<DIV>Dear Lorenzo Paulatto:</DIV><BR><PRE>> 1. Re: about relax in electric field (Lorenzo Paulatto)
> 2. Re: Pw_forum Digest, Vol 29, Issue 68 (Johnson, D. Ray)
> 3. Calculation of projected density of states for        individual
> atoms in chain (Dimpy Sharma)
> 4. Re: Calculation of projected density of states for individual
> atoms in chain (Lorenzo Paulatto)
> 5. cp.x code (Jiri Houska)
> 6. Re: cp.x code (Paolo Giannozzi)
> 7. Trouble with parallel running ld1.x (Pavel A. Akulov)
>
>
>----------------------------------------------------------------------
>
>Message: 1
>Date: Tue, 24 Nov 2009 16:26:13 +0100
>From: "Lorenzo Paulatto" <paulatto@sissa.it>
>Subject: Re: [Pw_forum] about relax in electric field
>To: "PWSCF Forum" <pw_forum@pwscf.org>
>Message-ID: <op.u3wntz0ua8x26q@paulax>
>Content-Type: text/plain; charset=utf-8; format=flowed; delsp=yes
>
>Replying to message "[Pw_forum] about relax in electric field" from
>Q.J.Wang (24/11/09):
>> PAY ATTENTION: EL FIELD AND OCCUPATIONS
>
>Dear Q.J. Wang,
>this message is not very explicative, but it means that Berry-phase
>electric field is not really well defined with smearing and can cause
>problems. In particular, it is not defined for metals, or any system that
>have fractional band occupations.
>
>If you have a band gap and your smearing is much smaller than that, then
>dipfield may work, but if some fractional occupation arises, the system
>may never converge, or converge to some meaningless state.
>
>In practice, if your system has a band gap, albeit small, you should try
>to use fixed occupations. Otherwise, you cannot use dipfield.</PRE><PRE>I only want to add a homogeneous finite electric field .In the instruction of the dipfield paremeter ,it said dipfield turn on only tefield is turn .So I don't know why you introduce the usage of parameter dipfield ? </PRE><PRE> I have see the code of forces_bp_efield.f90 in version 4.1(my cluster installed) .It seems that :"PAY ATTENTION: EL FIELD AND OCCUPATIONS" is not a error .My question is why when I turn on the lefield ,relax is so slower than without ? And one puzzled problem is ,at the start ,when the first iteration has not achieved the save file has turn up .</PRE><PRE> My input file is as following ,please give me some advice ,the problem has puzzled some days ,but don't find effective solution until now .</PRE><PRE>Best regards</PRE><PRE>>
>best regards
>
>--
>Lorenzo Paulatto
>SISSA & DEMOCRITOS (Trieste)
>phone: +39 040 3787 511
>skype: paulatz
>www: http://people.sissa.it/~paulatto/
>
> *** save italian brains ***
> <A href="http://saveitalianbrains.wordpress.com/>>>------------------------------>>Message">http://saveitalianbrains.wordpress.com/
>
>
>------------------------------
>
<FONT color="#000000"></FONT>
>Message</A>: 6
>Date: Tue, 24 Nov 2009 22:47:35 +0800 (CST)
>From: "Q.J.Wang" <wangqj1@126.com>
>Subject: [Pw_forum] about relax in electric field
>To: pw_forum <pw_forum@pwscf.org>
>Message-ID:
> <26724746.1247551259074055444.JavaMail.coremail@bj126app104.126.com>
>Content-Type: text/plain; charset="gbk"
>
>Dear all
> When I add external electric field to optimize the structure .It runs very very slower than not add the electric field , after 2 days ,the output file is still as following :
>
> iteration # 1 ecut= 30.00 Ry beta=0.20
> Davidson diagonalization with overlap
> ethr = 1.00E-02, avg # of iterations = 4.4
> PAY ATTENTION: EL FIELD AND OCCUPATIONS
> Expectation value of exp(iGx):
> (2.751012649613009E-016,-3.679346535033460E-017) 0.866025403784439
> Electronic Dipole per cell (a.u.) -0.482785724081663
> Ionic Dipole per cell (a.u.) 3794.73195575151
> PAY ATTENTION: EL FIELD AND OCCUPATIONS
>My input file adhere here:
>&CONTROL
> calculation = 'relax' ,
> restart_mode = 'from_scratch' ,
> pseudo_dir = './' ,
> prefix = 'ZnO' ,
> nstep = 250 ,
> tprnfor = .true. ,
> lelfield = .true. ,
> nberrycyc = 1 ,
> gdir = 3 ,
> /
> &SYSTEM
> ibrav = 14,
> celldm(1) = 18.6286,
> celldm(2) = 0.6666667,
> celldm(3) = 1.0733333,
> celldm(4) = 0,
> celldm(5) = 0,
> celldm(6) = -0.5,
> nat = 48,
> ntyp = 3,
> ecutwfc = 30 ,
> ecutrho = 300 ,
> nosym = .true. ,
> occupations = 'smearing' ,
> degauss = 0.008 ,
> smearing = 'gaussian' ,
> nspin = 2 ,
> starting_magnetization(1) = 0.5,
> starting_magnetization(2) = 0.5,
> starting_magnetization(3) = 0.5,
> lda_plus_u = .false. ,
> /
> &ELECTRONS
> conv_thr = 1.D-7 ,
> mixing_mode = 'plain' ,
> mixing_beta = 0.2 ,
> startingwfc = 'random' ,
> efield = 0 ,
> /
> &IONS
> ion_dynamics = 'bfgs' ,
> phase_space = 'full' ,
> /
>ATOMIC_SPECIES
> O 15.00000 O.pbe-rrkjus.UPF
> Zn 65.00000 Zn.pbe-van.UPF
> Mn 54.94000 Mn.pbe-sp-van.UPF
>ATOMIC_POSITIONS angstrom
> Zn 0.000000000 1.875966909 0.000000000 1 1 1
> O 0.000000000 1.875966909 1.991601344 1 1 1
>....................................................
>K_POINTS automatic
> 3 3 3 1 1 1
>I don't know why after I add the electric field ,the speed of relax is so slow ?
>Any advice will be appreciated !
>
>--
>
>Best regards
>
>Q.J.Wang
>
>XiangTan University
>-------------- next part --------------
>An HTML attachment was scrubbed...
>URL: http://www.democritos.it/pipermail/pw_forum/attachments/20091124/c751ca32/attachment.htm
>
>------------------------------
>
>_______________________________________________
>Pw_forum mailing list
>Pw_forum@pwscf.org
>http://www.democritos.it/mailman/listinfo/pw_forum
>
>
>End of Pw_forum Digest, Vol 29, Issue 68
>****************************************
>
>
>
>------------------------------
>
>Message: 3
>Date: Tue, 24 Nov 2009 17:39:39 -0000
>From: "Dimpy Sharma" <dimpy.sharma@tyndall.ie>
>Subject: [Pw_forum] Calculation of projected density of states for
>        individual atoms in chain
>To: <pw_forum@pwscf.org>
>Message-ID: <D4A761F73683694BBD69D634FF298015533C42@MAIL.tyndall.ie>
>Content-Type: text/plain; charset="iso-8859-1"
>
>
>Hi Quantum espresso users,
>
>I have calculated projected density of states (PDOS) for my system which contains 2 silicon atoms and 4 hydrogen atoms. Now I would like to calculate the PDOS for my individual atoms (i,e for Si1 , Si2 ) . I would like to know for calculating the PDOS for individual atom, which atomic coordinate shall I need to give in my input file , as I cannot give atomic coordinate for only one atom of Silicon ( as that would mean PDOS for single Silicon), what is the correct way of approaching ?
>
>Thanks a million
>
>Dimpy
>
>UCC
>
>Ireland
>-------------- next part --------------
>An HTML attachment was scrubbed...
>URL: http://www.democritos.it/pipermail/pw_forum/attachments/20091124/033345c5/attachment-0001.htm
>
>------------------------------
>
>Message: 4
>Date: Tue, 24 Nov 2009 18:48:42 +0100
>From: "Lorenzo Paulatto" <paulatto@sissa.it>
>Subject: Re: [Pw_forum] Calculation of projected density of states for
>        individual atoms in chain
>To: "PWSCF Forum" <pw_forum@pwscf.org>
>Message-ID: <op.u3wufgu1a8x26q@paulax>
>Content-Type: text/plain; charset=utf-8; format=flowed; delsp=yes
>
>Replying to message "[Pw_forum] Calculation of projected density of states
>for individual atoms in chain" from Dimpy Sharma (24/11/09):
>> I would like to know for calculating the PDOS for individual atom
>
>Dear Dimpy,
>pdos always computes the projected density of states for each individual
>atom.
>
>In the directory where you have executed projwfc.x you will have a bunch
>of file with names like this:
>prefix.pdos_atm#4(Pd)_wfc#3(p)
>
>This file does include projection on the 3rd wavefunction (which is of
>type p) belonging to the 4th atom (of type Pd). Inside the file itself you
>can find several columns that belong to the p_z, p_x and p_y orbitals and
>the total. It is all explained in Doc/INPUT_PROJWFC.txt, together with
>many more details.
>
>best 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/
>
>
>------------------------------
>
>Message: 5
>Date: Tue, 24 Nov 2009 20:04:28 +0100
>From: Jiri Houska <jhouska@kfy.zcu.cz>
>Subject: [Pw_forum] cp.x code
>To: pw_forum@pwscf.org
>Message-ID: <20091124200428.9ipx2zys00ws0wsc@webmail.zcu.cz>
>Content-Type: text/plain;        charset=ISO-8859-1;        DelSp="Yes";
>        format="flowed"
>
>Hello,
>please I am quite confused about the cp.x code which I an going to use:
>
>(I) is there a difference between (1) scf calculation using cp.x and
>(3) scf calculation using pw.x ?
>
>(II) is there a difference between (1) scf calculation using cp.x and
>(2) cp calculation using cp.x with electron_dynamics = (say) 'sd' and
>ion_dynamics = 'none' ?
>
>(III) how critical is the "ion_radius" choice ... e.g. is it clever to
>make it equal to (known) covalent radius?
>
>Thanks a lot,
>Jiri Houska
>University of West Bohemia
>
>
>
>
>------------------------------
>
>Message: 6
>Date: Tue, 24 Nov 2009 21:37:53 +0100
>From: Paolo Giannozzi <giannozz@democritos.it>
>Subject: Re: [Pw_forum] cp.x code
>To: PWSCF Forum <pw_forum@pwscf.org>
>Message-ID: <48987221-D34A-4B3C-95CB-D79ED30AC32A@democritos.it>
>Content-Type: text/plain; charset=US-ASCII; format=flowed
>
>
>On Nov 24, 2009, at 20:04 , Jiri Houska wrote:
>
>> (I) is there a difference between (1) scf calculation using cp.x and
>> (3) scf calculation using pw.x ?
>
>one is performed with damped (fictitious) dynamics on electrons,
>another with traditional self-consistency on the charge density
>
>> (II) is there a difference between (1) scf calculation using cp.x and
>> (2) cp calculation using cp.x with electron_dynamics = (say) 'sd' and
>> ion_dynamics = 'none' ?
>
>no, but steepest descent should be used only to start the calculation:
>it is very inefficient
>
>> (III) how critical is the "ion_radius" choice ... e.g. is it clever to
>> make it equal to (known) covalent radius?
>
>the defaults should be ok. The so called ion radius is used to broaden
>the point charges of the nuclei into a gaussian distribution of charges.
>It is a convenient computational trick.
>
>P.
>---
>Paolo Giannozzi, Dept of Physics, University of Udine
>via delle Scienze 208, 33100 Udine, Italy
>Phone +39-0432-558216, fax +39-0432-558222
>
>
>
>
>
>------------------------------
>
>Message: 7
>Date: Wed, 25 Nov 2009 08:52:43 +0600
>From: "Pavel A. Akulov" <apv544@phys.tsu.ru>
>Subject: [Pw_forum] Trouble with parallel running ld1.x
>To: pw_forum@pwscf.org
>Message-ID: <20091125025207.M9749@phys.tsu.ru>
>Content-Type: text/plain;        charset=koi8-r
>
>Dear PWSCF users and developers!
>
>I've faced a problem of parallel running ld1.x program.
>Stand-alone version works well (i've tried examples in atomic_doc),
>but parallel version gets me error:
>
>%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
>from ld1_readin : error # 1
>reading input namelist
>%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
>(even in examples).
>
>I run it with string(for example 2 procs h.in in /atomic_doc/all-electron)
>"mpirun -np 2 /path/ld1.x -in h.in > h.out"
>
>For pw.x this works well. But for ld1.x i don't know what's wrong.
>
>Help me please.
>
>
>--
>Akulov P.
>Student, Dept. of Physics
>Tomsk State University
>
>
>
>------------------------------
>
>_______________________________________________
>Pw_forum mailing list
>Pw_forum@pwscf.org
>http://www.democritos.it/mailman/listinfo/pw_forum
>
>
>End of Pw_forum Digest, Vol 29, Issue 69
>****************************************
</PRE>