 
> -0.250000000 )
> 
>       omega( 1) =       2.619122 [THz] =     
> 87.365086 [cm-1]
>       omega( 2) =       3.738457 [THz] =    
> 124.702317 [cm-1]
>       omega( 3) =       4.666853 [THz] =    
> 155.670490 [cm-1]
> 
>       Gaussian Broadening:   0.100 Ry, ngauss=   1
>       DOS =  4.376674 states/spin/Ry/Unit Cell at
> Ef=  8.063192 eV
>       double delta at Ef = 33.847218
>       lambda( 1)=  1.2622   gamma=   36.19 GHz
>       lambda( 2)=  0.9349   gamma=   54.61 GHz
>       lambda( 3)=  0.3583   gamma=   32.61 GHz
> 
>  
>
**************************************************************************
> 
>       q = (    0.250000000  -0.250000000  
> 0.750000000 )
> 
>       omega( 1) =       2.619570 [THz] =     
> 87.380025 [cm-1]
>       omega( 2) =       3.738907 [THz] =    
> 124.717352 [cm-1]
>       omega( 3) =       4.666851 [THz] =    
> 155.670416 [cm-1]
> 
>       Gaussian Broadening:   0.100 Ry, ngauss=   1
>       DOS =  4.376674 states/spin/Ry/Unit Cell at
> Ef=  8.063192 eV
>       double delta at Ef = 33.847218
>       lambda( 1)=  1.7620   gamma=   50.53 GHz
>       lambda( 2)=  0.9916   gamma=   57.94 GHz
>       lambda( 3)=  0.1641   gamma=   14.93 GHz
> 
> To obtain these results I have used
> 
> pw.x< Al.scf.in
> pw.x< Al-1.nscf.in
> ph.x< Al-1.ph_elph.in
> 
> for q =  0.250000000   0.750000000  -0.250000000
> 
> and
> 
> pw.x< Al.scf.in
> pw.x< Al-2.nscf.in
> ph.x< Al-2.ph_elph.in
> 
> for q =  0.250000000  -0.250000000   0.750000000
> 
>   = Al.scf.in
>
=============================================================
>   &control
>      calculation = 'scf',prefix='Al',
>      pseudo_dir =
>
'/home/marini/Sources/espresso/espresso-3.0/pseudo/',
>      wf_collect=.true.,
>   /
>   &system
>      ibrav=  2, celldm(1) =7.50, nat=  1, ntyp=
> 1,ecutwfc = 16, nbnd = 5,
>      occupations='smearing', smearing='mp',
> degauss=0.022
>   /
>   &electrons
>      diagonalization='david'
>      conv_thr = 1.0d-10
>   /
> ATOMIC_SPECIES
>   Al  26.98  Al.vbc.UPF
> ATOMIC_POSITIONS
>   Al 0.00 0.00 0.00
> K_POINTS {automatic}
>    4 4 4 0 0 0
> 
>   = Al-1.nscf.in
>
==========================================================
>   &control
>      calculation='phonon',prefix  = 'Al'
>      pseudo_dir =
>
'/home/marini/Sources/espresso/espresso-3.0/pseudo/',
>   /
>   &system
>      ibrav=  2, celldm(1) =7.50, nat=  1, ntyp=
> 1,ecutwfc = 16, nbnd = 5,
>      occupations='smearing', smearing='mp',
> degauss=0.022
>   /
>   &electrons
>    diago_thr_init =  1.0d-6
>   /
>   &phonon
>      xqq(1)= 0.25  , xqq(2)= 0.75 , xqq(3)= -0.25
>   /
> ATOMIC_SPECIES
>   Al  26.98  Al.vbc.UPF
> ATOMIC_POSITIONS
>   Al 0.00 0.00 0.00
> K_POINTS {automatic}
>    4 4 4 0 0 0
> 
>   = Al-1.ph_elph.in
>
=======================================================
> 
> phonons of Al
>   &inputph
>    tr2_ph=1.0d-14,
>    prefix='Al',
>    amass(1)=26.98,
>    fildyn='Al.dyn',
>    fildvscf='Al.dvscf',
>    filelph='Al-1.elph',
>    iverbosity = 1,
>    elph=.true.,
>    trans=.true.,
>   /
>   0.25  0.75 -0.25
> 
>   = Al-2.nscf.in
>
==========================================================
> 
>   &control
>      calculation='phonon',prefix  = 'Al'
>      pseudo_dir =
>
'/home/marini/Sources/espresso/espresso-3.0/pseudo/',
>   /
>   &system
>      ibrav=  2, celldm(1) =7.50, nat=  1, ntyp=
> 1,ecutwfc = 16, nbnd = 5,
>      occupations='smearing', smearing='mp',
> degauss=0.022
>   /
>   &electrons
>    diago_thr_init =  1.0d-6
>   /
>   &phonon
>      xqq(1)= 0.25  , xqq(2)= -0.25 , xqq(3)= 0.75
>   /
> ATOMIC_SPECIES
>   Al  26.98  Al.vbc.UPF
> ATOMIC_POSITIONS
>   Al 0.00 0.00 0.00
> K_POINTS {automatic}
>    4 4 4 0 0 0
> 
>   = Al-2.ph_elph.in
>
=======================================================
> 
> phonons of Al
>   &inputph
>    tr2_ph=1.0d-14,
>    prefix='Al',
>    amass(1)=26.98,
>    fildyn='Al.dyn',
>    fildvscf='Al.dvscf',
>    filelph='Al-2.elph',
>    iverbosity = 1,
>    elph=.true.,
>    trans=.true.,
>   /
>   0.25  -0.25 0.75
> 
> 
>
--------------------------------------------------------------------------
> 
=== message truncated ===


__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 


From Andrea.Marini at roma2.infn.it  Mon May 29 14:06:47 2006
From: Andrea.Marini at roma2.infn.it (Andrea Marini)
Date: Mon, 29 May 2006 14:06:47 +0200 (CEST)
Subject: [Pw_forum] Re: Different Gamma factors for Q-points in the same star
In-Reply-To: <Pine.LNX.4.64.0605291229230.23347@anomaly.roma2.infn.it>
References: <Pine.LNX.4.64.0605291229230.23347@anomaly.roma2.infn.it>
Message-ID: <Pine.LNX.4.64.0605291405240.29266@anomaly.roma2.infn.it>

> Hi Andrea,
>
> At least your k-points set looks very-very small for
> electron-phonon coupling constants calculations. Try
> increase number of k-points.
> It is worth increasing ecutwfc, too (say, up to 20-25
> Ry).
>
> Bests,
> Eyvaz.

Dear Eyvaz,

 	the results do  not change if I increase the cutoff to 32 Ry and 
use a 6 6 6 MP grid.


Andrea

--------------------------------------------------------------------------
                                Andrea MARINI

        Physics Department, University of Rome "Tor Vergata" (Italy)
               - phone: +39-0672594894 - fax: +39-062023507 -

-- andrea.marini at roma2.infn.it - http://www.fisica.uniroma2.it/~marini/ --


From eyvaz_isaev at yahoo.com  Mon May 29 14:31:19 2006
From: eyvaz_isaev at yahoo.com (Eyvaz Isaev)
Date: Mon, 29 May 2006 05:31:19 -0700 (PDT)
Subject: [Pw_forum] Re: Different Gamma factors for Q-points in the same star
In-Reply-To: <Pine.LNX.4.64.0605291405240.29266@anomaly.roma2.infn.it>
Message-ID: <20060529123119.95930.qmail@web60320.mail.yahoo.com>

> the results do  not change if I increase the
> cutoff to 32 Ry and 
> use a 6 6 6 MP grid.
>
Actually 25 Ry should be enough. I think there is not
too much difference between 4x4x4 and 6x6x6 k-grid (8,
and 16 k-points including gamma-point). How about
12x12x12 or 16x16x16?

I would try to look at you input files again.

Bests,
Eyvaz.
 
> Andrea
> 
>
--------------------------------------------------------------------------
>                                 Andrea MARINI
> 
>         Physics Department, University of Rome "Tor
> Vergata" (Italy)
>                - phone: +39-0672594894 - fax:
> +39-062023507 -
> 
> -- andrea.marini at roma2.infn.it -
> http://www.fisica.uniroma2.it/~marini/ --
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
> 


__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 


From Andrea.Marini at roma2.infn.it  Mon May 29 14:40:01 2006
From: Andrea.Marini at roma2.infn.it (Andrea Marini)
Date: Mon, 29 May 2006 14:40:01 +0200 (CEST)
Subject: [Pw_forum] Re: Different Gamma factors for Q-points in the same
 star
In-Reply-To: <20060529123119.95930.qmail@web60320.mail.yahoo.com>
References: <20060529123119.95930.qmail@web60320.mail.yahoo.com>
Message-ID: <Pine.LNX.4.64.0605291438470.29717@anomaly.roma2.infn.it>

On Mon, 29 May 2006, Eyvaz Isaev wrote:

>> the results do  not change if I increase the
>> cutoff to 32 Ry and
>> use a 6 6 6 MP grid.
>>
> Actually 25 Ry should be enough. I think there is not
> too much difference between 4x4x4 and 6x6x6 k-grid (8,
> and 16 k-points including gamma-point). How about
> 12x12x12 or 16x16x16?

No better results. These are the .elph files using a 12 12 12 grid.

      0.25000000     0.75000000    -0.25000000      10       3
   0.271070E-05  0.427079E-05  0.734232E-05
      Gaussian Broadening:   0.100 Ry, ngauss=   1
      DOS =  2.626188 states/spin/Ry/Unit Cell at Ef=  8.283160 eV
      lambda( 1)=  0.1696   gamma=   12.48 GHz
      lambda( 2)=  0.1114   gamma=   12.91 GHz
      lambda( 3)=  0.1656   gamma=   33.01 GHz


      0.25000000    -0.25000000     0.75000000      10       3
   0.271070E-05  0.427086E-05  0.734185E-05
      Gaussian Broadening:   0.100 Ry, ngauss=   1
      DOS =  2.626188 states/spin/Ry/Unit Cell at Ef=  8.283160 eV
      lambda( 1)=  0.0859   gamma=    6.32 GHz
      lambda( 2)=  0.2025   gamma=   23.47 GHz
      lambda( 3)=  0.1435   gamma=   28.60 GHz

Thanks for your help

Andrea

--------------------------------------------------------------------------
                                Andrea MARINI

        Physics Department, University of Rome "Tor Vergata" (Italy)
               - phone: +39-0672594894 - fax: +39-062023507 -

-- andrea.marini at roma2.infn.it - http://www.fisica.uniroma2.it/~marini/ --


From baroni at sissa.it  Mon May 29 22:10:15 2006
From: baroni at sissa.it (Stefano Baroni)
Date: Mon, 29 May 2006 15:10:15 -0500
Subject: [Pw_forum] Re: Different Gamma factors for Q-points in the same star
In-Reply-To: <Pine.LNX.4.64.0605291438470.29717@anomaly.roma2.infn.it>
References: <20060529123119.95930.qmail@web60320.mail.yahoo.com> <Pine.LNX.4.64.0605291438470.29717@anomaly.roma2.infn.it>
Message-ID: <69DAE2A0-2AEB-4AD2-9FF6-FC00590D0DD7@sissa.it>

Andrea, Eyvaz: I know nothing about the way the Eliashberg function  
is calculated in Quantum ESPRESSO, but symmetry issues have (should  
have) little (if anything) to do with k-point grids, cutoff, and the  
like. What I mean is that if symmetrization is well implemented (and  
well understood by those who use the code) an inadequate cutoff and/ 
or k-point grid may well give inaccurate results, but these results  
should be the same (to _very_ high accuracy) whether calculated by  
restricting the BZ sampling to the irreducible wedge (and properly  
symmetrizing the result), or by the extending the sum to the full BZ.  
Sorry for not being more helpful than this - Stefano

On May 29, 2006, at 7:40 AM, Andrea Marini wrote:

> On Mon, 29 May 2006, Eyvaz Isaev wrote:
>
>>> the results do  not change if I increase the
>>> cutoff to 32 Ry and
>>> use a 6 6 6 MP grid.
>>>
>> Actually 25 Ry should be enough. I think there is not
>> too much difference between 4x4x4 and 6x6x6 k-grid (8,
>> and 16 k-points including gamma-point). How about
>> 12x12x12 or 16x16x16?
>
> No better results. These are the .elph files using a 12 12 12 grid.
>
>      0.25000000     0.75000000    -0.25000000      10       3
>   0.271070E-05  0.427079E-05  0.734232E-05
>      Gaussian Broadening:   0.100 Ry, ngauss=   1
>      DOS =  2.626188 states/spin/Ry/Unit Cell at Ef=  8.283160 eV
>      lambda( 1)=  0.1696   gamma=   12.48 GHz
>      lambda( 2)=  0.1114   gamma=   12.91 GHz
>      lambda( 3)=  0.1656   gamma=   33.01 GHz
>
>
>      0.25000000    -0.25000000     0.75000000      10       3
>   0.271070E-05  0.427086E-05  0.734185E-05
>      Gaussian Broadening:   0.100 Ry, ngauss=   1
>      DOS =  2.626188 states/spin/Ry/Unit Cell at Ef=  8.283160 eV
>      lambda( 1)=  0.0859   gamma=    6.32 GHz
>      lambda( 2)=  0.2025   gamma=   23.47 GHz
>      lambda( 3)=  0.1435   gamma=   28.60 GHz
>
> Thanks for your help
>
> Andrea
>
> ---------------------------------------------------------------------- 
> ----
>                                Andrea MARINI
>
>        Physics Department, University of Rome "Tor Vergata" (Italy)
>               - phone: +39-0672594894 - fax: +39-062023507 -
>
> -- andrea.marini at roma2.infn.it - http://www.fisica.uniroma2.it/ 
> ~marini/ --
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum

---
Stefano Baroni - SISSA  &  DEMOCRITOS National Simulation Center -  
Trieste
[+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype)

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: /pipermail/attachments/20060529/4dbbb321/attachment.htm 

From eyvaz_isaev at yahoo.com  Tue May 30 01:21:34 2006
From: eyvaz_isaev at yahoo.com (Eyvaz Isaev)
Date: Mon, 29 May 2006 16:21:34 -0700 (PDT)
Subject: [Pw_forum] Re: Different Gamma factors for Q-points in the same star
In-Reply-To: <69DAE2A0-2AEB-4AD2-9FF6-FC00590D0DD7@sissa.it>
Message-ID: <20060529232134.93641.qmail@web60322.mail.yahoo.com>

Dear Stefano, 

I am quite agree with your comment, but I suggested
(and hopefully it is) that the symmetrization is done
correctly in the code.  As you know the gamma
approximately is a sum of production of two delta
functions, connecting two parts of the Fermi surface,
and is believed to converge slowly wrt number of
k-sampling. That is why I suggested the use of  more
k-points.

Bests,
Eyvaz.

--- Stefano Baroni <baroni at sissa.it> wrote:

> Andrea, Eyvaz: I know nothing about the way the
> Eliashberg function  
> is calculated in Quantum ESPRESSO, but symmetry
> issues have (should  have) little (if anything) to
do > with k-point grids,  cutoff, and the  like. What
I 
> mean is that if  symmetrization is well implemented 
> (and   well understood by those who use the code) an
> inadequate cutoff and/  or k-point grid may well
give > inaccurate results,  but these results  
> should be the same (to _very_ high accuracy) whether
> calculated by restricting the BZ sampling to the 
> irreducible wedge (and properly  
> symmetrizing the result), or by the extending the
> sum to the full BZ.  
> Sorry for not being more helpful than this - Stefano
 

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 


From ding at sissa.it  Tue May 30 10:38:16 2006
From: ding at sissa.it (Xunlei Ding)
Date: Tue, 30 May 2006 10:38:16 +0200
Subject: [Pw_forum] Re: about electric field potential in pp.x
In-Reply-To: <200605282046.58873.giannozz@nest.sns.it>
References: <20060526214301.12197.83726.Mailman@democritos.sissa.it> <4479635E.1060800@sissa.it> <200605282046.58873.giannozz@nest.sns.it>
Message-ID: <1148978296.447c04781e304@webmail.sissa.it>

Dear Paolo,

Thank you very much!
I have tested the same system on that paper.
With dipole correction, the energy and geometry are more close to those by
doulbe H2O layers whose total dipole is zero. And the energy difference between
with and without dipole correction is the same as that in the paper. So I think
my calculation is correct now.

The only thing now, is how to draw the electrostatic potential in pwscf?
I have tested v3.0 and v3.1. 

Your sincerely,
Ding
 
Quoting Paolo Giannozzi <giannozz at nest.sns.it>:

> On Sunday 28 May 2006 10:46, Xunlei Ding wrote:
> 
> > Another question is, how to check whether the dipole correction is added
> > correctly for the dipole molecular calculation such as H2O? I add the
> above
> > lines to the input file for pw.x calculation for H2O molecule, but I don't
> > how to check whether the correction is correct.
> 
> read the paper by Bengtsson (PRB 59, 12301 (1999)) and make the
> same test calculation he made
> -- 
> Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
> Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
> Piazza dei Cavalieri 7      I-56126 Pisa, Italy
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
> 




----------------------------------------------------------------
  SISSA Webmail https://webmail.sissa.it/
  Powered by Horde http://www.horde.org/



From coldwind1978 at yahoo.com.cn  Tue May 30 11:09:37 2006
From: coldwind1978 at yahoo.com.cn (=?gb2312?q?=D5=EB=D2=B6=20=D6=EC?=)
Date: Tue, 30 May 2006 17:09:37 +0800 (CST)
Subject: [Pw_forum] what' the meaning of *.axsf file by dynmat.x
Message-ID: <20060530090937.94972.qmail@web15806.mail.cnb.yahoo.com>

Dear all users:
    i have calculate LO-TO splitting of PbTiO3 by dynmat.x. the result is good. but i donnot know the meaning of pt.axsf file. i have known there is the eignvector of modes in  pt.molden. the eignvector is as follows:
  vibration     1
   0.07571  -0.03804   0.00000
   0.17546  -0.08817   0.00000
  -0.53433   0.26849   0.00000
  -0.53433   0.21953   0.00000
  -0.43689   0.26849   0.00000
 vibration     2
  -0.03804  -0.07571   0.00000
  -0.08817  -0.17546   0.00000
   0.26849   0.53433   0.00000
   0.26849   0.43689   0.00000
   0.21953   0.53433   0.00000
 vibration     3
   0.40346   0.19188  -0.02001
   0.40346   0.19188  -0.02001
   0.40346   0.19188  -0.02001
   0.40346   0.19188  -0.02001
   0.40346   0.19188  -0.02001
 vibration     4
   0.19217  -0.40381   0.00248
   0.19217  -0.40381   0.00248
   0.19217  -0.40381   0.00248
   0.19217  -0.40381   0.00248
   0.19217  -0.40381   0.00248
 vibration     5
  -0.01700  -0.01083  -0.44676
  -0.01700  -0.01083  -0.44676
  -0.01700  -0.01083  -0.44676
  -0.01700  -0.01083  -0.44676
  -0.01700  -0.01083  -0.44676
 vibration     6
   0.00000   0.00000  -0.24779
   0.00000   0.00000   0.67105
   0.00000   0.00000   0.32923
   0.00000   0.00000   0.43583
   0.00000   0.00000   0.43583
 vibration     7
   0.21986   0.12155   0.00000
  -0.81192  -0.44887   0.00000
  -0.14516  -0.08025   0.00000
  -0.14516  -0.07039   0.00000
  -0.12732  -0.08025   0.00000
 vibration     8
  -0.12155   0.21986   0.00000
   0.44887  -0.81192   0.00000
   0.08025  -0.14516   0.00000
   0.08025  -0.12732   0.00000
   0.07039  -0.14516   0.00000
 vibration     9
   0.00000   0.00000   0.00000
   0.00000   0.00000   0.00000
   0.01732   0.30779   0.00000
  -0.01732   0.00000  -0.63637
   0.00000  -0.30779   0.63637
 vibration    10
   0.00000   0.00000   0.00000
   0.00000   0.00000   0.00000
  -0.70667  -0.00834   0.00000
   0.70667   0.00000  -0.02327
   0.00000   0.00834   0.02327
 vibration    11
   0.00000   0.00000   0.00000
   0.00000   0.00000   0.00000
   0.01764  -0.63655   0.00000
  -0.01764   0.00000  -0.30740
   0.00000   0.63655   0.30740
 vibration    12
   0.00000   0.00000  -0.01510
   0.00000   0.00000  -0.24344
   0.00000   0.00000  -0.35320
   0.00000   0.00000   0.63865
   0.00000   0.00000   0.63865
 vibration    13
   0.00064   0.00770   0.00000
   0.00088   0.01059   0.00000
   0.03016   0.36183   0.00000
   0.03016  -0.85504   0.00000
  -0.07127   0.36183   0.00000
 vibration    14
  -0.00770   0.00064   0.00000
  -0.01059   0.00088   0.00000
  -0.36183   0.03016   0.00000
  -0.36183  -0.07127   0.00000
   0.85504   0.03016   0.00000
 vibration    15
   0.00000   0.00000  -0.03183
   0.00000   0.00000  -0.25598
   0.00000   0.00000   0.95294
   0.00000   0.00000   0.11264
   0.00000   0.00000   0.11264
  however pt.axsf  include:
  PRIMCOORD  1
      5   1
   Pb     0.00000   0.00000   0.00000   0.00757  -0.00380   0.00000
   Ti     1.94208   1.94208   1.94208   0.01755  -0.00882   0.00000
   O      1.94208   1.94208   0.00000  -0.05343   0.02685   0.00000
   O      1.94208   0.00000   1.94208  -0.05343   0.02195   0.00000
   O      0.00000   1.94208   1.94208  -0.04369   0.02685   0.00000
PRIMCOORD  2
      5   1
   Pb     0.00000   0.00000   0.00000  -0.00380  -0.00757   0.00000
   Ti     1.94208   1.94208   1.94208  -0.00882  -0.01755   0.00000
   O      1.94208   1.94208   0.00000   0.02685   0.05343   0.00000
   O      1.94208   0.00000   1.94208   0.02685   0.04369   0.00000
   O      0.00000   1.94208   1.94208   0.02195   0.05343   0.00000
PRIMCOORD  3
      5   1
   Pb     0.00000   0.00000   0.00000   0.04035   0.01919  -0.00200
   Ti     1.94208   1.94208   1.94208   0.04035   0.01919  -0.00200
   O      1.94208   1.94208   0.00000   0.04035   0.01919  -0.00200
   O      1.94208   0.00000   1.94208   0.04035   0.01919  -0.00200
   O      0.00000   1.94208   1.94208   0.04035   0.01919  -0.00200
PRIMCOORD  4
      5   1
   Pb     0.00000   0.00000   0.00000   0.01922  -0.04038   0.00025
   Ti     1.94208   1.94208   1.94208   0.01922  -0.04038   0.00025
   O      1.94208   1.94208   0.00000   0.01922  -0.04038   0.00025
   O      1.94208   0.00000   1.94208   0.01922  -0.04038   0.00025
   O      0.00000   1.94208   1.94208   0.01922  -0.04038   0.00025
PRIMCOORD  5
      5   1
   Pb     0.00000   0.00000   0.00000  -0.00170  -0.00108  -0.04468
   Ti     1.94208   1.94208   1.94208  -0.00170  -0.00108  -0.04468
   O      1.94208   1.94208   0.00000  -0.00170  -0.00108  -0.04468
   O      1.94208   0.00000   1.94208  -0.00170  -0.00108  -0.04468
   O      0.00000   1.94208   1.94208  -0.00170  -0.00108  -0.04468
PRIMCOORD  6
      5   1
   Pb     0.00000   0.00000   0.00000   0.00000   0.00000  -0.02478
   Ti     1.94208   1.94208   1.94208   0.00000   0.00000   0.06711
   O      1.94208   1.94208   0.00000   0.00000   0.00000   0.03292
   O      1.94208   0.00000   1.94208   0.00000   0.00000   0.04358
   O      0.00000   1.94208   1.94208   0.00000   0.00000   0.04358
PRIMCOORD  7
      5   1
   Pb     0.00000   0.00000   0.00000   0.02199   0.01215   0.00000
   Ti     1.94208   1.94208   1.94208  -0.08119  -0.04489   0.00000
   O      1.94208   1.94208   0.00000  -0.01452  -0.00803   0.00000
   O      1.94208   0.00000   1.94208  -0.01452  -0.00704   0.00000
   O      0.00000   1.94208   1.94208  -0.01273  -0.00803   0.00000
PRIMCOORD  8
      5   1
   Pb     0.00000   0.00000   0.00000  -0.01215   0.02199   0.00000
   Ti     1.94208   1.94208   1.94208   0.04489  -0.08119   0.00000
   O      1.94208   1.94208   0.00000   0.00803  -0.01452   0.00000
   O      1.94208   0.00000   1.94208   0.00803  -0.01273   0.00000
   O      0.00000   1.94208   1.94208   0.00704  -0.01452   0.00000
PRIMCOORD  9
      5   1
   Pb     0.00000   0.00000   0.00000   0.00000   0.00000   0.00000
   Ti     1.94208   1.94208   1.94208   0.00000   0.00000   0.00000
   O      1.94208   1.94208   0.00000   0.00173   0.03078   0.00000
   O      1.94208   0.00000   1.94208  -0.00173   0.00000  -0.06364
   O      0.00000   1.94208   1.94208   0.00000  -0.03078   0.06364
PRIMCOORD 10
      5   1
   Pb     0.00000   0.00000   0.00000   0.00000   0.00000   0.00000
   Ti     1.94208   1.94208   1.94208   0.00000   0.00000   0.00000
   O      1.94208   1.94208   0.00000  -0.07067  -0.00083   0.00000
   O      1.94208   0.00000   1.94208   0.07067   0.00000  -0.00233
   O      0.00000   1.94208   1.94208   0.00000   0.00083   0.00233
PRIMCOORD 11
      5   1
   Pb     0.00000   0.00000   0.00000   0.00000   0.00000   0.00000
   Ti     1.94208   1.94208   1.94208   0.00000   0.00000   0.00000
   O      1.94208   1.94208   0.00000   0.00176  -0.06366   0.00000
   O      1.94208   0.00000   1.94208  -0.00176   0.00000  -0.03074
   O      0.00000   1.94208   1.94208   0.00000   0.06366   0.03074
PRIMCOORD 12
      5   1
   Pb     0.00000   0.00000   0.00000   0.00000   0.00000  -0.00151
   Ti     1.94208   1.94208   1.94208   0.00000   0.00000  -0.02434
   O      1.94208   1.94208   0.00000   0.00000   0.00000  -0.03532
   O      1.94208   0.00000   1.94208   0.00000   0.00000   0.06387
   O      0.00000   1.94208   1.94208   0.00000   0.00000   0.06387
PRIMCOORD 13
      5   1
   Pb     0.00000   0.00000   0.00000   0.00006   0.00077   0.00000
   Ti     1.94208   1.94208   1.94208   0.00009   0.00106   0.00000
   O      1.94208   1.94208   0.00000   0.00302   0.03618   0.00000
   O      1.94208   0.00000   1.94208   0.00302  -0.08550   0.00000
   O      0.00000   1.94208   1.94208  -0.00713   0.03618   0.00000
PRIMCOORD 14
      5   1
   Pb     0.00000   0.00000   0.00000  -0.00077   0.00006   0.00000
   Ti     1.94208   1.94208   1.94208  -0.00106   0.00009   0.00000
   O      1.94208   1.94208   0.00000  -0.03618   0.00302   0.00000
   O      1.94208   0.00000   1.94208  -0.03618  -0.00713   0.00000
   O      0.00000   1.94208   1.94208   0.08550   0.00302   0.00000
PRIMCOORD 15
      5   1
   Pb     0.00000   0.00000   0.00000   0.00000   0.00000  -0.00318
   Ti     1.94208   1.94208   1.94208   0.00000   0.00000  -0.02560
   O      1.94208   1.94208   0.00000   0.00000   0.00000   0.09529
   O      1.94208   0.00000   1.94208   0.00000   0.00000   0.01126
   O      0.00000   1.94208   1.94208   0.00000   0.00000   0.01126

   in pt.axsf file, i found that the last three columns in every mode is as large as 1/10 of eignvector in corresponding mode. i want to know if the last columns in every  mode are displacements of every atoms(fraction cooridinate)? please give me some instruction.
  Furthermore, the mode frequency in pt.out is very good as follows:
  #  mode   [cm-1]     [THz]       IR
    1   -145.12   -4.3504   74.6882
    2   -145.12   -4.3504   74.6882
    3      0.00    0.0000    0.0000
    4      0.00    0.0000    0.0000
    5      0.00    0.0000    0.0000
    6    102.83    3.0828    0.0759
    7    117.18    3.5128    6.5095
    8    117.18    3.5128    6.5095
    9    227.14    6.8095    0.0000
   10    227.14    6.8095    0.0000
   11    227.14    6.8095    0.0000
   12    416.50   12.4863   12.1192
   13    504.65   15.1288   15.3889
   14    504.65   15.1288   15.3889
   15    688.49   20.6403   84.3914
but in pt.molden the frequency is obtained:
      0.00
    0.00
    0.00
    0.00
    0.00
  102.83
  117.18
  117.18
  227.14
  227.14
  227.14
  416.50
  504.65
  504.65
  688.49
  in pt.axsf file, why the soft mode become zero?
  Regards
   
   
                            zhuzhenye 

		
---------------------------------
 ??????-3.5G???20M??
-------------- next part --------------
An HTML attachment was scrubbed...
URL: /pipermail/attachments/20060530/55dd34c7/attachment.htm 

From jasonsun98 at hotmail.com  Tue May 30 11:35:47 2006
From: jasonsun98 at hotmail.com (sun jason)
Date: Tue, 30 May 2006 09:35:47 +0000
Subject: [Pw_forum] about "free" structure
In-Reply-To: <20060529201101.24788.90243.Mailman@democritos.sissa.it>
Message-ID: <BAY107-F26E3024EB6DB3936484160DF920@phx.gbl>

Dear all,

The dynmat.x will not work if I run the scf and phonon calculation with a 
"free" structure by ibrav=0 and specified lattice parameters. for instance, 
if I use 
-------------------------
 &SYSTEM
                       ibrav = 2,
                           A = 3.612 ,
                           B = 3.612 ,
                           C = 3.612 ,
                       cosAB = 0.0 ,
                       cosAC = 0.0 ,
                       cosBC = 0.0 ,
                         nat = 2,
                        ntyp = 2,
                     ecutwfc = 70.0 ,
 /
 &ELECTRONS
                    conv_thr = 1.0d-10 ,
                 mixing_beta = 0.7 ,
 /
ATOMIC_SPECIES
    B   10.81100  B.pbe.tm.UPF
    N   14.00674  N.pbe.tm.UPF
 
ATOMIC_POSITIONS (crystal)
B        0.000000000   0.000000000   0.000000000
N        0.250000000   0.250000000   0.250000000
K_POINTS automatic
  6 6 6   0 0 0
------------------------------------------

but the dynmat.x will not work if I use this structure.
------------------------------------------
 &SYSTEM
                       ibrav = 0,
                    celldm(1)= 6.831,
                         nat = 2,
                        ntyp = 2,
                     ecutwfc = 70.0 ,
 /
 &ELECTRONS
                    conv_thr = 1.0d-10 ,
                 mixing_beta = 0.7 ,
 /
ATOMIC_SPECIES
    B   10.81100  B.pbe.tm.UPF
    N   14.00674  N.pbe.tm.UPF
 
ATOMIC_POSITIONS (crystal)
B        0.000000000   0.000000000   0.000000000
N        0.250000000   0.250000000   0.250000000
K_POINTS automatic
  6 6 6   1 1 1
CELL_PARAMETERS
0.50 0.00 0.50
0.00 0.50 0.50
0.50 0.50 0.00
-------------------------------------------
 
the dynamical matrix file bn.dynG are almost same except the first line
>   2    2  2  6.8256935  1.0000000  1.0000000  0.0000000  0.0000000  
0.0000000
---
<   2    2  0  6.8310000  0.0000000  0.0000000  0.0000000  0.0000000  
0.0000000


so should I manually change this line if I want to use the second method 
using "ibrav=0 and cell parameters" to describe the crystal structures for 
some low symmetry ones, such as Triclinic; and cell parameters are more 
convenient to get from relaxtion result.

Thank you in advance


Best,
===============================================
Jian SUN 
Physics Dept. of Nanjing University 
National Lab. of Solid State Microstructures 
22 Hankou Road, Gulou District 
Nanjing, Jiangsu Province 
210093 
China 
===============================================




From cyrille.barreteau at cea.fr  Tue May 30 11:39:49 2006
From: cyrille.barreteau at cea.fr (Cyrille Barreteau)
Date: Tue, 30 May 2006 11:39:49 +0200
Subject: [Pw_forum] Monkhorst Pack k points
Message-ID: <447C12E5.5060709@cea.fr>

Dear Pwscf users,

I am doing surface calculations with pw.x and I have a very simple
question concerning the automatic kpoints. I am generating the k points
by the following input:

K_POINTS automatic
n1 n2  1 0 0 0

According to the manuel the code is using MP k points but I am a bit 
surprised to find that
the Gamma= 0,0,0  point is always present in the k points list whatever the
parity of the integers n1 and n2. I was expecting to get the Gamma point 
only
when n1 and n2 are odd numbers (for example n1=n2=25).
But I also get the Gamma point when n1 and and n2 are even .

Is the code really using MP k points or another scheme?

cyrille
-- 
==================================================================

Cyrille Barreteau

CEA Saclay DSM/DRECAM/SPCSI

Batiment 462

91191 Gif sur Yvette Cedex FRANCE
~~~~~~~~~~~~~~~~~~~~~~~~
phone : +33 (0)1 69 08 29 51
fax : +33 (0)1 69 08 84 46
email : cyrille.barreteau at cea.fr

~~~~~~~~~~~~~~~~~~~~~~~~

http://www-drecam.cea.fr/Images/Pisp/cbarreteau/cbarreteau_fr.html

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



From ding at sissa.it  Tue May 30 11:29:53 2006
From: ding at sissa.it (Xunlei Ding)
Date: Tue, 30 May 2006 11:29:53 +0200
Subject: [Pw_forum] Monkhorst Pack k points
In-Reply-To: <447C12E5.5060709@cea.fr>
References: <447C12E5.5060709@cea.fr>
Message-ID: <447C1091.9080303@sissa.it>

Gamma is included in (6 6 1, 0 0 0) or (5 5 1, 0 0 0).
And no Gamma point is in (6 6 1, 1 1 0) because (1 1 0) means there are 
shifts in the first and second directions.


Cyrille Barreteau wrote:

> Dear Pwscf users,
>
> I am doing surface calculations with pw.x and I have a very simple
> question concerning the automatic kpoints. I am generating the k points
> by the following input:
>
> K_POINTS automatic
> n1 n2  1 0 0 0
>
> According to the manuel the code is using MP k points but I am a bit 
> surprised to find that
> the Gamma= 0,0,0  point is always present in the k points list 
> whatever the
> parity of the integers n1 and n2. I was expecting to get the Gamma 
> point only
> when n1 and n2 are odd numbers (for example n1=n2=25).
> But I also get the Gamma point when n1 and and n2 are even .
>
> Is the code really using MP k points or another scheme?
>
> cyrille



From jasonsun98 at hotmail.com  Tue May 30 11:55:06 2006
From: jasonsun98 at hotmail.com (sun jason)
Date: Tue, 30 May 2006 09:55:06 +0000
Subject: [Pw_forum] how to convert crystal description
In-Reply-To: <20060530054957.5637.73260.Mailman@democritos.sissa.it>
Message-ID: <BAY107-F121C0DD93B9BF42734F429DF920@phx.gbl>

dear all,

I have a silly question about structure description conversion,

for example, the relaxation result is like this:
-----------------------------------------------------------
     Final estimate of lattice vectors (input alat units)
   3.711193881  -0.138643395  -0.133568528
  -0.138262060   3.711099663   0.133354920
  -0.131446546   0.131591848   3.713501399
  final unit-cell volume =    343.8141 (a.u.)^3
  input alat =       1.8897 (a.u.)
-----------------------------------------------------------

then how to calculate a, b, c, and alpha, beta, gamma, if I want to 
describe the structure ?
I didn't find it in the solid state physics textbook, please give me some 
help.

Thank you very much!

Best regards,
===============================================
Jian SUN 
Physics Dept. of Nanjing University 
National Lab. of Solid State Microstructures 
22 Hankou Road, Gulou District 
Nanjing, Jiangsu Province 
210093 
China 
===============================================




From giannozz at nest.sns.it  Tue May 30 11:59:35 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Tue, 30 May 2006 11:59:35 +0200
Subject: [Pw_forum] Monkhorst Pack k points
In-Reply-To: <447C12E5.5060709@cea.fr>
References: <447C12E5.5060709@cea.fr>
Message-ID: <200605301159.35369.giannozz@nest.sns.it>

On Tuesday 30 May 2006 11:39, Cyrille Barreteau wrote:

> I am generating the k points by the following input:
>
> K_POINTS automatic
> n1 n2  1 0 0 0
>
> According to the manual the code is using MP k points but I am a bit
> surprised to find that the Gamma= 0,0,0  point is always present in 
> the k points list whatever the parity of the integers n1 and n2.

K_POINTS automatic
  n1 n2 n3 j1 j2 j3
yields a grid of k-points:
  k(i1,i2,i3)  = (i1+j1/2) * g1/n1 + (i2+j2/2) * g2/n2 + (i3+j3/2) * g3/n3
where g1, g2, g3 are the vectors that generate the reciprocal lattice;
the indices i1, i2, i3, run from 0 to n1-1, n2-1, n3-1 respectively;
j1, j2, j3 can be either 0 (no offset, k=0 in the grid) or 1 (with offset,
k=0 not in the grid). What is usually referred to as "Monkhorst-Pack" 
grid has offset and does not contain k=0 . Of course, only inequivalent
k-points are actually used

Paolo
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From giannozz at nest.sns.it  Tue May 30 12:00:45 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Tue, 30 May 2006 12:00:45 +0200
Subject: [Pw_forum] about "free" structure
In-Reply-To: <BAY107-F26E3024EB6DB3936484160DF920@phx.gbl>
References: <BAY107-F26E3024EB6DB3936484160DF920@phx.gbl>
Message-ID: <200605301200.45921.giannozz@nest.sns.it>

On Tuesday 30 May 2006 11:35, sun jason wrote:

> The dynmat.x will not work if I run the scf and phonon calculation
> with a "free" structure by ibrav=0 and specified lattice parameters.

it is a known problem, but fixing it requires some care in order
to avoid breaking something else

Paolo
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From cyrille.barreteau at cea.fr  Tue May 30 12:07:25 2006
From: cyrille.barreteau at cea.fr (Cyrille Barreteau)
Date: Tue, 30 May 2006 12:07:25 +0200
Subject: [Pw_forum] Monkhorst Pack k points
In-Reply-To: <200605301159.35369.giannozz@nest.sns.it>
References: <447C12E5.5060709@cea.fr> <200605301159.35369.giannozz@nest.sns.it>
Message-ID: <447C195D.5010307@cea.fr>

Ok so you are not using the "original" Monkhorst-Pack grid but a slightly
modified version always centered on the Gamma point (if  j1=j2=j3=0).

thanks
    cyrille

>On Tuesday 30 May 2006 11:39, Cyrille Barreteau wrote:
>
>  
>
>>I am generating the k points by the following input:
>>
>>K_POINTS automatic
>>n1 n2  1 0 0 0
>>
>>According to the manual the code is using MP k points but I am a bit
>>surprised to find that the Gamma= 0,0,0  point is always present in 
>>the k points list whatever the parity of the integers n1 and n2.
>>    
>>
>
>K_POINTS automatic
>  n1 n2 n3 j1 j2 j3
>yields a grid of k-points:
>  k(i1,i2,i3)  = (i1+j1/2) * g1/n1 + (i2+j2/2) * g2/n2 + (i3+j3/2) * g3/n3
>where g1, g2, g3 are the vectors that generate the reciprocal lattice;
>the indices i1, i2, i3, run from 0 to n1-1, n2-1, n3-1 respectively;
>j1, j2, j3 can be either 0 (no offset, k=0 in the grid) or 1 (with offset,
>k=0 not in the grid). What is usually referred to as "Monkhorst-Pack" 
>grid has offset and does not contain k=0 . Of course, only inequivalent
>k-points are actually used
>
>Paolo
>  
>


-- 
==================================================================

Cyrille Barreteau

CEA Saclay DSM/DRECAM/SPCSI

Batiment 462

91191 Gif sur Yvette Cedex FRANCE
~~~~~~~~~~~~~~~~~~~~~~~~
phone : +33 (0)1 69 08 29 51
fax : +33 (0)1 69 08 84 46
email : cyrille.barreteau at cea.fr

~~~~~~~~~~~~~~~~~~~~~~~~

http://www-drecam.cea.fr/Images/Pisp/cbarreteau/cbarreteau_fr.html

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



From giannozz at nest.sns.it  Tue May 30 12:34:35 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Tue, 30 May 2006 12:34:35 +0200
Subject: [Pw_forum] how to convert crystal description
In-Reply-To: <BAY107-F121C0DD93B9BF42734F429DF920@phx.gbl>
References: <BAY107-F121C0DD93B9BF42734F429DF920@phx.gbl>
Message-ID: <200605301234.35255.giannozz@nest.sns.it>

On Tuesday 30 May 2006 11:55, sun jason wrote:

> how to calculate a, b, c, and alpha, beta, gamma, if I want to
> describe the structure ?

a, b, c = modules of first, second, third vector respectively
(in A, not a.u.)
cos(ab) = cosine of the angle between first and second vector
and so on
ibrav=14 (triclinic) always works
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From eyvaz_isaev at yahoo.com  Tue May 30 13:25:14 2006
From: eyvaz_isaev at yahoo.com (Eyvaz Isaev)
Date: Tue, 30 May 2006 04:25:14 -0700 (PDT)
Subject: [Pw_forum] about "free" structure
In-Reply-To: <200605301200.45921.giannozz@nest.sns.it>
Message-ID: <20060530112514.86873.qmail@web60311.mail.yahoo.com>

Hi, 

The way to overcome the troubles with ibrav=0 
for phonon calculations is first to get cartesian
coordinates   using a specified basis vectors and then
turn to your "native" lattice and use "alat" units.

Bests,
Eyvaz.

--- Paolo Giannozzi <giannozz at nest.sns.it> wrote:

> On Tuesday 30 May 2006 11:35, sun jason wrote:
> 
> > The dynmat.x will not work if I run the scf and
> phonon calculation
> > with a "free" structure by ibrav=0 and specified
> lattice parameters.
> 
> it is a known problem, but fixing it requires some
> care in order
> to avoid breaking something else
> 
> Paolo
> -- 
> Paolo Giannozzi             e-mail: 
> giannozz at nest.sns.it
> Scuola Normale Superiore    Phone:   +39/050-509876,
> Fax:-563513 
> Piazza dei Cavalieri 7      I-56126 Pisa, Italy
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
> 


__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 


From ezadshojaee at hotmail.com  Tue May 30 15:27:43 2006
From: ezadshojaee at hotmail.com (Ezad Shojaee)
Date: Tue, 30 May 2006 13:27:43 +0000
Subject: [Pw_forum] about GGA implementation
Message-ID: <BAY19-F304C91FBCFB1AA6C9F399AA920@phx.gbl>

Hi
i want to implement GGA approximation in my performance for TiO2 & i have 
used GGA pseudopotentials .
i want to know what  can i do more?  ( in order to have a better band gap )
thanx

_________________________________________________________________
Express yourself instantly with MSN Messenger! Download today it's FREE! 
http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/



From marzari at MIT.EDU  Tue May 30 15:41:53 2006
From: marzari at MIT.EDU (Nicola Marzari)
Date: Tue, 30 May 2006 09:41:53 -0400
Subject: [Pw_forum] about GGA implementation
In-Reply-To: <BAY19-F304C91FBCFB1AA6C9F399AA920@phx.gbl>
References: <BAY19-F304C91FBCFB1AA6C9F399AA920@phx.gbl>
Message-ID: <447C4BA1.9000509@mit.edu>

Dear Ezad,

the gap is a property that DFT is not meant to describe
or predict. There is a long literature on the topic. You can
even have a real insulator described correctly (in exact DFT)
by a Kohn-Sham metal.

Notwithstanding this, hybrid functionals often seem to predict
Kohn-Sham gaps in closer agreement with experiments. As far as I know,
there aren't hybrid functionals implemented in pwscf yet.

Gaussian03 has an extensive choice (e.g. the HSE hybrid),
and I believe that now CPMD also has several of them.

Best luck,

			nicola

Ezad Shojaee wrote:

> 
> Hi
> i want to implement GGA approximation in my performance for TiO2 & i 
> have used GGA pseudopotentials .
> i want to know what  can i do more?  ( in order to have a better band gap )
> thanx
> 
---------------------------------------------------------------------
Prof Nicola Marzari   Department of Materials Science and Engineering
13-5066   MIT   77 Massachusetts Avenue   Cambridge MA 02139-4307 USA
tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu


From ezadshojaee at hotmail.com  Tue May 30 16:17:48 2006
From: ezadshojaee at hotmail.com (Ezad Shojaee)
Date: Tue, 30 May 2006 14:17:48 +0000
Subject: [Pw_forum] about GGA implementation
Message-ID: <BAY19-F21A70C0BA0AB382D62583AAA920@phx.gbl>

thank  you , dear Nicola

_________________________________________________________________
FREE pop-up blocking with the new MSN Toolbar - get it now! 
http://toolbar.msn.click-url.com/go/onm00200415ave/direct/01/



From akohlmey at vitae.cmm.upenn.edu  Tue May 30 16:50:58 2006
From: akohlmey at vitae.cmm.upenn.edu (Axel Kohlmeyer)
Date: Tue, 30 May 2006 10:50:58 -0400 (EDT)
Subject: [Pw_forum] about GGA implementation
In-Reply-To: <447C4BA1.9000509@mit.edu>
Message-ID: <Pine.LNX.4.44.0605301042531.12088-100000@localhost.localdomain>

On Tue, 30 May 2006, Nicola Marzari wrote:

hi,

[...]

NM> Notwithstanding this, hybrid functionals often seem to predict
NM> Kohn-Sham gaps in closer agreement with experiments. As far as I know,
NM> there aren't hybrid functionals implemented in pwscf yet.

to the best of my knowledge, exact exchange is an experimental
feature in the PW code, but needs to be ifdef'd in via -DEXX.
IIRC, it has been mentioned in this mailing list a couple of 
times. please search the mailing list archive.

NM> Gaussian03 has an extensive choice (e.g. the HSE hybrid),
NM> and I believe that now CPMD also has several of them.

correct. CPMD had exact exchange via ACM0/ACM1/ACM3
implemented for a while and as of the current version (3.11.1)
several hybrid functionals are implemented 
(B3LYP,X3LYP,B1LYP,PBE0,PBE1W).
note however, that this implementation has only seen
some limited testing and is computationally _extremely_
expensive (unlike in codes with gaussian basis sets). 

regards,
    axel.

NM> 
NM> Best luck,
NM> 
NM> 			nicola
NM> 
NM> Ezad Shojaee wrote:
NM> 
NM> > 
NM> > Hi
NM> > i want to implement GGA approximation in my performance for TiO2 & i 
NM> > have used GGA pseudopotentials .
NM> > i want to know what  can i do more?  ( in order to have a better band gap )
NM> > thanx
NM> > 
NM> ---------------------------------------------------------------------
NM> Prof Nicola Marzari   Department of Materials Science and Engineering
NM> 13-5066   MIT   77 Massachusetts Avenue   Cambridge MA 02139-4307 USA
NM> tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu
NM> _______________________________________________
NM> Pw_forum mailing list
NM> Pw_forum at pwscf.org
NM> http://www.democritos.it/mailman/listinfo/pw_forum
NM> 

-- 
=======================================================================
Axel Kohlmeyer   akohlmey at cmm.chem.upenn.edu   http://www.cmm.upenn.edu
   Center for Molecular Modeling   --   University of Pennsylvania
Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323
tel: 1-215-898-1582,  fax: 1-215-573-6233,  office-tel: 1-215-898-5425
=======================================================================
If you make something idiot-proof, the universe creates a better idiot.



From lijun_physics at yahoo.com.cn  Wed May 31 08:40:05 2006
From: lijun_physics at yahoo.com.cn (Lijun Zhang)
Date: Wed, 31 May 2006 14:40:05 +0800 (CST)
Subject: [Pw_forum] Questions about a soft Raman active phonon
Message-ID: <20060531064005.66095.qmail@web15603.mail.cnb.yahoo.com>

Dear all,
  I am doing some calculations on the phase transition induced by lattice 
  instability. With increasing pressures, I found that a Raman active phonon 
  softens to zero, the same as behaviors in the ferroelectricity.  An elastic 
  instability also appears upon compression, signaled by a sound velocity 
  going to zero. However, the former extremely precedes the latter, thus I 
  conclude that the phase transition is driven by the phonon instability. Is it 
  right? If true, how to search for the stable high pressure phase through the 
  eigenvectors of the unstable mode?
  Any reply will be appreciated.     
  Lijun Zhang

		
---------------------------------
 Mp3???-???????   
-------------- next part --------------
An HTML attachment was scrubbed...
URL: /pipermail/attachments/20060531/4a294578/attachment.htm 

From giannozz at nest.sns.it  Wed May 31 11:04:37 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Wed, 31 May 2006 11:04:37 +0200
Subject: [Pw_forum] question on random_matrix
In-Reply-To: <4471E3A5.2060500@polytechnique.edu>
References: <4471E3A5.2060500@polytechnique.edu>
Message-ID: <200605311104.37561.giannozz@nest.sns.it>

On Monday 22 May 2006 18:15, Virginie Trinit? wrote:

> I have a question about the subroutine random_matrix in PH
> [...]  I didn't understand why if nb is never the transform of na
> by some symmetry then the dynamical matrix is zero?

the comment in the header of that subroutine reads:

  !   Create a random hermitian matrix with non zero elements
  !   similar to the dynamical matrix of the system

but this is somewhat misleading. From what I remember the 
random matrix is used to find the irreducible representations 
of the small group of q (the phonon wavevector). If no symmetry 
sends atom na into atom nb, displacements of atoms na and nb 
will belong to different irreps.

Disclaimer: I didn't write, and never really understood, how
symmetry works in the phonon code, so the above might be 
all wrong!

Paolo
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From giannozz at nest.sns.it  Wed May 31 11:53:02 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Wed, 31 May 2006 11:53:02 +0200
Subject: [Pw_forum] what' the meaning of *.axsf file by dynmat.x
In-Reply-To: <20060530090937.94972.qmail@web15806.mail.cnb.yahoo.com>
References: <20060530090937.94972.qmail@web15806.mail.cnb.yahoo.com>
Message-ID: <200605311153.02581.giannozz@nest.sns.it>

On Tuesday 30 May 2006 11:09, ?? ? wrote:

>i have calculate LO-TO splitting of PbTiO3 by dynmat.x.
>the result is good. but i do not know the meaning of pt.axsf file.

the *.axsf file contains data to be plotted by XcrySDen, not for
visual inspection

>   in pt.axsf file, why the soft mode become zero?

because negative (i.e. imaginary) frequencies are set to zero.
You can change it if you want.

Paolo
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From Andrea.Marini at roma2.infn.it  Wed May 31 12:23:35 2006
From: Andrea.Marini at roma2.infn.it (Andrea Marini)
Date: Wed, 31 May 2006 12:23:35 +0200 (CEST)
Subject: [Pw_forum] Re: Different Gamma factors for Q-points in the same
 star
In-Reply-To: <20060529232134.93641.qmail@web60322.mail.yahoo.com>
References: <20060529232134.93641.qmail@web60322.mail.yahoo.com>
Message-ID: <Pine.LNX.4.64.0605311157001.28792@anomaly.roma2.infn.it>

On Mon, 29 May 2006, Eyvaz Isaev wrote:

> I am quite agree with your comment, but I suggested
> (and hopefully it is) that the symmetrization is done
> correctly in the code.  As you know the gamma
> approximately is a sum of production of two delta
> functions, connecting two parts of the Fermi surface,
> and is believed to converge slowly wrt number of
> k-sampling. That is why I suggested the use of  more
> k-points.

Dear Eyvaz & Stefano,

 	I do agree as well about what Stefano commented. And if we agree 
this means that using also a small set of k-points symmetry relations 
between the different gamma factors should be respected.

So, unless I am doing something wrong in my calculations (perfectly 
possible) there could be something wrong in the code.

Now, after studying the symmetrization of the dynamical matrix 
(symdynph_gq.f90) I understood (also thanks to Paolo comments) that it 
simply expands the sum on the k-points entering the gamma factor 
expression  to the whole BZ starting from the k-points reduced using the 
small group of q.

To my understanding, however, this  symmetrization of the electron-phonon 
matrix elements is not necessary as the electron-phonon matrix elements 
are more symmetric than the dynamical matrix (being a dot product).

Strictly speaking if I remove the call to

call symdynph_gq (xq, phi, s, invs, rtau, irt, irgq, nsymq, nat, &
                          irotmq, minus_q)

in symdyn_munu.f90 ONLY when this is called from elphon.f90 (in order to 
leave unchanged the symmetrization of the dynamical matrix) than I get the 
correct result for 2 the q-points of the same star.

Andrea

**************************************************************************
      q = (    0.250000000   0.750000000  -0.250000000 )

      omega( 1) =       2.619114 [THz] =      87.364805 [cm-1]
      omega( 2) =       3.738462 [THz] =     124.702500 [cm-1]
      omega( 3) =       4.666846 [THz] =     155.670255 [cm-1]

       Gaussian Broadening:   0.010 Ry, ngauss= -99
      DOS =  6.804424 states/spin/Ry/Unit Cell at Ef=  7.742912 eV
      double delta at Ef = 58.673351
      lambda( 1)=  2.1676   gamma=   96.62 GHz
      lambda( 2)=  1.0954   gamma=   99.47 GHz
      lambda( 3)=  0.1137   gamma=   16.09 GHz

**************************************************************************
      q = (    0.250000000  -0.250000000   0.750000000 )

      omega( 1) =       2.619204 [THz] =      87.367818 [cm-1]
      omega( 2) =       3.738232 [THz] =     124.694832 [cm-1]
      omega( 3) =       4.666853 [THz] =     155.670507 [cm-1]

      Gaussian Broadening:   0.010 Ry, ngauss= -99
      DOS =  6.804424 states/spin/Ry/Unit Cell at Ef=  7.742912 eV
      double delta at Ef = 58.673351
      lambda( 1)=  2.1675   gamma=   96.62 GHz
      lambda( 2)=  1.0955   gamma=   99.47 GHz
      lambda( 3)=  0.1137   gamma=   16.09 GHz

--------------------------------------------------------------------------
                                Andrea MARINI

        Physics Department, University of Rome "Tor Vergata" (Italy)
               - phone: +39-0672594894 - fax: +39-062023507 -

-- andrea.marini at roma2.infn.it - http://www.fisica.uniroma2.it/~marini/ --


From giannozz at nest.sns.it  Wed May 31 14:02:17 2006
From: giannozz at nest.sns.it (Paolo Giannozzi)
Date: Wed, 31 May 2006 14:02:17 +0200
Subject: [Pw_forum] Re: Different Gamma factors for Q-points in the same star
In-Reply-To: <Pine.LNX.4.64.0605311157001.28792@anomaly.roma2.infn.it>
References: <20060529232134.93641.qmail@web60322.mail.yahoo.com> <Pine.LNX.4.64.0605311157001.28792@anomaly.roma2.infn.it>
Message-ID: <200605311402.17227.giannozz@nest.sns.it>

On Wednesday 31 May 2006 12:23, Andrea Marini wrote:

> So, unless I am doing something wrong in my calculations 
> (perfectly possible) there could be something wrong in the code.

there is. If you perform the calculation of the Delta V and the 
calculation of electron-phonon coefficients in two separate steps,
you get correct results (or, more exactly, results that are not 
obviously wrong). If however you perform the two calculations in 
a single step, you get incorrect results. This is true for all versions
from 2.1 to 3.0 I think: version 2.0 didn't allow the single-step
calculation, version 3.1 is correct in this respect (no warranty that
it is correct in other respects, though).

I think I know why this happen: the small group of q and the crystal
group are stored in the same set of variables, recalculated and 
overwritten 2n+1 times. Since with the old algorithm all serious
electron-phonon calculations needed a two-step calculation 
(phonons and Delta V with a smaller k-point grid, electron-phonon
coefficients with a much larger grid) I hope that nobody published
wrong results obtained with this code.

As a side note: for some reason the Methefessel-Paxton smearing
(ngauss1=1, hardcoded in PH/elphon.f90) was used, but I don't 
see any good reason to prefer it to simple gaussian (ngauss1=0).

Other questions about the symmetrization requires some more
thought. Thank you Andrea for noticing the above problem.

Paolo
-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy


From Andrea.Marini at roma2.infn.it  Wed May 31 15:56:18 2006
From: Andrea.Marini at roma2.infn.it (Andrea Marini)
Date: Wed, 31 May 2006 15:56:18 +0200 (CEST)
Subject: [Pw_forum] Re: Different Gamma factors for Q-points in the same
 star
In-Reply-To: <200605311402.17227.giannozz@nest.sns.it>
References: <20060529232134.93641.qmail@web60322.mail.yahoo.com>
 <Pine.LNX.4.64.0605311157001.28792@anomaly.roma2.infn.it>
 <200605311402.17227.giannozz@nest.sns.it>
Message-ID: <Pine.LNX.4.64.0605311527040.11214@anomaly.roma2.infn.it>

On Wed, 31 May 2006, Paolo Giannozzi wrote:

> On Wednesday 31 May 2006 12:23, Andrea Marini wrote:
>
>> So, unless I am doing something wrong in my calculations
>> (perfectly possible) there could be something wrong in the code.
>
> there is. If you perform the calculation of the Delta V and the
> calculation of electron-phonon coefficients in two separate steps,
> you get correct results (or, more exactly, results that are not
> obviously wrong). If however you perform the two calculations in
> a single step, you get incorrect results. This is true for all versions
> from 2.1 to 3.0 I think: version 2.0 didn't allow the single-step
> calculation, version 3.1 is correct in this respect (no warranty that
> it is correct in other respects, though).

Perfect ! Thanks. Now it works :) ... but ....

> Other questions about the symmetrization requires some more
> thought.

... again if you remove the call to the symmetrization routine symdynph_gq 
from elphon.f90 you get exactly the same Eliashberg function (tested 
with 4 4 4 and 6 6 6 MP grid). The gamma factors, instead,  are rotated in 
such a way to be the same for degenerate modes).

My problem remains. I need the gkkp elements and not their square modulus. 
While I can perfectly get them if the symdynph_gq routine is not called I 
have no idea how to define them from the output of symdynph_gq :(

Andrea

--------------------------------------------------------------------------
                                Andrea MARINI

        Physics Department, University of Rome "Tor Vergata" (Italy)
               - phone: +39-0672594894 - fax: +39-062023507 -

-- andrea.marini at roma2.infn.it - http://www.fisica.uniroma2.it/~marini/ --


From baroni at sissa.it  Wed May 31 17:55:33 2006
From: baroni at sissa.it (Stefano Baroni)
Date: Wed, 31 May 2006 08:55:33 -0700
Subject: [Pw_forum] Questions about a soft Raman active phonon
In-Reply-To: <20060531064005.66095.qmail@web15603.mail.cnb.yahoo.com>
References: <20060531064005.66095.qmail@web15603.mail.cnb.yahoo.com>
Message-ID: <07C5CCF8-91DA-4E1B-AF34-9FAB140D7975@sissa.it>


On May 30, 2006, at 11:40 PM, Lijun Zhang wrote:

> Dear all,
> I am doing some calculations on the phase transition induced by  
> lattice
> instability. With increasing pressures, I found that a Raman active  
> phonon
> softens to zero, the same as behaviors in the ferroelectricity.  An  
> elastic
> instability also appears upon compression, signaled by a sound  
> velocity
> going to zero. However, the former extremely precedes the latter,  
> thus I
> conclude that the phase transition is driven by the phonon  
> instability. Is it
> right?

NO, it is not. For two reasons, which are somewhat related to one  
another. First, the phonon instability is usually coupled to the  
phonon instability, so that the phase transformation usually results  
from a distortion pattern which involves both the shape of the unit  
cell (elastic) and the atomic positions within the unit cell  
(phonon). As a consequence of this, the transition is usuall first- 
order, i.e. it occurs before any mode (either phonon-like or elastic)  
actually softens. Phonon or sound-velocity softening just signal the  
onset of an instability. In order to study the structural stability  
of the crystal as a function of pressure, you should study how the  
enthalpy (E+PV) of the systems depends on the "order parameters" of  
the transition for different values of the pressure. Good order  
parameter would obviously be the atomic distortion patterns  
corresponding to the soft phonon mode and the components of the  
strain corresponding to the vanishing sound velocity.

I am sure this is discussed in many different places. The references  
were what I have understood in this business is explained is:

M. Buongiorno Nardelli, S. Baroni, and P. Giannozzi, Phys. Rev. B 51,  
8060 (1995).

M. Buongiorno Nardelli, S. Baroni, and P. Giannozzi, Phys. Rev.  
Lett., 69, 1069 (1992).

Hope this helps
SB


> If true, how to search for the stable high pressure phase through the
> eigenvectors of the unstable mode?
> Any reply will be appreciated.
> Lijun Zhang
>
> Mp3???-???????

---
Stefano Baroni - SISSA  &  DEMOCRITOS National Simulation Center -  
Trieste
[+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype)

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: /pipermail/attachments/20060531/f34a5953/attachment.htm 

From baroni at sissa.it  Wed May 31 17:55:33 2006
From: baroni at sissa.it (Stefano Baroni)
Date: Wed, 31 May 2006 08:55:33 -0700
Subject: [Pw_forum] Questions about a soft Raman active phonon
In-Reply-To: <20060531064005.66095.qmail@web15603.mail.cnb.yahoo.com>
References: <20060531064005.66095.qmail@web15603.mail.cnb.yahoo.com>
Message-ID: <07C5CCF8-91DA-4E1B-AF34-9FAB140D7975@sissa.it>


On May 30, 2006, at 11:40 PM, Lijun Zhang wrote:

> Dear all,
> I am doing some calculations on the phase transition induced by  
> lattice
> instability. With increasing pressures, I found that a Raman active  
> phonon
> softens to zero, the same as behaviors in the ferroelectricity.  An  
> elastic
> instability also appears upon compression, signaled by a sound  
> velocity
> going to zero. However, the former extremely precedes the latter,  
> thus I
> conclude that the phase transition is driven by the phonon  
> instability. Is it
> right?

NO, it is not. For two reasons, which are somewhat related to one  
another. First, the phonon instability is usually coupled to the  
phonon instability, so that the phase transformation usually results  
from a distortion pattern which involves both the shape of the unit  
cell (elastic) and the atomic positions within the unit cell  
(phonon). As a consequence of this, the transition is usuall first- 
order, i.e. it occurs before any mode (either phonon-like or elastic)  
actually softens. Phonon or sound-velocity softening just signal the  
onset of an instability. In order to study the structural stability  
of the crystal as a function of pressure, you should study how the  
enthalpy (E+PV) of the systems depends on the "order parameters" of  
the transition for different values of the pressure. Good order  
parameter would obviously be the atomic distortion patterns  
corresponding to the soft phonon mode and the components of the  
strain corresponding to the vanishing sound velocity.

I am sure this is discussed in many different places. The references  
were what I have understood in this business is explained is:

M. Buongiorno Nardelli, S. Baroni, and P. Giannozzi, Phys. Rev. B 51,  
8060 (1995).

M. Buongiorno Nardelli, S. Baroni, and P. Giannozzi, Phys. Rev.  
Lett., 69, 1069 (1992).

Hope this helps
SB


> If true, how to search for the stable high pressure phase through the
> eigenvectors of the unstable mode?
> Any reply will be appreciated.
> Lijun Zhang
>
> Mp3???-???????

---
Stefano Baroni - SISSA  &  DEMOCRITOS National Simulation Center -  
Trieste
[+39] 040 3787 406 (tel) -528 (fax) / stefanobaroni (skype)

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: /pipermail/attachments/20060531/f34a5953/attachment-0001.htm