[Pw_forum] Problem with the Band Structure Calculation

[潘登]

> You have to be sure that you're using the same atomic configuration, same 
> k-point path,
thanks for your reply first.I got a question here.The same k-point
path means that the k-points in the band calculation are the points
which were generated in the scf run.

[Pw_forum] Problem with the Band Structure Calculation

[潘登]

Dear  Gabriele Sclauzero,
 I have tried the K_POINTS with the option fo 'CRYSTAL'.It is not the
problem.With the "CRYSTAL" the band structure seems to far more wrong.

[Pw_forum] Problem with the Band Structure Calculation

[lan haiping]

OK, i just check the INPUT, there is no problem for calculating eigenvalues.

On Thu, Mar 5, 2009 at 8:29 PM, lan haiping  wrote:

> Why do you set different weight numbers for K-path ? Is there any standing
> point for this ?
>
>
>
> On Thu, Mar 5, 2009 at 6:25 PM, ??  wrote:
>
>> Dear All,
>>
>>   I was trying to calculate the band structure of the BaNi2As2 which
>> had been calculated in the paper 0809.0499v2 on the arxiv.I could not
>> repeat the same struture.The  BaNi2As2 occurs in a body centered
>> tetragonal structure (I4/mmm) with Ba,Ni and As at the positions
>> 2a(0,0,0),4d (0.5,0,0.25)and 4e(0,0,z).Here z equels 0.3476 by the
>> experimental result.The lattice parameters are a=4.112 A and c=11.54
>> A.
>> Any comment about my input of this system?
>> my input of scf run:
>>
>>  
>> calculation = 'scf'
>> restart_mode='from_scratch'
>> prefix='BaNiAs',
>> pseudo_dir = '/disk2/xgwan/Quantum-Espresso/espresso-4.0.3/pseudo/',
>> outdir='/disk2/xgwan/tmp/'
>>  /
>>  
>> ibrav=6,
>> celldm(1)=7.7702,
>> celldm(3)=2.806,
>> nat=10,
>> ntyp=3,
>> ecutwfc = 40.0,
>> ecutrho = 400.0,
>> occupations= 'tetrahedra'
>>  /
>>  
>> diagonalization = 'cg'
>> mixing_beta = 0.7
>> conv_thr =  1.0d-6
>>  /
>> ATOMIC_SPECIES
>>  Ba 137.327 Ba.pbe-nsp-van.UPF
>>  Ni 58.6934 Ni.pbe-nd-rrkjus.UPF
>>  As 74.9216 As.pbe-n-van.UPF
>> ATOMIC_POSITIONS crystal
>>  Ba 0. 0.  0.
>>  Ba 0.5000 0.5000  0.5000
>>  Ni 0.0   0.50.25
>>  Ni 0.5   0.00.75
>>  Ni 0.5   0.00.25
>>  Ni 0.0   0.50.75
>>  As 0. 0.  0.3476
>>  As 0. 0.  0.6524
>>  As 0.5000 0.5000  0.8476
>>  As 0.5000 0.5000  0.1524
>> K_POINTS {automatic}
>>  8 8 8 0 0 0
>>
>> And my input of band calculation ,bands.x,plotband.x if helps:
>> input of band calculation
>>  
>> calculation = 'bands'
>> prefix='BaNiAs',
>> pseudo_dir = '/disk2/xgwan/Quantum-Espresso/espresso-4.0.3/pseudo/',
>> outdir='/disk2/xgwan/tmp/'
>> wf_collect=.true.
>>  /
>>  
>> ibrav=6,
>> celldm(1)=7.7702,
>> celldm(3)=2.806,
>> nat=10,
>> ntyp=3,
>> ecutwfc = 40.0,
>> ecutrho = 400.0,
>> occupations= 'tetrahedra'
>>  /
>>  
>> diagonalization = 'cg'
>> mixing_beta = 0.7
>> conv_thr =  1.0d-6
>>  /
>> ATOMIC_SPECIES
>>  Ba 137.327 Ba.pbe-nsp-van.UPF
>>  Ni 58.6934 Ni.pbe-nd-rrkjus.UPF
>>  As 74.9216 As.pbe-n-van.UPF
>> ATOMIC_POSITIONS crystal
>> Ba 0.5000 0.5000 0.5000
>> Ni 0.5000 0. 0.2500
>> Ni 0. 0.5000 0.2500
>> Ni 0. 0.5000 0.7500
>> Ni 0.5000 0. 0.7500
>> As 0. 0. 0.3476
>> As 0. 0. 0.6524
>> As 0.5000 0.5000 0.8476
>> As 0.5000 0.5000 0.1524
>> K_POINTS
>> 31
>> -0.50  0.50  0.50  1
>> -0.45  0.45  0.45  2
>> -0.40  0.40  0.40  3
>> -0.35  0.35  0.35  4
>> -0.30  0.30  0.30  5
>> -0.25  0.25  0.25  6
>> -0.20  0.20  0.20  7
>> -0.15  0.15  0.15  8
>> -0.10  0.10  0.10  9
>> -0.05  0.05  0.05  10
>>  0.00  0.00  0.00  11
>>  0.00  0.00  0.05  12
>>  0.00  0.00  0.10  13
>>  0.00  0.00  0.15  14
>>  0.00  0.00  0.20  15
>>  0.00  0.00  0.25  16
>>  0.00  0.00  0.30  17
>>  0.00  0.00  0.35  18
>>  0.00  0.00  0.40  19
>>  0.00  0.00  0.45  20
>>  0.00  0.00  0.50  21
>>  0.05  0.05  0.45  22
>>  0.10  0.10  0.40  23
>>  0.15  0.15  0.35  24
>>  0.20  0.20  0.30  25
>>  0.25  0.25  0.25  26
>>  0.20  0.30  0.20  27
>>  0.15  0.35  0.15  28
>>  0.10  0.40  0.10  29
>>  0.05  0.45  0.05  30
>>  0.0   0.50  0.0   31
>>
>> input of bands.x
>>  
>>prefix  = 'BaNiAs'
>>outdir = '/disk2/xgwan/tmp/'
>>filband = 'bands.dat'
>>  /
>>  input of plotband.x
>> bands.dat
>> 2 17
>> bands.xmgr
>> bands.ps
>> 10.5264
>> 1.0 10.5264
>>
>> With these inputs I could not get the same band structure as the paper
>> did.I am appreciate for any suggustion and comments.
>> Please tell me where I was wrong or which detail I ignored.
>>
>> Have a nice day.
>>
>> Deng Pan
>> Nanjing Unversity
>> ___
>> Pw_forum mailing list
>> Pw_forum at pwscf.org
>> http://www.democritos.it/mailman/listinfo/pw_forum
>>
>
>
>
> --
> Hai-Ping Lan
> Department of Electronics ,
> Peking University , Bejing, 100871
> lanhaiping at gmail.com, hplan at pku.edu.cn
>



-- 
Hai-Ping Lan
Department of Electronics ,
Peking University , Bejing, 100871
lanhaiping at gmail.com, hplan at pku.edu.cn
-- next part --
An HTML attachment was scrubbed...
URL: 
http://www.democritos.it/pipermail/pw_forum/attachments/20090305/8c6b9991/attachment-0001.htm
 

[Pw_forum] Problem with the Band Structure Calculation

[lan haiping]

Why do you set different weight numbers for K-path ? Is there any standing
point for this ?


On Thu, Mar 5, 2009 at 6:25 PM, ??  wrote:

> Dear All,
>
>   I was trying to calculate the band structure of the BaNi2As2 which
> had been calculated in the paper 0809.0499v2 on the arxiv.I could not
> repeat the same struture.The  BaNi2As2 occurs in a body centered
> tetragonal structure (I4/mmm) with Ba,Ni and As at the positions
> 2a(0,0,0),4d (0.5,0,0.25)and 4e(0,0,z).Here z equels 0.3476 by the
> experimental result.The lattice parameters are a=4.112 A and c=11.54
> A.
> Any comment about my input of this system?
> my input of scf run:
>
>  
> calculation = 'scf'
> restart_mode='from_scratch'
> prefix='BaNiAs',
> pseudo_dir = '/disk2/xgwan/Quantum-Espresso/espresso-4.0.3/pseudo/',
> outdir='/disk2/xgwan/tmp/'
>  /
>  
> ibrav=6,
> celldm(1)=7.7702,
> celldm(3)=2.806,
> nat=10,
> ntyp=3,
> ecutwfc = 40.0,
> ecutrho = 400.0,
> occupations= 'tetrahedra'
>  /
>  
> diagonalization = 'cg'
> mixing_beta = 0.7
> conv_thr =  1.0d-6
>  /
> ATOMIC_SPECIES
>  Ba 137.327 Ba.pbe-nsp-van.UPF
>  Ni 58.6934 Ni.pbe-nd-rrkjus.UPF
>  As 74.9216 As.pbe-n-van.UPF
> ATOMIC_POSITIONS crystal
>  Ba 0. 0.  0.
>  Ba 0.5000 0.5000  0.5000
>  Ni 0.0   0.50.25
>  Ni 0.5   0.00.75
>  Ni 0.5   0.00.25
>  Ni 0.0   0.50.75
>  As 0. 0.  0.3476
>  As 0. 0.  0.6524
>  As 0.5000 0.5000  0.8476
>  As 0.5000 0.5000  0.1524
> K_POINTS {automatic}
>  8 8 8 0 0 0
>
> And my input of band calculation ,bands.x,plotband.x if helps:
> input of band calculation
>  
> calculation = 'bands'
> prefix='BaNiAs',
> pseudo_dir = '/disk2/xgwan/Quantum-Espresso/espresso-4.0.3/pseudo/',
> outdir='/disk2/xgwan/tmp/'
> wf_collect=.true.
>  /
>  
> ibrav=6,
> celldm(1)=7.7702,
> celldm(3)=2.806,
> nat=10,
> ntyp=3,
> ecutwfc = 40.0,
> ecutrho = 400.0,
> occupations= 'tetrahedra'
>  /
>  
> diagonalization = 'cg'
> mixing_beta = 0.7
> conv_thr =  1.0d-6
>  /
> ATOMIC_SPECIES
>  Ba 137.327 Ba.pbe-nsp-van.UPF
>  Ni 58.6934 Ni.pbe-nd-rrkjus.UPF
>  As 74.9216 As.pbe-n-van.UPF
> ATOMIC_POSITIONS crystal
> Ba 0.5000 0.5000 0.5000
> Ni 0.5000 0. 0.2500
> Ni 0. 0.5000 0.2500
> Ni 0. 0.5000 0.7500
> Ni 0.5000 0. 0.7500
> As 0. 0. 0.3476
> As 0. 0. 0.6524
> As 0.5000 0.5000 0.8476
> As 0.5000 0.5000 0.1524
> K_POINTS
> 31
> -0.50  0.50  0.50  1
> -0.45  0.45  0.45  2
> -0.40  0.40  0.40  3
> -0.35  0.35  0.35  4
> -0.30  0.30  0.30  5
> -0.25  0.25  0.25  6
> -0.20  0.20  0.20  7
> -0.15  0.15  0.15  8
> -0.10  0.10  0.10  9
> -0.05  0.05  0.05  10
>  0.00  0.00  0.00  11
>  0.00  0.00  0.05  12
>  0.00  0.00  0.10  13
>  0.00  0.00  0.15  14
>  0.00  0.00  0.20  15
>  0.00  0.00  0.25  16
>  0.00  0.00  0.30  17
>  0.00  0.00  0.35  18
>  0.00  0.00  0.40  19
>  0.00  0.00  0.45  20
>  0.00  0.00  0.50  21
>  0.05  0.05  0.45  22
>  0.10  0.10  0.40  23
>  0.15  0.15  0.35  24
>  0.20  0.20  0.30  25
>  0.25  0.25  0.25  26
>  0.20  0.30  0.20  27
>  0.15  0.35  0.15  28
>  0.10  0.40  0.10  29
>  0.05  0.45  0.05  30
>  0.0   0.50  0.0   31
>
> input of bands.x
>  
>prefix  = 'BaNiAs'
>outdir = '/disk2/xgwan/tmp/'
>filband = 'bands.dat'
>  /
>  input of plotband.x
> bands.dat
> 2 17
> bands.xmgr
> bands.ps
> 10.5264
> 1.0 10.5264
>
> With these inputs I could not get the same band structure as the paper
> did.I am appreciate for any suggustion and comments.
> Please tell me where I was wrong or which detail I ignored.
>
> Have a nice day.
>
> Deng Pan
> Nanjing Unversity
> ___
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
>



-- 
Hai-Ping Lan
Department of Electronics ,
Peking University , Bejing, 100871
lanhaiping at gmail.com, hplan at pku.edu.cn
-- next part --
An HTML attachment was scrubbed...
URL: 
http://www.democritos.it/pipermail/pw_forum/attachments/20090305/b1b1481d/attachment.htm
 

[Pw_forum] Character energy band.

[Li YL]

Dear all users, 
How to plot character energy band using PWSCF code?  That is to say, I want to 
know how to classify the energy bands across Fermi level. For example, when 
both s-electron of A atom and p-electrons of B atom have contributions to Fermi 
level, how to distinguish which band is from s-electron of A atom and which 
band is from the p-electron of B atom. 
Many thanks. 
Best regards,
Yanling Li
Institute of solid states physics, CAS, China.


-- next part --
An HTML attachment was scrubbed...
URL: 
http://www.democritos.it/pipermail/pw_forum/attachments/20090305/971c04b7/attachment.htm
 

[Pw_forum] newbie update

[Carlo Nervi]

> Carlo Nervi wrote:
>
>> Total wall time (s) spent in this run:  5274.73
>> Reference:  720.04
>>
>> There are few discrepancies, but they seems
>> reasonables..
>> or not?
>
> they do. What doesn't seem reasonable to me is 2 1/2 h for
> a series of tests that take 12 min not on the "ExaFlop
> Machine", but on a rather standard 3GHz IntelCore2 E6850
> (the one I have in front). There is something really
> strange in your setup.

This is exactly what I was wondering. During the tests 4
CPUs were running at 10%, while the other four were
running at 1-5%.
It must be something wrong in my settings, but I cannot
figure it out.
Anyone could please provide a sort of "semi-standard"
make.sys for ifort and mkl with a working mpi?
Cheers,
  Carlo

--
--
Carlo Nervi carlo.nervi at unito.it Tel:+39 011 6707507/8
Fax: +39 011 6707855 - Dipartimento di Chimica IFM
via P. Giuria 7, 10125 Torino, Italy
http://lem.ch.unito.it/

[Pw_forum] newbie update

[Axel Kohlmeyer]

On Thu, 5 Mar 2009, Carlo Nervi wrote:

CN> time is almost 1/3 of the wall time. I suspect that Axel
CN> is right pointing to the disk I/O. I'm afraid that to
CN> increase the performance I should set the RAID 10...

no. this cannot be I/O. this is why i suggested this example.
unlike with pw.x in some configurations, there usually is no 
significant i/o when running cp.x. 
also you have almost exactly 4x the wall time relative to 
the cpu time. are you _sure_ you are running on a machine 
with 8 cores? _and_ is that machine otherwise empty?

when running on a local 2x quad-core intel xeon E5430 2.66GHz
node, i get 42.30s CPU time,43,87s wall time. 
with serial MKL and 8 MPI tasks.

CN> OMP=2, MPI=8
CN> v0:  CP   :  0m46.98s CPU time, 1m36.08s
CN> wall time
CN> little worse than OMP=1 and MPI=8

there is no point in oversubscribing a node. you rather
want to test, whether less than the maximum number of total
tasks per node is not giving you a better speedup.

cheers,
  axel.

[...]


-- 
===
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] DOS calculation for non collinear magnetic ordering

[Xun-Wang Yan]

auzero, PhD Student  |
> | c/o:   SISSA & CNR-INFM Democritos,  |
> |    via Beirut 2-4, 34014 Trieste (Italy) |
> | email: sclauzer at sissa.it |
> | phone: +39 040 3787 511  |
> | skype: gurlonotturno |
> o  o
> ___
> 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: 
http://www.democritos.it/pipermail/pw_forum/attachments/20090305/87feb39a/attachment-0001.htm
 

[Pw_forum] newbie update

[Paolo Giannozzi]

Carlo Nervi wrote:

> Total wall time (s) spent in this run:  5274.73
> Reference:  720.04
> 
> There are few discrepancies, but they seems reasonables..
> or not?

they do. What doesn't seem reasonable to me is 2 1/2 h for
a series of tests that take 12 min not on the "ExaFlop
Machine", but on a rather standard 3GHz IntelCore2 E6850
(the one I have in front). There is something really
strange in your setup.

> P.S.: I hope I did not post a message too much long...

messages longer tha 40Kb are blocked by the mailing list
software, so no, you didn't :-)

Paolo
-- 
Paolo Giannozzi, Democritos and University of Udine, Italy

[Pw_forum] Re[Psi] and Im[Psi]

[Aritz Leonardo Liceranzu]

Dear Baris Malcioglu,

THANK YOU for such a detailed explanation. I am going through it and I  
am sure that it is going to be very helpful.

Cheers,
Aritz

PD: Paolo, the integrations that I need to perform are in the real space.

- Forwarded message from baris.malcioglu at gmail.com -
 Date: Thu, 5 Mar 2009 10:16:28 +0100
 From: "O. Baris Malcioglu" 
Reply-To: PWSCF Forum 
  Subject: Re: [Pw_forum] Re[Psi] and Im[Psi]
   To: PWSCF Forum 

Dear Aritz,

First of all, you must notice that norm of the wavefunctions stored
will vary depending on what type of pseudopotential you use (Norm
conserving Ultrasoft etc.) The most straightforward in terms of
mathematical handling is the norm conserving one.

I prefer writing new subroutines rather than modifying old ones, since
I have witnessed doing that makes things get really messy after a
while.

Anyhow, here are some exempts from my personal notes, I hope you find
them useful. They are a bit unmaintained, thus as a disclaimer, please
report any error in them, so that I end up with better documentation.


1) How to calculate the real space grid points:

In this exempt, the aim is to find a set of grid points within a
certain distance from an atom.

The first step is to start an index rolling on available grid points

  DO ir = 1, nrxxs

nrxxs is explained as ?the total dimension of the smooth grid? in
gsmooth (PW/pwcom.f90) . It is the total number of grid points in the
smooth grid.

 !
 ! ... three dimensional indexes
 !
 index = index0 + ir - 1
 k = index / (nrx1s*nrx2s)
 index = index - (nrx1s*nrx2s)*k
 j = index / nrx1s
 index = index - nrx1s*j
 i = index
 !

Index0 is an offset for parallelism issues. It separates the grid into
processors and each processor starts processing from this offset.

nrx1s and nrx2s and nrx3s are again from gsmooth, and defined as
?maximum dimension of the smooth grid? in the corresponding direction,
they are used as the reduced forms of nrXs (X=1,2,3) dividing the grid
points among the nodes. In the serial case nrxxs=nrx1s*nrx2s*nrx3s but
this might or might not always be true for different parallel cases.
index (a temporary integer), nrx1s and nrx2s are all integers, so the
divisions above are integer divisions (otherwise the algorithm should
return 0 always, look at how the k and the index are calculated
consecutively).

As a convention, the segmentation of grid points should always start
from the 3rd dimension (z). The goal is to achieve an arrangement such
that the first 1..nrx1s elements correspond to ith element where j=0
k=0; the second 1..nrx1s element correspond to ith element where j=1
k=0 and etc.. This would make an index of higher rank increase only
when the cycles of all the lower ranks are complete. So the k index
(third dimension) is the total number of i and j cycles completed
(that would be nrx1s*nrx2s number of elements for each cycle). When k
times the number of elements per cycle is removed from the index
count, the excess only refers to i and j components. Using the same
trick to subtract the cycles of i (nrx1s elements each) the j index
value is obtained, and removing the number of elements per cycle for
each j increment from the index, the index for i is obtained.

 DO ipol = 1, 3
posi(ipol) = DBLE( i )*inv_nr1s*at(ipol,1) + &
 DBLE( j )*inv_nr2s*at(ipol,2) + &
 DBLE( k )*inv_nr3s*at(ipol,3)
 END DO

this might be on a non-orthogonal basis and we should be extra careful
in the next step. Here ipol is an index running over the basis vectors
(these will be called x y and z directions for further reference)
at(ipol,1..3) are the simulation cell base vectors (in real space)
divided by alat (lattice parameter), in terms of Cartesian components
(second index). nr1s, nr2s and nr3s are the actual number of grid
points (where nrxXs are reduced in a node specific way to conserve
memory) thus at/nr is the vector increment for the particular
direction. Multiplying this with the increment counter in that
direction, a first iteration for the position vector is obtained.

 !
 posi(:) = posi(:) - tau_ia(:)

 !

This line offsets the position such that current atom is at the
center. tau_ia is just a rename for the tau from Modules/ions_base.f90
which contains the positions of the atomic cores.

 ! ... minimum image convenction
 !
 CALL cryst_to_cart( 1, posi, bg, -1 )
 !
 posi(:) = posi(:) - ANINT( posi(:) )
 !
 CALL cryst_to_cart( 1, posi, at, 1 )
 !

cryst_to_cart is a subroutine responsible for conversions of vectors
from cartesian coordinates to crystal coordinates (+1) and viceversa
(-1). This part ensures that any distance more than 0.5 

[Pw_forum] PBE0 in PWSCF

[xinzheng li]

Dear friends,

My name is Xinzheng Li. I am a postdoc in University College London. I 
am using PWSCF to perform some PBE0 calculations for water cluster on 
metal surface recently. In order to make sure that the PBE0 number I got 
from PWSCF is correct. I performed a test for the binding energy of a 
water-dimer in gas phase (total energy of 2 water moleculars minus the 
total energy of a water dimer) taking the all-electron PBE0 results from 
Gaussian03 as reference.  The converged binding energy from  Gaussian03 
is around 207 meV. While after PBE0 calculation of PWSCF is performed, I 
have these lines in the output file:

!total energy  =   -68.48225787 Ry
Harris-Foulkes estimate   =   -72.38797788 Ry
estimated scf accuracy< 0.0008 Ry

The total energy is the sum of the following terms:

one-electron contribution =  -144.72933450 Ry
hartree contribution  =71.54274579 Ry
xc contribution   =   -12.87564636 Ry
ewald contribution=13.67425725 Ry
Fock energy 1 =-7.81160613 Ry
Fock energy 2 =-7.81177236 Ry
Half Fock energy 2=-3.90588618 Ry

I have two questions about this output:

1. Shall I take the '!   total energy' as the final total energy, or the 
'Harris-Foulkes estimate' as the final total energy? For the comment: 
'The total energy is the sum of the following terms', if I sum the 
numbers below, I don't get either 'Harris-Foulkes estimeate', or '! 
total energy'. Only if I sum the one-electron contribution, hartree 
contibution, xc contribution and the ewald contribution, can I get the 
'Harris-Foulkes estimate'. Does it mean that I should take 
'Harris-Foulkes estimate' as the final energy? If so, what is the 
meaning of the '! total energy'?

3. If I take Harris-Foulke estimate' as the total energy, the binding 
energy I get is 184 meV. If I take '! total energy' as total energy, the 
binding energy is 158 meV. Both of these two number are > 20 meV from 
the Gaussian all-electron benchmark. What is the expectation for the 
accuracy of such a value in the PWSCF code compared with all-electron 
Gaussian (when pseudopotentials, and supercell-size convergence are 
well-tested)?

Thanks a lot for your patiency in advance!

Sincere
Xinzheng



-
Dr. Xinzheng Li
London Centre for Nanotechnology, University College London
17-19 Gordon Street, London WC1H, 0AH
ucanxli at ucl.ac.uk
--

[Pw_forum] Problem with the Band Structure Calculation

[潘登]

Dear All,

   I was trying to calculate the band structure of the BaNi2As2 which
had been calculated in the paper 0809.0499v2 on the arxiv.I could not
repeat the same struture.The  BaNi2As2 occurs in a body centered
tetragonal structure (I4/mmm) with Ba,Ni and As at the positions
2a(0,0,0),4d (0.5,0,0.25)and 4e(0,0,z).Here z equels 0.3476 by the
experimental result.The lattice parameters are a=4.112 A and c=11.54
A.
Any comment about my input of this system?
my input of scf run:

 
 calculation = 'scf'
 restart_mode='from_scratch'
 prefix='BaNiAs',
 pseudo_dir = '/disk2/xgwan/Quantum-Espresso/espresso-4.0.3/pseudo/',
 outdir='/disk2/xgwan/tmp/'
 /
 
 ibrav=6,
 celldm(1)=7.7702,
 celldm(3)=2.806,
 nat=10,
 ntyp=3,
 ecutwfc = 40.0,
 ecutrho = 400.0,
 occupations= 'tetrahedra'
 /
 
 diagonalization = 'cg'
 mixing_beta = 0.7
 conv_thr =  1.0d-6
 /
ATOMIC_SPECIES
 Ba 137.327 Ba.pbe-nsp-van.UPF
 Ni 58.6934 Ni.pbe-nd-rrkjus.UPF
 As 74.9216 As.pbe-n-van.UPF
ATOMIC_POSITIONS crystal
 Ba 0. 0.  0.
 Ba 0.5000 0.5000  0.5000
 Ni 0.0   0.50.25
 Ni 0.5   0.00.75
 Ni 0.5   0.00.25
 Ni 0.0   0.50.75
 As 0. 0.  0.3476
 As 0. 0.  0.6524
 As 0.5000 0.5000  0.8476
 As 0.5000 0.5000  0.1524
K_POINTS {automatic}
  8 8 8 0 0 0

And my input of band calculation ,bands.x,plotband.x if helps:
input of band calculation
 
 calculation = 'bands'
 prefix='BaNiAs',
 pseudo_dir = '/disk2/xgwan/Quantum-Espresso/espresso-4.0.3/pseudo/',
 outdir='/disk2/xgwan/tmp/'
 wf_collect=.true.
 /
 
 ibrav=6,
 celldm(1)=7.7702,
 celldm(3)=2.806,
 nat=10,
 ntyp=3,
 ecutwfc = 40.0,
 ecutrho = 400.0,
 occupations= 'tetrahedra'
 /
 
 diagonalization = 'cg'
 mixing_beta = 0.7
 conv_thr =  1.0d-6
 /
ATOMIC_SPECIES
 Ba 137.327 Ba.pbe-nsp-van.UPF
 Ni 58.6934 Ni.pbe-nd-rrkjus.UPF
 As 74.9216 As.pbe-n-van.UPF
ATOMIC_POSITIONS crystal
Ba 0.5000 0.5000 0.5000
Ni 0.5000 0. 0.2500
Ni 0. 0.5000 0.2500
Ni 0. 0.5000 0.7500
Ni 0.5000 0. 0.7500
As 0. 0. 0.3476
As 0. 0. 0.6524
As 0.5000 0.5000 0.8476
As 0.5000 0.5000 0.1524
K_POINTS
31
-0.50  0.50  0.50  1
-0.45  0.45  0.45  2
-0.40  0.40  0.40  3
-0.35  0.35  0.35  4
-0.30  0.30  0.30  5
-0.25  0.25  0.25  6
-0.20  0.20  0.20  7
-0.15  0.15  0.15  8
-0.10  0.10  0.10  9
-0.05  0.05  0.05  10
 0.00  0.00  0.00  11
 0.00  0.00  0.05  12
 0.00  0.00  0.10  13
 0.00  0.00  0.15  14
 0.00  0.00  0.20  15
 0.00  0.00  0.25  16
 0.00  0.00  0.30  17
 0.00  0.00  0.35  18
 0.00  0.00  0.40  19
 0.00  0.00  0.45  20
 0.00  0.00  0.50  21
 0.05  0.05  0.45  22
 0.10  0.10  0.40  23
 0.15  0.15  0.35  24
 0.20  0.20  0.30  25
 0.25  0.25  0.25  26
 0.20  0.30  0.20  27
 0.15  0.35  0.15  28
 0.10  0.40  0.10  29
 0.05  0.45  0.05  30
 0.0   0.50  0.0   31

input of bands.x
 
prefix  = 'BaNiAs'
outdir = '/disk2/xgwan/tmp/'
filband = 'bands.dat'
 /
 input of plotband.x
bands.dat
2 17
bands.xmgr
bands.ps
10.5264
1.0 10.5264

With these inputs I could not get the same band structure as the paper
did.I am appreciate for any suggustion and comments.
Please tell me where I was wrong or which detail I ignored.

Have a nice day.

Deng Pan
Nanjing Unversity

[Pw_forum] Problem with the Band Structure Calculation

[Paolo Giannozzi]

?? wrote:

> Any comment about my input of this system?

>  occupations= 'tetrahedra'

tetrahedra should be used if you want to calculate the DOS.
Forces calculated with tetrahedra are not very accurate
(i.e. they are not exactly equal to the derivative of
the energy)

>  diagonalization = 'cg'

Davidson is usually faster

> With these inputs I could not get the same band structure 
> as the paper did

then look more carefully at your data and at your output.
Note that k-points are by default given in units of 2pi/a
(I mean: for all cartesian components, including the c axis)

Paolo
-- 
Paolo Giannozzi, Democritos and University of Udine, Italy

[Pw_forum] Problem with the Band Structure Calculation

[Gabriele Sclauzero]

I can't tell you if your structure is the same as that of the paper, I don't 
know these 
kind of compounds. You have to ensure that your structure is correct, I can 
only say that 
it seems reasonable.
Please try to point out clearly where your problem really is, otherwise noone 
will be able 
to help you.

GS

?? wrote:
> The structure in paper have the sysmmetry of the space group
> I4/mmm(139) with Ba 2a(0 0 0) Ni 4d (0.5 0 0.25) As (0 0 0.3476).
> and my input are:
> 
> 
>  
> calculation = 'scf'
> restart_mode='from_scratch'
> prefix='BaNiAs',
> pseudo_dir = '/disk2/xgwan/Quantum-Espresso/espresso-4.0.3/pseudo/',
> outdir='/disk2/xgwan/tmp/'
>  /
>  
> ibrav=6,
> celldm(1)=7.7702,
> celldm(3)=2.806,
> nat=10,
> ntyp=3,
> ecutwfc = 40.0,
> ecutrho = 400.0,
> occupations= 'tetrahedra'
>  /
>  
> diagonalization = 'cg'
> mixing_beta = 0.7
> conv_thr =  1.0d-6
>  /
> ATOMIC_SPECIES
>  Ba 137.327 Ba.pbe-nsp-van.UPF
>  Ni 58.6934 Ni.pbe-nd-rrkjus.UPF
>  As 74.9216 As.pbe-n-van.UPF
> ATOMIC_POSITIONS crystal
>  Ba 0. 0.  0.
>  Ba 0.5000 0.5000  0.5000
>  Ni 0.0   0.50.25
>  Ni 0.5   0.00.75
>  Ni 0.5   0.00.25
>  Ni 0.0   0.50.75
>  As 0. 0.  0.3476
>  As 0. 0.  0.6524
>  As 0.5000 0.5000  0.8476
>  As 0.5000 0.5000  0.1524
> K_POINTS {automatic}
>  8 8 8 0 0 0
> 
> 
> Could you tell the structure I inputed are the same of that is in the
> paper.I compare with XCryden to Wien2k input of this structure,it
> seems the same.
> ___
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
> 

-- 


o  o
| Gabriele Sclauzero, PhD Student  |
| c/o:   SISSA & CNR-INFM Democritos,  |
|via Beirut 2-4, 34014 Trieste (Italy) |
| email: sclauzer at sissa.it |
| phone: +39 040 3787 511  |
| skype: gurlonotturno |
o  o

[Pw_forum] "charge is wrong" error

[Aihua Zhang]

On Tue, Mar 3, 2009 at 9:35 PM, Paolo Giannozzi wrote:

> Gabriele Sclauzero wrote:
>
> > I think that CG style diagonalization has not been parallelized
> > like the Davidson one, so these lines at the beginning
> > of your output sound strange to me.
> >
> >>  >  a parallel distributed memory algorithm will be used,
> >>  >  eigenstates matrixes will be distributed block like on
> >>  >  ortho sub-group =2*   2 procs
>
> they aren't strange. This is for "subspace diagonalization",
> which is not directly used in CG diagonalization, but it is
> used in i) calculating starting wavefunctions, and ii) rotating
> wavefunctions before starting a new diagonalization (it is
> equivalent to first-order perturbation theory and improves
> quite a bit the subsequent diagonalization).
>
> Anyway I verified that the problem with charge is present only
> in parallel execution, both for CG and Davidson, irrespective
> whether subspace diagonalization is parallel or not
>

Is there any work-around now? Parallel execution is indispensable for large
systems.


>
> Paolo
> --
> Paolo Giannozzi, Democritos and University of Udine, Italy
> ___
> - Show quoted text -
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
>



-- 
--
Research Fellow,
Dept. of Physics, National Univ. of Singapore
10 Kent Ridge Crescent, Singapore, 119260
Tel: +65 6516 7844
--
------ next part --
An HTML attachment was scrubbed...
URL: 
http://www.democritos.it/pipermail/pw_forum/attachments/20090305/970d83ac/attachment-0001.htm
 

[Pw_forum] separate the surface state and surface resonance state from the bulk state

[yaoyugui]

Dear all,
 
I am simulating the Be(0001) surface through N-layer slab construction. 
But how can I decompose the calculated bands into surface states and bulk 
states. 

[Pw_forum] newbie update

[Axel Kohlmeyer]

On Thu, 5 Mar 2009, Carlo Nervi wrote:

CN> > Carlo Nervi wrote:
CN> >
CN> >> Total wall time (s) spent in this run:  5274.73
CN> >> Reference:  720.04
CN> >>
CN> >> There are few discrepancies, but they seems
CN> >> reasonables..
CN> >> or not?
CN> >
CN> > they do. What doesn't seem reasonable to me is 2 1/2 h for
CN> > a series of tests that take 12 min not on the "ExaFlop
CN> > Machine", but on a rather standard 3GHz IntelCore2 E6850
CN> > (the one I have in front). There is something really
CN> > strange in your setup.
CN> 
CN> This is exactly what I was wondering. During the tests 4
CN> CPUs were running at 10%, while the other four were
CN> running at 1-5%.
CN> It must be something wrong in my settings, but I cannot
CN> figure it out.

how much memory does your machine have? 
what else is running on that machine?

CN> Anyone could please provide a sort of "semi-standard"
CN> make.sys for ifort and mkl with a working mpi?

if you could run the tests, your make.sys should be ok.
either you do something really strange that limits the
MPI communications (and i cannot imagine what that would
be) or there is something else going on that you have not
paid attention to.
how does the machine behave, if you compile a serial executable?

cheers,
   axel.

CN> Cheers,
CN>   Carlo
CN> 
CN> --
CN> --
CN> Carlo Nervi carlo.nervi at unito.it Tel:+39 011 6707507/8
CN> Fax: +39 011 6707855 - Dipartimento di Chimica IFM
CN> via P. Giuria 7, 10125 Torino, Italy
CN> http://lem.ch.unito.it/
CN> 
CN> 
CN> ___
CN> Pw_forum mailing list
CN> Pw_forum at pwscf.org
CN> http://www.democritos.it/mailman/listinfo/pw_forum
CN> 

-- 
===
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] Character energy band.

[Gabriele Sclauzero]

Prasenjit Ghosh wrote:
> Dear Yanling,
> 
> You can use projwfc.x post-processing tool in the PP directory which 
> projects the DOS on the wavefunction of each atom. Then you can compare 
> the total DOS with the projected dos and you can tell which band is from 
> which wave function of which atom. Regarding the input file for it, you 
> can find the details in Doc/INPUT_PROJWFC.txt

The PDOS can be obtained with lsym=.TRUE. (which is the default), but if you 
want to know 
the projection on individual KS states (the PDOS is build up on these, by 
summing up on 
k-point and bands with correct weights), you need lsym=.FALSE.

cheers,

GS

> 
> With regards,
> Prasenjit.
> 
> 2009/3/5 Li YL mailto:ylli at theory.issp.ac.cn>>
> 
> Dear all users, 
> 
> How to plot character energy band using PWSCF code?  That is to
> say, I want to know how to classify the energy bands across Fermi level. 
> For example,
> when both /s/-electron of A atom and /p/-electrons of B atom have
> contributions to Fermi
> level, how to distinguish which band is from /s/-electron of A atom and 
> which band is from the p-electron of B atom. 
> 
> Many thanks. 
> 
> Best regards,
> 
> Yanling Li
> 
> Institute of solid states physics, CAS, China.
> 
> 
> ___
> Pw_forum mailing list
> Pw_forum at pwscf.org 
> http://www.democritos.it/mailman/listinfo/pw_forum
> 
> 
> 
> 
> -- 
> PRASENJIT GHOSH,
> POST-DOC,
> ROOM NO: 265, MAIN BUILDING,
> CM SECTION, ICTP,
> STRADA COSTERIA 11,
> TRIESTE, 34104,
> ITALY
> PHONE: +39 040 2240 369 (O)
>  +39 3891833716 (M)
> 
> 
> 
> 
> ___
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum

-- 


o  o
| Gabriele Sclauzero, PhD Student  |
| c/o:   SISSA & CNR-INFM Democritos,  |
|via Beirut 2-4, 34014 Trieste (Italy) |
| email: sclauzer at sissa.it |
| phone: +39 040 3787 511  |
| skype: gurlonotturno |
o  o

[Pw_forum] Problem with the Band Structure Calculation

[Gabriele Sclauzero]

?? wrote:
>> You have to be sure that you're using the same atomic configuration, same 
>> k-point path,
> thanks for your reply first.I got a question here.The same k-point
> path means that the k-points in the band calculation are the points
> which were generated in the scf run.

No, I meant the same k-point path used for bands in the paper. If you specified 
k-points 
coordinates for bands using the wrong option (e.g. alat instead of crystal), 
you may not 
get the k-points that you really want (not always, in some lattices the two 
options give 
the same k-points). Use the verbosity='high' in the pw run and you will see the 
difference 
(if any).

GS


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

-- 


o  o
| Gabriele Sclauzero, PhD Student  |
| c/o:   SISSA & CNR-INFM Democritos,  |
|via Beirut 2-4, 34014 Trieste (Italy) |
| email: sclauzer at sissa.it |
| phone: +39 040 3787 511  |
| skype: gurlonotturno |
o  o

[Pw_forum] Character energy band.

[Gabriele Sclauzero]

Li YL wrote:
> Dear all users, 
> 
> How to plot character energy band using PWSCF code?  That is to 
> say, I want to know how to classify the energy bands across Fermi level. For 
> example, 
> when both /s/-electron of A atom and /p/-electrons of B atom have 
> contributions to Fermi 
> level, how to distinguish which band is from /s/-electron of A atom and which 
> band is from the p-electron of B atom. 

The projwfc.x postprocessing program is all what you need. If things haven't 
benn changed 
much since last time I used it for such kind of study, you have to use an input 
file like 
this:


   
 prefix= "CoatPtwire",
 outdir= "/local_scratch/sclauzer/tmp/",
 lsym= .FALSE.,
 filproj= "COatPtwire.proj",
   /

using your values for prefix and filenames. In the  filproj file you should 
find, for each 
atomic state available in your system, the projection of the atomic states onto 
all KS 
states produced by the calculation, indexed by k-point and band (so that you 
can relate 
the KS eigenstates to the corresponding eigenvalue from the scf calculation 
output).

HTH

GS


> 
> Many thanks. 
> 
> Best regards,
> 
> Yanling Li
> 
> Institute of solid states physics, CAS, China.
> 
> 
> 
> 
> ___
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum

-- 


o  o
| Gabriele Sclauzero, PhD Student  |
| c/o:   SISSA & CNR-INFM Democritos,  |
|via Beirut 2-4, 34014 Trieste (Italy) |
| email: sclauzer at sissa.it |
| phone: +39 040 3787 511  |
| skype: gurlonotturno |
o  o

[Pw_forum] strange bandstructure

[idoldog]

dear Lex Kemper, Gabriele Sclauzero, Marcel Mohr,

I appreciate your help very much. I will try it follow your suggestions.

all best.

idoldog
2009.3.5


come from: Lex Kemper
send date: 2009-03-05 00:12:31
send to: PWSCF Forum
subject:Re: [Pw_forum] strange bandstructure

Hi idoldog,

I presume what you're referring to as 'strange' is the lines jumping
from one band to another. This is, as I understand it, the bands.x code
not seeing enough overlap between the kpoints to order them properly.
There are two ways to fix this:

(a) Run with more points per line in k-space
(b) If you're only interested in a few bands, fix them by hand.

Good luck!

Lex Kemper
University of Florida

idoldog wrote:
> Dear PWSCF user,
>  
> I just finished a bandstructure calculation about semiconductor SrTiO3. I get 
> a strange bandstructure.
>  
> In order to see clearly, I redraw the graph and mark same bands, which may be 
> wrong, with different colors.
>  
> Some friends tell me it would be better without lines.  Another friends tell 
> me to change the path of markde lines.
>  
> I want to know that whether these bands are correct and if they are wrong, 
> where my mistakes are and how to solve them.
>  
> The server blocked my files attched, so please visit 
> graph with colorful 
> lines:http://pic.emuch.net/200903/04/33e7b7b230bfd248104b8ba1ada65361.jpg
> graph without 
> line:http://pic.emuch.net/200903/04/4a2fdd6cbf9f7d0ce2df725c579e57ea.jpg
> 
> If you couldn't visit them successfully, mail me. 
>  
> I need your help.
>  
> all the best.  
>  
> Li Fei
> 2009-03-04
> 
> ___
> 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] Problem with the Band Structure Calculation

[Gabriele Sclauzero]

?? wrote:
> Dear  Gabriele Sclauzero,
>  I have tried the K_POINTS with the option fo 'CRYSTAL'.It is not the
> problem.With the "CRYSTAL" the band structure seems to far more wrong.

Before going on, if you want to use safely espresso you should understand those 
kind of 
differences (like those between alat or crystal options). You should judge 
which of the 
two options is needed in your case, mine was only a suggestion of a common 
mistake in BS 
calculations with pw.x.

Your input files do not have anything strange at first glance, but I don't know 
if they 
will compute the system you actually have in mind.
You have to be sure that you're using the same atomic configuration, same 
k-point path, 
but also the same functional and to be converged with respect to the relevant 
parameters 
(I suppose they have done it in the paper, if they want to publish it).

GS


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

-- 


o  o
| Gabriele Sclauzero, PhD Student  |
| c/o:   SISSA & CNR-INFM Democritos,  |
|via Beirut 2-4, 34014 Trieste (Italy) |
| email: sclauzer at sissa.it |
| phone: +39 040 3787 511  |
| skype: gurlonotturno |
o  o

[Pw_forum] Character energy band.

[Prasenjit Ghosh]

Dear Yanling,

You can use projwfc.x post-processing tool in the PP directory which
projects the DOS on the wavefunction of each atom. Then you can compare the
total DOS with the projected dos and you can tell which band is from which
wave function of which atom. Regarding the input file for it, you can find
the details in Doc/INPUT_PROJWFC.txt

With regards,
Prasenjit.

2009/3/5 Li YL 

>  Dear all users,
>
> How to plot character energy band using PWSCF code?  That is to
> say, I want to know how to classify the energy bands across Fermi level. For 
> example,
> when both *s*-electron of A atom and *p*-electrons of B atom have
> contributions to Fermi level, how to distinguish which band is from *s*
> -electron of A atom and which band is from the p-electron of B atom.
>
> Many thanks.
>
> Best regards,
>
> Yanling Li
>
> Institute of solid states physics, CAS, China.
>
> ___
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
>
>


-- 
PRASENJIT GHOSH,
POST-DOC,
ROOM NO: 265, MAIN BUILDING,
CM SECTION, ICTP,
STRADA COSTERIA 11,
TRIESTE, 34104,
ITALY
PHONE: +39 040 2240 369 (O)
 +39 3891833716 (M)
-- next part --
An HTML attachment was scrubbed...
URL: 
http://www.democritos.it/pipermail/pw_forum/attachments/20090305/f140cff0/attachment.htm
 

[Pw_forum] Problem with the Band Structure Calculation

[Lex Kemper]

I have a passing familiarity with your system and this problem.
Comparing band structure is hard, for a number of reasons. One is an
inconsistency in the notation -- for simple tetragonal units cells, for
example, there are at least two different notations for the
high-symmetry points. Secondly, the primitive units cell for system
you're looking at is not simple tetragonal, it's body centered
tetragonal. This has a whole different set of symmetry points. Now, you
can use k-points from the conventional (tetragonal) unit cell, but the
number of bands you get will vary, depending on whether you're using the
primitive zone or the conventional one.

Good luck,

Lex Kemper
University of Florida

?? wrote:
> The structure in paper have the sysmmetry of the space group
> I4/mmm(139) with Ba 2a(0 0 0) Ni 4d (0.5 0 0.25) As (0 0 0.3476).
> and my input are:
> 
> 
>  
> calculation = 'scf'
> restart_mode='from_scratch'
> prefix='BaNiAs',
> pseudo_dir = '/disk2/xgwan/Quantum-Espresso/espresso-4.0.3/pseudo/',
> outdir='/disk2/xgwan/tmp/'
>  /
>  
> ibrav=6,
> celldm(1)=7.7702,
> celldm(3)=2.806,
> nat=10,
> ntyp=3,
> ecutwfc = 40.0,
> ecutrho = 400.0,
> occupations= 'tetrahedra'
>  /
>  
> diagonalization = 'cg'
> mixing_beta = 0.7
> conv_thr =  1.0d-6
>  /
> ATOMIC_SPECIES
>  Ba 137.327 Ba.pbe-nsp-van.UPF
>  Ni 58.6934 Ni.pbe-nd-rrkjus.UPF
>  As 74.9216 As.pbe-n-van.UPF
> ATOMIC_POSITIONS crystal
>  Ba 0. 0.  0.
>  Ba 0.5000 0.5000  0.5000
>  Ni 0.0   0.50.25
>  Ni 0.5   0.00.75
>  Ni 0.5   0.00.25
>  Ni 0.0   0.50.75
>  As 0. 0.  0.3476
>  As 0. 0.  0.6524
>  As 0.5000 0.5000  0.8476
>  As 0.5000 0.5000  0.1524
> K_POINTS {automatic}
>  8 8 8 0 0 0
> 
> 
> Could you tell the structure I inputed are the same of that is in the
> paper.I compare with XCryden to Wien2k input of this structure,it
> seems the same.
> ___
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum

[Pw_forum] Problem with the Band Structure Calculation

[Lorenzo Paulatto]

On Thu, 05 Mar 2009 13:29:34 +0100, lan haiping  wrote:

> Why do you set different weight numbers for K-path ? Is there any  
> standing point for this ?

Weights are irrelevant for "bands" calculations.

regards


-- 
Lorenzo Paulatto
SISSA  &  DEMOCRITOS (Trieste)
phone: +39 040 3787 511
skype: paulatz
www:   http://people.sissa.it/~paulatto/

 *** save italian brains ***
 http://saveitalianbrains.wordpress.com/

[Pw_forum] newbie update

[Carlo Nervi]

> but this does NOT work:
> OMP_NUM_THREAD=1
> ./myprogram
>
> mysteries of bash shell...
>
I think this is due to the fact that bash temporarely set
the variables for the single command...

Anyway, I did few tests, as suggested by Axel, on 32 water
molecule (example 21).
I started to play with the version 4.1 and forgot that it
is under testing.
The following results refers to the version 4.1CVS ...
sorry...
I did two exacutables, v0 is without -ipo, v1 is compiled
with -ipo
I do have to try to set the sequential, but in my mkl lib
it does not exists any lib-sequential...
therefore I'm going to try to use some compiler switches
Here are the results (read OMP as OMP_NUM_THREADS and MPI
as number of cpu used with mpiexec):

The test using OMP=8 and no MPI:
v0:CP   :  4m 1.98s CPU time, 3m 0.83s
wall time
v1:CP   :  3m49.07s CPU time, 2m57.33s
wall time

Using OMP=1 and MPI=8
v0:CP   :  0m36.43s CPU time, 1m31.59s
wall time
v1:CP   :  0m35.39s CPU time, 1m30.67s
wall time

There is a clear advantage in using MPI, even though each
%CPU (with MPI) is only between 40-50%. In fact the CPU
time is almost 1/3 of the wall time. I suspect that Axel
is right pointing to the disk I/O. I'm afraid that to
increase the performance I should set the RAID 10...

OMP=2, MPI=8
v0:  CP   :  0m46.98s CPU time, 1m36.08s
wall time
little worse than OMP=1 and MPI=8

OMP=4. MPI=4
v0: CP   :  1m25.82s CPU time, 2m 4.96s
wall time

OMP=8, MPI=8
v0: CP   :  1m17.36s CPU time, 2m11.83s
wall time
Here there is probably some interferences between parallel
approaches.


I did also a couple of run on 64 water molecules:

OMP=1, MPI=8
v0: CP   :  3m41.49s CPU time, 7m14.67s
wall time

OMP=8, no MPI
v0: CP   : 15m27.18s CPU time,12m56.67s
wall time

In this case the ratio CPU time/wall time is slightly
better, and in fact the %CPU usage were between 50-60%

Furthermore, I did also the tests/check-pw.x.j on the
version v1 only:

./check-pw.x.j
Checking atom-lsda...passed
Checking atom-pbe...discrepancy in number of scf
iterations detected
Reference: 7, You got: 5
discrepancy in pressure detected
Reference: -14.44, You got: -14.48
Checking atom-sigmapbe...discrepancy in pressure detected
Reference: -15.02, You got: -15.01
Checking atom...passed
Checking berry...passed
Checking berry, step 2 ...passed
Checking electric0...discrepancy in number of scf
iterations detected
Reference: 8, You got: 9
Checking electric1...passed
Checking electric2...passed
Checking eval_infix...passed
Checking eval_infix, step 2 ...discrepancy in HOMO detected
Reference: -8.4542, You got: -8.4554
discrepancy in LUMO detected
Reference: -0.4297, You got: -0.4300
Checking lattice-ibrav0-abc...passed
Checking lattice-ibrav0-cell_parameters+a...passed
Checking lattice-ibrav0-cell_parameters+celldm...passed
Checking lattice-ibrav0-cell_parameters...passed
Checking lattice-ibrav1-kauto...passed
Checking lattice-ibrav1...passed
Checking lattice-ibrav10-kauto...passed
Checking lattice-ibrav10...passed
Checking lattice-ibrav11-kauto...passed
Checking lattice-ibrav11...passed
Checking lattice-ibrav12-kauto...passed
Checking lattice-ibrav12...passed
Checking lattice-ibrav13-kauto...passed
Checking lattice-ibrav13...passed
Checking lattice-ibrav14-kauto...passed
Checking lattice-ibrav14...passed
Checking lattice-ibrav2-kauto...passed
Checking lattice-ibrav2...passed
Checking lattice-ibrav3-kauto...passed
Checking lattice-ibrav3...passed
Checking lattice-ibrav4-kauto...passed
Checking lattice-ibrav4...passed
Checking lattice-ibrav5-kauto...passed
Checking lattice-ibrav5...passed
Checking lattice-ibrav6-kauto...passed
Checking lattice-ibrav6...passed
Checking lattice-ibrav7-kauto...passed
Checking lattice-ibrav7...passed
Checking lattice-ibrav8-kauto...passed
Checking lattice-ibrav8...passed
Checking lattice-ibrav9-kauto...passed
Checking lattice-ibrav9...passed
Checking lda+U-noU...passed
Checking lda+U-user_ns...passed
Checking lda+U...passed
Checking lsda-cg...passed
Checking lsda-mixing_TF...passed
Checking lsda-mixing_localTF...passed
Checking lsda-mixing_ndim...passed
Checking lsda-nelup+neldw...passed
Checking lsda-tot_magnetization...passed
Checking lsda...passed
Checking lsda, step 2 ...passed
Checking md-pot_extrap1...passed
Checking md-pot_extrap2...passed
Checking md-wfc_extrap1...passed
Checking md-wfc_extrap2...passed
Checking md...passed
Checking metaGGA...passed
Checking metadyn...passed
Checking metal-fermi_dirac...passed
Checking metal-gaussian...passed
Checking metal-tetrahedra...passed
Checking metal-tetrahedra, step 2 ...passed
Checking metal...passed
Checking metal, step 2 ...passed
Checking neb1-H2+H...passed
Checking neb2-H2+H-symm...passed
Checking neb3-H2+H-asym...passed
Checking noncolin-cg...passed
Checking noncolin-constrain_angle...passed
Checking noncolin-constrain_atomic...discrepancy in total
energy 

[Pw_forum] Xcrsyden

[Álvaro Alves]

Hello All. I got to the spin polarization using pp.x, creating a file
with extension xsf. However when trying to open in Xcrysden, the system
is deformed. How to solve this?

??? Sincerely, Alvaro

Universidade Federal Fluminense - Niteroi - Rio de Janeiro - Brasil



  Veja quais s?o os assuntos do momento no Yahoo! +Buscados
http://br.maisbuscados.yahoo.com
-- next part --
An HTML attachment was scrubbed...
URL: 
http://www.democritos.it/pipermail/pw_forum/attachments/20090305/3c844521/attachment.htm
 

[Pw_forum] parallelization issues

[Mahmoud Payami]

Dear Gabriele and Stefano,

Thank you very much for your explanations.
Bests,
 Mahmoud
 


> It means that since very little communication is needed across different 
> images this parallelization does not suffer much (at all) if the 
> communication network is not very fast... however it does not 
> distribute  the memory and I/O ... which may be bad for the efficiency 
> of the calculation as we are learning these days from the discussion 
> going on on pool parallelization (which also is less-communication and 
> more-memory/O intensive than R parallelization)...
> My today understanding of this is that one could increase the number of 
> processors using R parallelization as far as the network at hand 
> allows to get a good scaling... then one can increase further the number 
> of processors by adding pool parallelization and, in case of  NEB 
> calculations, some image parallelization.
> 
> Hope this helps,
>  Stefano de Gironcoli
> 
> Mahmoud Payami wrote:
>> Dear ALL,
>>  
>> In sec. of "parallelization issues" of UG, is mentioned that: "As a 
>> general rule, image parallelization may give good scaling..."
>> Could anybody please explain what exactly it means in terms of say, 
>> static linking, using more npools, or so.
>>  
>> Bests,
>>Mahmoud Payami
>> Physics Group, AEOI.
>>  
>>  
>>  
>>  
>>  
>> 
>>
>> ___
>> 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] parallelization issues

[Mahmoud Payami]

Dear ALL,

In sec. of "parallelization issues" of UG, is mentioned that: "As a general 
rule, image parallelization may give good scaling..."
Could anybody please explain what exactly it means in terms of say, static 
linking, using more npools, or so.

Bests,
   Mahmoud Payami
Physics Group, AEOI.




-- next part --
An HTML attachment was scrubbed...
URL: 
http://www.democritos.it/pipermail/pw_forum/attachments/20090305/ab8dd5f2/attachment.htm
 

[Pw_forum] band and dos input files in xcrysden

[Eduardo Ariel Menendez Proupin]

Well, I also like the way the tasks are distributed, it is conceptually very
clear. I know a different code that saves a DOS file after a band
calculation and the newbies come to me with strange peaks in DOS that are
caused by a flat band along a line.
Calling dos from pw.x  or pw from dos.x masks that conceptual clarity.
However, we have to be practical and run the frequent calculations as easily
as possible. I would prepend a nscf calculation with "standard options" ,
and give warning in the otput file that non-standard options can be done
doing the nscf calculation separately.

Also, I would alway make pw.x to write the self-consistent DOS in a file
after an scf calculation, no matter how coarse is the k-points grid. I find
this useful for checking purposes (e.g. to know if an impurity have levels
in the gaps). Also, for simple metals the k-points grid needed to have
converged energies and forces also is enough to have a good DOS. Then let
the user the choice to do post processing if want a DOS with an occupation
scheme different than the SCF. I would put this file in the .save directory
and name it prefix.scfdos or something like that.

Boris, thank for your script. I will soon test it. Maybe it can be embedded
in PWGUI.

Best regards
Eduardo

-- Mensaje reenviado --

> From: Paolo Giannozzi 
> To: PWSCF Forum 
> Date: Thu, 26 Feb 2009 10:34:30 +0100
> Subject: Re: [Pw_forum] band and dos input files in xcrysden
>
> On Feb 21, 2009, at 17:05 , Eduardo Ariel Menendez Proupin wrote:
>
>  Let me remind how is the process to obtain the DOS. It has three
>> calculations
>> and 3 input files
>>
>> pw.x < si.scf.in > si.scf.out   # self consistent
>> pw.x  si.dos.out  # non selfconsistent dense k-points mesh
>> and a few options
>> dos.x   si.dos2.out  # postprocessing
>>
>
> Hi Eduardo,  the issue you raise is a serious one and requires some though.
> Right now several standard calculations in q-e are clumsy because they
> require separate steps. While I think that it is a good idea to keep
> separate
> steps separate, I also think that at least the most common calculations
> might
> be streamlined. I am not convinced that the PWGui is the right tool for
> this,
> though, at least not in the present form (a tool to produce input data).
> The ideal  solution would be the usage of a high-level scripting language
> like
> python to "glue" the various pieces together, but this is highly nontrivial
> (at least for me and for 99% of q-e users).
> A simpler option could be to collapse some calculations into the same
> executable:
> for instance
> - add a call to dos after a nscf calculation in pw.x,
>  add a call to bands after a bands calculation in pw.x
> or
> - prepend a nscf calculation to dos.x,
>  prepend a band calculation to bands.x
> Suggestions are welcome
>
> Paolo
> ---
> Paolo Giannozzi, Democritos and University of Udine, Italy
>
>
>
>
>
-- next part --
An HTML attachment was scrubbed...
URL: 
http://www.democritos.it/pipermail/pw_forum/attachments/20090305/e9b64e3d/attachment-0001.htm
 

[Pw_forum] Problem with the Band Structure Calculation

[Gabriele Sclauzero]

Are you absolutely sure that the atomic configuration and k-point path are the 
same as 
those in the paper?

Maybe K_POINTS in the bands calculation needs the 'CRYSTAL' option?

GS

?? wrote:
> Dear All,
> 
>I was trying to calculate the band structure of the BaNi2As2 which
> had been calculated in the paper 0809.0499v2 on the arxiv.I could not
> repeat the same struture.The  BaNi2As2 occurs in a body centered
> tetragonal structure (I4/mmm) with Ba,Ni and As at the positions
> 2a(0,0,0),4d (0.5,0,0.25)and 4e(0,0,z).Here z equels 0.3476 by the
> experimental result.The lattice parameters are a=4.112 A and c=11.54
> A.
> Any comment about my input of this system?
> my input of scf run:
> 
>  
>  calculation = 'scf'
>  restart_mode='from_scratch'
>  prefix='BaNiAs',
>  pseudo_dir = '/disk2/xgwan/Quantum-Espresso/espresso-4.0.3/pseudo/',
>  outdir='/disk2/xgwan/tmp/'
>  /
>  
>  ibrav=6,
>  celldm(1)=7.7702,
>  celldm(3)=2.806,
>  nat=10,
>  ntyp=3,
>  ecutwfc = 40.0,
>  ecutrho = 400.0,
>  occupations= 'tetrahedra'
>  /
>  
>  diagonalization = 'cg'
>  mixing_beta = 0.7
>  conv_thr =  1.0d-6
>  /
> ATOMIC_SPECIES
>  Ba 137.327 Ba.pbe-nsp-van.UPF
>  Ni 58.6934 Ni.pbe-nd-rrkjus.UPF
>  As 74.9216 As.pbe-n-van.UPF
> ATOMIC_POSITIONS crystal
>  Ba 0. 0.  0.
>  Ba 0.5000 0.5000  0.5000
>  Ni 0.0   0.50.25
>  Ni 0.5   0.00.75
>  Ni 0.5   0.00.25
>  Ni 0.0   0.50.75
>  As 0. 0.  0.3476
>  As 0. 0.  0.6524
>  As 0.5000 0.5000  0.8476
>  As 0.5000 0.5000  0.1524
> K_POINTS {automatic}
>   8 8 8 0 0 0
> 
> And my input of band calculation ,bands.x,plotband.x if helps:
> input of band calculation
>  
>  calculation = 'bands'
>  prefix='BaNiAs',
>  pseudo_dir = '/disk2/xgwan/Quantum-Espresso/espresso-4.0.3/pseudo/',
>  outdir='/disk2/xgwan/tmp/'
>  wf_collect=.true.
>  /
>  
>  ibrav=6,
>  celldm(1)=7.7702,
>  celldm(3)=2.806,
>  nat=10,
>  ntyp=3,
>  ecutwfc = 40.0,
>  ecutrho = 400.0,
>  occupations= 'tetrahedra'
>  /
>  
>  diagonalization = 'cg'
>  mixing_beta = 0.7
>  conv_thr =  1.0d-6
>  /
> ATOMIC_SPECIES
>  Ba 137.327 Ba.pbe-nsp-van.UPF
>  Ni 58.6934 Ni.pbe-nd-rrkjus.UPF
>  As 74.9216 As.pbe-n-van.UPF
> ATOMIC_POSITIONS crystal
> Ba 0.5000 0.5000 0.5000
> Ni 0.5000 0. 0.2500
> Ni 0. 0.5000 0.2500
> Ni 0. 0.5000 0.7500
> Ni 0.5000 0. 0.7500
> As 0. 0. 0.3476
> As 0. 0. 0.6524
> As 0.5000 0.5000 0.8476
> As 0.5000 0.5000 0.1524
> K_POINTS
> 31
> -0.50  0.50  0.50  1
> -0.45  0.45  0.45  2
> -0.40  0.40  0.40  3
> -0.35  0.35  0.35  4
> -0.30  0.30  0.30  5
> -0.25  0.25  0.25  6
> -0.20  0.20  0.20  7
> -0.15  0.15  0.15  8
> -0.10  0.10  0.10  9
> -0.05  0.05  0.05  10
>  0.00  0.00  0.00  11
>  0.00  0.00  0.05  12
>  0.00  0.00  0.10  13
>  0.00  0.00  0.15  14
>  0.00  0.00  0.20  15
>  0.00  0.00  0.25  16
>  0.00  0.00  0.30  17
>  0.00  0.00  0.35  18
>  0.00  0.00  0.40  19
>  0.00  0.00  0.45  20
>  0.00  0.00  0.50  21
>  0.05  0.05  0.45  22
>  0.10  0.10  0.40  23
>  0.15  0.15  0.35  24
>  0.20  0.20  0.30  25
>  0.25  0.25  0.25  26
>  0.20  0.30  0.20  27
>  0.15  0.35  0.15  28
>  0.10  0.40  0.10  29
>  0.05  0.45  0.05  30
>  0.0   0.50  0.0   31
> 
> input of bands.x
>  
> prefix  = 'BaNiAs'
> outdir = '/disk2/xgwan/tmp/'
> filband = 'bands.dat'
>  /
>  input of plotband.x
> bands.dat
> 2 17
> bands.xmgr
> bands.ps
> 10.5264
> 1.0 10.5264
> 
> With these inputs I could not get the same band structure as the paper
> did.I am appreciate for any suggustion and comments.
> Please tell me where I was wrong or which detail I ignored.
> 
> Have a nice day.
> 
> Deng Pan
> Nanjing Unversity
> ___
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
> 

-- 


o  o
| Gabriele Sclauzero, PhD Student  |
| c/o:   SISSA & CNR-INFM Democritos,  |
|via Beirut 2-4, 34014 Trieste (Italy) |
| email: sclauzer at sissa.it |
| phone: +39 040 3787 511  |
| skype: gurlonotturno |
o  o

[Pw_forum] newbie update

[Henning Glawe]

On Thu, Mar 05, 2009 at 10:51:43AM +0100, Lorenzo Paulatto wrote:
> You have to export the variable, for it to work.
> e.g. this works:
> OMP_NUM_THREAD=1 ./myprogram
> 
> this works as well:
> export OMP_NUM_THREAD=1
> ./myprogram
> 
> but this does NOT work:
> OMP_NUM_THREAD=1
> ./myprogram
> 
> mysteries of bash shell...

general property of shell variables:
if you set a variable without exporting it, this variable is set only for
this shell, but not for its child processes (such as the pw.x).
a child process only "sees" environment variables, which you create with the
export command :)

-- 
c u
henning

[Pw_forum] newbie update

[Lorenzo Paulatto]

On Wed, 04 Mar 2009 22:32:48 +0100, JR Schmidt  wrote:
> (I set it in my profile, so it should have stuck around
> when executed by mpirun). 


You have to export the variable, for it to work.
e.g. this works:
OMP_NUM_THREAD=1 ./myprogram

this works as well:
export OMP_NUM_THREAD=1
./myprogram

but this does NOT work:
OMP_NUM_THREAD=1
./myprogram

mysteries of bash shell...

regards


-- 
Lorenzo Paulatto
SISSA  &  DEMOCRITOS (Trieste)
phone: +39 040 3787 511
skype: paulatz
www:   http://people.sissa.it/~paulatto/

 *** save italian brains ***
 http://saveitalianbrains.wordpress.com/

[Pw_forum] Re[Psi] and Im[Psi]

[O. Baris Malcioglu]

Dear Aritz,

First of all, you must notice that norm of the wavefunctions stored
will vary depending on what type of pseudopotential you use (Norm
conserving Ultrasoft etc.) The most straightforward in terms of
mathematical handling is the norm conserving one.

I prefer writing new subroutines rather than modifying old ones, since
I have witnessed doing that makes things get really messy after a
while.

Anyhow, here are some exempts from my personal notes, I hope you find
them useful. They are a bit unmaintained, thus as a disclaimer, please
report any error in them, so that I end up with better documentation.


1) How to calculate the real space grid points:

In this exempt, the aim is to find a set of grid points within a
certain distance from an atom.

The first step is to start an index rolling on available grid points

 DO ir = 1, nrxxs

nrxxs is explained as ?the total dimension of the smooth grid? in
gsmooth (PW/pwcom.f90) . It is the total number of grid points in the
smooth grid.

!
! ... three dimensional indexes
!
index = index0 + ir - 1
k = index / (nrx1s*nrx2s)
index = index - (nrx1s*nrx2s)*k
j = index / nrx1s
index = index - nrx1s*j
i = index
!

Index0 is an offset for parallelism issues. It separates the grid into
processors and each processor starts processing from this offset.

nrx1s and nrx2s and nrx3s are again from gsmooth, and defined as
?maximum dimension of the smooth grid? in the corresponding direction,
they are used as the reduced forms of nrXs (X=1,2,3) dividing the grid
points among the nodes. In the serial case nrxxs=nrx1s*nrx2s*nrx3s but
this might or might not always be true for different parallel cases.
index (a temporary integer), nrx1s and nrx2s are all integers, so the
divisions above are integer divisions (otherwise the algorithm should
return 0 always, look at how the k and the index are calculated
consecutively).

As a convention, the segmentation of grid points should always start
from the 3rd dimension (z). The goal is to achieve an arrangement such
that the first 1..nrx1s elements correspond to ith element where j=0
k=0; the second 1..nrx1s element correspond to ith element where j=1
k=0 and etc.. This would make an index of higher rank increase only
when the cycles of all the lower ranks are complete. So the k index
(third dimension) is the total number of i and j cycles completed
(that would be nrx1s*nrx2s number of elements for each cycle). When k
times the number of elements per cycle is removed from the index
count, the excess only refers to i and j components. Using the same
trick to subtract the cycles of i (nrx1s elements each) the j index
value is obtained, and removing the number of elements per cycle for
each j increment from the index, the index for i is obtained.

DO ipol = 1, 3
   posi(ipol) = DBLE( i )*inv_nr1s*at(ipol,1) + &
DBLE( j )*inv_nr2s*at(ipol,2) + &
DBLE( k )*inv_nr3s*at(ipol,3)
END DO

this might be on a non-orthogonal basis and we should be extra careful
in the next step. Here ipol is an index running over the basis vectors
(these will be called x y and z directions for further reference)
at(ipol,1..3) are the simulation cell base vectors (in real space)
divided by alat (lattice parameter), in terms of Cartesian components
(second index). nr1s, nr2s and nr3s are the actual number of grid
points (where nrxXs are reduced in a node specific way to conserve
memory) thus at/nr is the vector increment for the particular
direction. Multiplying this with the increment counter in that
direction, a first iteration for the position vector is obtained.

!
posi(:) = posi(:) - tau_ia(:)

!

This line offsets the position such that current atom is at the
center. tau_ia is just a rename for the tau from Modules/ions_base.f90
which contains the positions of the atomic cores.

! ... minimum image convenction
!
CALL cryst_to_cart( 1, posi, bg, -1 )
!
posi(:) = posi(:) - ANINT( posi(:) )
!
CALL cryst_to_cart( 1, posi, at, 1 )
!

cryst_to_cart is a subroutine responsible for conversions of vectors
from cartesian coordinates to crystal coordinates (+1) and viceversa
(-1). This part ensures that any distance more than 0.5 crystal unit
far is treated according to the closest atomic center, not the one it
is further away. It is actually the minimum Brillouin cell for the
atom at hand.

distsq = posi(1)**2 + posi(2)**2 + posi(3)**2
 END DO
  END DO

the distance from the center...

2) How to FFT a given band into real space (most simplistic version,
no gamma point tricks, no non-collinear, etc.)

psic(nls(igk(:))) = evc(:,ibnd)
psic is the buffer where this kind of 

[Pw_forum] separate the surface state and surface resonance state from the bulk state

[Stefano de Gironcoli]

Dear Wenmei Ming,

surface states are, by definition, localized at the surface and, in term
of band energy, they are located in gaps of the bulk-projected bund
structure.
as such you can identify them in the list of eigenvalues and then plot
their three dimensional density using (for instance) the post processing
tool pp.x (plot_num=7)...
resonant surface states are more tricky as they do not correspond to a
single state but to an energy window and extend inside your slab ...
they can be visualized by looking at the integrated local density of
states (ILDOS) plot_num=10 in the energy window of interest.

Hope this helps,

stefano de Gironcoli -SISSA & DEMOCRITOS


yaoyugui wrote:
> Dear all,
> I am simulating the Be(0001) surface through N-layer slab construction.
> But how can I decompose the calculated bands into surface states and
> bulk states.
> From a few publications concerning DFT calculation of surface structure,
> the final bands are showed as surface projected bulk bands,
> distinctively pointing out
> the surface states and the surface resonance states.
> How can I separate the surface state and surface resonance state from
> the bulk state
> using PWSCF package? or additional codes are required ?
> Could anyone give me suggestion?
> Thanks a million.
> Wenmei Ming
> Institue of Physics,
> Chinese Academy of Sciences
> Beijing 100190
> People's Republic of China
>
> 
> ? Windows Live Messenger  ?
> 
> 
>
> ___
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
>   

[Pw_forum] parallelization issues

[Stefano de Gironcoli]

It means that since very little communication is needed across different 
images this parallelization does not suffer much (at all) if the 
communication network is not very fast... however it does not 
distribute  the memory and I/O ... which may be bad for the efficiency 
of the calculation as we are learning these days from the discussion 
going on on pool parallelization (which also is less-communication and 
more-memory/O intensive than R parallelization)...
My today understanding of this is that one could increase the number of 
processors using R parallelization as far as the network at hand 
allows to get a good scaling... then one can increase further the number 
of processors by adding pool parallelization and, in case of  NEB 
calculations, some image parallelization.

Hope this helps,
  Stefano de Gironcoli

Mahmoud Payami wrote:
> Dear ALL,
>  
> In sec. of "parallelization issues" of UG, is mentioned that: "As a 
> general rule, image parallelization may give good scaling..."
> Could anybody please explain what exactly it means in terms of say, 
> static linking, using more npools, or so.
>  
> Bests,
>Mahmoud Payami
> Physics Group, AEOI.
>  
>  
>  
>  
>  
> 
>
> ___
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
>   

[Pw_forum] parallelization issues

[Gabriele Sclauzero]

Mahmoud Payami wrote:
> Dear ALL,
>  
> In sec. of "parallelization issues" of UG, is mentioned that: "As a 
> general rule, image parallelization may give good scaling..."

These is relevant to you only if you are performing NEB calculations.

> Could anybody please explain what exactly it means in terms of say, 
> static linking, using more npools, or so.

I think it has nothing to do with static linking and in general with any 
compilation 
issue. Once you correctly compiled a parallel version, you're free to chose 
which kind of 
parallelization to eploit, depending of your type of system and calculation.
If you're doing NEB, then the image parallelization is the more efficients 
since different 
images have to "speak" very few one to the other, while if not doing NEB you 
can still use 
the other levels of parallelization described in the user guide. They do not 
scale as well 
as image parallelization, but still do a good job to speed up your code or let 
you study 
very big systems.

cheers,

GS

>  
> Bests,
>Mahmoud Payami
> Physics Group, AEOI.
>  
>  
>  
>  
>  
> 
> 
> 
> 
> ___
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum

-- 


o  o
| Gabriele Sclauzero, PhD Student  |
| c/o:   SISSA & CNR-INFM Democritos,  |
|via Beirut 2-4, 34014 Trieste (Italy) |
| email: sclauzer at sissa.it |
| phone: +39 040 3787 511  |
| skype: gurlonotturno |
o  o

[Pw_forum] newbie update

[Henning Glawe]

On Wed, Mar 04, 2009 at 07:03:27PM -0500, Axel Kohlmeyer wrote:
> also one has to consider that debian based systems (incl ubuntu), 
> don't use /bin/bash as /bin/sh, but have a version of /bin/ash as 
> /bin/sh and that one only reads .profile (which many bash setups 
> in turn don't) but, of course, not .bashrc.

actually, it is only ubuntu doing this by default. even debian lenny has bash
as /bin/sh (although, of course, one should never rely on sh==bash,
especially on machines running on commercial unix variants). therefore when
relying on bash features (such as reading .bashrc), one should explicitely
write #!/bin/bash and not #!/bin/sh into the job scripts.
in case the queuing system does not call jobscript not directly, but instead
forces /bin/sh as an interpreter, set the appropriate job options to use
/bin/bash instead (in torque, this is job option '-S').

-- 
c u
henning

[Pw_forum] Fw: nimage is larger than the available number of images?

[Gabriele Sclauzero]

J. J. Ramsey wrote:
>> I don't get it. nimage = num_of_images, so why is there a problem?
> 
> Ok, I figured out what I did to myself. 

Well done! ;-)

> Unless the first and last images are being
> optimized (done by setting first_last_opt to .TRUE. in the  namelist), 
> then the
> number of images available to be parallelized is two less than num_of_images, 
> so if
> num_of_images = 7, then nimage <= 5.

Good that you have shared the piece of information you got by yourself with all 
the 
community of users.

Thanks!

Cheers,

GS

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

-- 


o  o
| Gabriele Sclauzero, PhD Student  |
| c/o:   SISSA & CNR-INFM Democritos,  |
|via Beirut 2-4, 34014 Trieste (Italy) |
| email: sclauzer at sissa.it |
| phone: +39 040 3787 511  |
| skype: gurlonotturno |
o  o

[Pw_forum] Re[Psi] and Im[Psi]

[Paolo Giannozzi]

On Mar 4, 2009, at 17:52 , Aritz Leonardo Liceranzu wrote:

> For a tddft problem that I am involved into, I need to numerically
> integrate the equilibrium KS wavefunctions. For this purpose I would
> like to obtain for each (k,i) Bloch state a chart containing:
>
> Rx,Ry,Rz, Re[psi], Im[psi]

in real or reciprocal space? in the latter case, you could modify the
"pw2casino" utility. In the former case, you can get the wavefunctions
on the FFT grid by Fourier-transfoming them. The correspondence
between points in real space and indices in the FFT grid is explained
in the FAQ of the user guide. You might try to modify code "elf.f90",
for instance: calculate \psi(r) iinstead of grad\psi(r)

P.
---
Paolo Giannozzi, Democritos and University of Udine, Italy

[Pw_forum] Re[Psi] and Im[Psi]

[Aritz Leonardo Liceranzu]

No hints on what to modify on the code to write the wave-function?

:(

Thanks again
Aritz

- Forwarded message from swblelia at ehu.es -
 Date: Wed, 4 Mar 2009 17:52:05 +0100
 From: Aritz Leonardo Liceranzu 
Reply-To: PWSCF Forum 
  Subject: [Pw_forum] Re[Psi] and Im[Psi]
   To: pw_forum at pwscf.org

Hi pwscf users,

My question is regarding the output of the KS wave-functions.

For a tddft problem that I am involved into, I need to numerically
integrate the equilibrium KS wavefunctions. For this purpose I would
like to obtain for each (k,i) Bloch state a chart containing:

Rx,Ry,Rz, Re[psi], Im[psi]

I surfed through old emails (like the one below) and I found that this
feature is not implemented in PP codes (am I right?).

Ab-init produces this kind of output but I am trying to avoid using it
(I am loyal to espresso :) ). On the other hand, it shouldn't be very
difficult to write a small program that does it, any hints so that I
do not have to use ab-init

Thank you in advance
Aritz




Lorenzo Paulatto paulatto at sissa.it
Sat Feb 23 12:48:15 CET 2008

On Sab, 23 Febbraio 2008 5:42 am, nafise rezaei wrote:
>> However, I want to extract and to plot parts of real and imaginary
>> wave function.
>> Do you know any program to generate wave function? If there is no
>> program, I can modify what program to write it?

The postprocessing utility pp.x can extract the square modulus of a
wavefunction (you choose it by specifying kpoint and band index), plotting
real and imaginary parts is not implemented, but you may be able to do it
youself.

On the other hand, if nobody have felt the need to implement it in the
last not so few years there could be a reason. If you tell us what you
want to do, instead of how you want to do it, you will probably get some
more useful advice.

Regards, LP



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


- End forwarded message -

[Pw_forum] strange bandstructure

[idoldog]

Dear PWSCF user,
 
I just finished a bandstructure calculation about semiconductor SrTiO3. I get a 
strange bandstructure.
 
In order to see clearly, I redraw the graph and mark same bands, which may be 
wrong, with different colors.
 
Some friends tell me it would be better without lines.  Another friends tell me 
to change the path of markde lines.
 
I want to know that whether these bands are correct and if they are wrong, 
where my mistakes are and how to solve them.
 
The server blocked my files attched, so please visit 
graph with colorful 
lines:http://pic.emuch.net/200903/04/33e7b7b230bfd248104b8ba1ada65361.jpg
graph without 
line:http://pic.emuch.net/200903/04/4a2fdd6cbf9f7d0ce2df725c579e57ea.jpg

If you couldn't visit them successfully, mail me. 
 
I need your help.
 
all the best.  
 
Li Fei
2009-03-04