From brad.malone@gmail.com Wed May 2 20:42:32 2007 From: brad.malone@gmail.com (Brad Malone) Date: Wed, 2 May 2007 12:42:32 -0700 Subject: [Wannier] Accuracy of fermi surface plot Message-ID: ------=_Part_11364_16378960.1178134952556 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Content-Disposition: inline Hi, I was wondering how accurate the eigenvalues given in the .BXSF file when outputting the fermi surface are. Does increasing the number of iterations (num_iter) increase the accuracy, or does this number only affect the Wannierization? Thanks, Brad ------=_Part_11364_16378960.1178134952556 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline Hi, I was wondering how accurate the eigenvalues given in the .BXSF file when outputting the fermi surface are.  Does increasing the number  of iterations (num_iter) increase the accuracy, or does this number only affect the Wannierization?


Thanks,
Brad
------=_Part_11364_16378960.1178134952556-- From jry20@cam.ac.uk Wed May 2 21:22:21 2007 From: jry20@cam.ac.uk (Jonathan Yates) Date: Wed, 2 May 2007 21:22:21 +0100 (BST) Subject: [Wannier] Accuracy of fermi surface plot In-Reply-To: References: Message-ID: On Wed, 2 May 2007, Brad Malone wrote: > Hi, I was wondering how accurate the eigenvalues given in the .BXSF file > when outputting the fermi surface are. Does increasing the number of > iterations (num_iter) increase the accuracy, or does this number only affect > the Wannierization? Brad, num_iter sets the maximum number of steps to take when minimising the spread of the Wannier functions. If you set this to a small number you may find that the spread of the WF is converged to only, say, 1x10-6 Ang^2, rather than machine precision. However, this should have no significant effect on the quality of the interpolated eigenvalues.[*] The Wannier interpolation procedure requires having well localised WF, rather than maximally localised WF (ie it doesn't matter if you're not at the exact minimum of the spread). The quality of the interpolated eigenvalues is determined by the spacing of the ab-initio k-point mesh used to compute the WF. In fact the quality of the interpolation increases exponentially with the spacing of this mesh.(denser k-grid => better interpolation). See cond-mat/0702554 particularly the Appendix (this will appear in PRB shortly) In order for you to assess the quality of the interpolation I suggest the following procedure: Plot a band structure along various high symmetry directions using Wannier90. Plot an equivalent band-structure using your ab-initio code. Note: if you're using pwscf, Wannier90 will write a file _band.kpt, which contains the path kpoints in pwscf format. Compare the two band-structures. Increase the density of the k-point mesh used to compute the WF until the interpolated bands have an accuracy sufficient for your needs. Yours Jonathan [*] Of course this is a simple thing to check out numerically - so you don't have to take my word for it. -- Dr Jonathan Yates | Theory of Condensed Matter Group Corpus Christi College | Cavendish Laboratory Cambridge, CB2 1RH, UK | Cambridge, CB3 OHE, UK email jry20@cam.ac.uk | Tel +44 (0)1223 337461 From ogitsu@llnl.gov Sat May 5 01:34:55 2007 From: ogitsu@llnl.gov (Tadashi Ogitsu) Date: Fri, 4 May 2007 17:34:55 -0700 Subject: [Wannier] a little problem with pwscf 3.2.1? In-Reply-To: References: Message-ID: Hi Wannier90s, I've found a small problem with, I think, pw2wannier90 came with pwscf3.2.1 (3.2 plus the distributed patch on their wabpage). Somewhat they have changed the output format of .eig file so that the last data, the weight of each eigenvalues, does not exist anymore, while the current distribution of wannier90 try to read and complains that the data is not consistent. I just removed the last read as follows, old: read(eig_unit,*,err=106,end=106) i,j,eigval(n,k),wg(n,k) new: read(eig_unit,*,err=106,end=106) i,j,eigval(n,k) First of all, is my understanding correct? (i.e., this is a problem caused by the change in pw2wannier90 in 3.2.1 distribution) Second, did I do the right correction? (I have no idea if wg is used anywhere, although it does not seem to be...) Thanking in advance! Tadashi PS. Jonathan, I've started to work on the gold transport and as the first try, I'm playing with the example6. From mostofi@MIT.EDU Sat May 5 03:18:43 2007 From: mostofi@MIT.EDU (Arash A Mostofi) Date: Fri, 4 May 2007 22:18:43 -0400 (EDT) Subject: [Wannier] a little problem with pwscf 3.2.1? In-Reply-To: References: Message-ID: Dear Tadashi I'm slightly confused: the eigenvalues have always (ie, since the original incarnation of pw2wannier90 in December 2005) been written by pw2wannier90 as: write (iun_band,'(2i5,f18.12)') ibnd1, ikevc, et(ibnd,ik)*rytoev (ie, without any reference to weights, wg). This is easily verified by checking the espresso cvs repository. Consequently, Wannier90 has always read the eigenvalues as: read(eig_unit,*,err=106,end=106) i,j,eigval(n,k) as I've verified by checking our Wannier90 cvs repository. Can you remember from where you got your copy of the Wannier90 source? Was it from the www.wannier.org or from someone who may have made some modifications for his/her own purposes? In any case, the change you've made is correct. Best regards Arash ----------------------------------------- Dr Arash A Mostofi Dept of Materials Science and Engineering 13-4084 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA Tel: 617.2536026 ----------------------------------------- On Fri, 4 May 2007, Tadashi Ogitsu wrote: > Hi Wannier90s, > > I've found a small problem with, I think, pw2wannier90 came with > pwscf3.2.1 (3.2 plus the distributed patch on their wabpage). > > Somewhat they have changed the output format of .eig file so that the > last data, the weight of each eigenvalues, does not exist anymore, while > the current distribution of wannier90 try to read and complains that the > data is not consistent. > > I just removed the last read as follows, > > old: > read(eig_unit,*,err=106,end=106) i,j,eigval(n,k),wg(n,k) > new: > read(eig_unit,*,err=106,end=106) i,j,eigval(n,k) > > First of all, is my understanding correct? (i.e., this is a problem > caused by the change in pw2wannier90 in 3.2.1 distribution) > > Second, did I do the right correction? (I have no idea if wg is used > anywhere, although it does not seem to be...) > > Thanking in advance! > > Tadashi > PS. Jonathan, I've started to work on the gold transport and as the first > try, I'm playing with the example6. > _______________________________________________ > Wannier mailing list > Wannier@quantum-espresso.org > http://www.democritos.it/mailman/listinfo/wannier From ogitsu@llnl.gov Mon May 7 19:13:44 2007 From: ogitsu@llnl.gov (Tadashi Ogitsu) Date: Mon, 7 May 2007 11:13:44 -0700 Subject: [Wannier] a little problem with pwscf 3.2.1? In-Reply-To: References: Message-ID: <34D087AF-F363-4BC6-8A54-92218C68A490@llnl.gov> Hi Arash, Thanks for the response. Well, then, I got a non-official version probably...? I thought, I downloaded from the webpage, but I cannot be 100% sure. I'll download the official distribution and replace with the one I have now just in case. Sorry about the confusion. Tadashi On May 4, 2007, at 7:18 PM, Arash A Mostofi wrote: > Dear Tadashi > > I'm slightly confused: the eigenvalues have always (ie, since the > original incarnation of pw2wannier90 in December 2005) been written > by pw2wannier90 as: > > write (iun_band,'(2i5,f18.12)') ibnd1, ikevc, et(ibnd,ik)*rytoev > > (ie, without any reference to weights, wg). This is easily verified > by checking the espresso cvs repository. > > Consequently, Wannier90 has always read the eigenvalues as: > > read(eig_unit,*,err=106,end=106) i,j,eigval(n,k) > > as I've verified by checking our Wannier90 cvs repository. > > Can you remember from where you got your copy of the Wannier90 > source? Was it from the www.wannier.org or from someone who may > have made some modifications for his/her own purposes? In any case, > the change you've made is correct. > > Best regards > > Arash > > ----------------------------------------- > Dr Arash A Mostofi > Dept of Materials Science and Engineering > 13-4084 MIT > 77 Massachusetts Avenue > Cambridge MA 02139-4307 USA > Tel: 617.2536026 > ----------------------------------------- > > On Fri, 4 May 2007, Tadashi Ogitsu wrote: > >> Hi Wannier90s, >> >> I've found a small problem with, I think, pw2wannier90 came with >> pwscf3.2.1 (3.2 plus the distributed patch on their wabpage). >> >> Somewhat they have changed the output format of .eig file so that >> the last data, the weight of each eigenvalues, does not exist >> anymore, while the current distribution of wannier90 try to read >> and complains that the data is not consistent. >> >> I just removed the last read as follows, >> >> old: >> read(eig_unit,*,err=106,end=106) i,j,eigval(n,k),wg(n,k) >> new: >> read(eig_unit,*,err=106,end=106) i,j,eigval(n,k) >> >> First of all, is my understanding correct? (i.e., this is a >> problem caused by the change in pw2wannier90 in 3.2.1 distribution) >> >> Second, did I do the right correction? (I have no idea if wg is >> used anywhere, although it does not seem to be...) >> >> Thanking in advance! >> >> Tadashi >> PS. Jonathan, I've started to work on the gold transport and as >> the first try, I'm playing with the example6. >> _______________________________________________ >> Wannier mailing list >> Wannier@quantum-espresso.org >> http://www.democritos.it/mailman/listinfo/wannier > _______________________________________________ > Wannier mailing list > Wannier@quantum-espresso.org > http://www.democritos.it/mailman/listinfo/wannier From joseph99077@yahoo.com.tw Tue May 22 15:10:12 2007 From: joseph99077@yahoo.com.tw (joseph99077@yahoo.com.tw) Date: Tue, 22 May 2007 22:10:12 +0800 (CST) Subject: [Wannier] about Amn Message-ID: <846698.46490.qm@web72301.mail.tp2.yahoo.com> --0-1740024784-1179843012=:46490 Content-Type: text/plain; charset=big5 Content-Transfer-Encoding: 8bit hello I got a question about the matrix Amn. in the pwscf package, we can generate Mmn and Amn. there is a term in the source code pw2wannier90.f90 around line 1090 , amn = ZDOTC(npw,evc(1,ibnd),1,sgf(1,iw),1). Can you please explain more about what kind of role do evc and sgf play in this statement. From the user guide, I would say it's the inner product of Bloch states and trail functions. so, is sgf the trail function? ( I hope this question is not beyond the scope of the website.). I hope that you can tell me how to generate wannier functions from the wannier package which you provided. have a good day. 杜絕網路駭客，保障帳號安全 - 馬上設定 Yahoo!奇摩安全圖章！ --0-1740024784-1179843012=:46490 Content-Type: text/html; charset=big5 Content-Transfer-Encoding: 8bit hello
       I got a question about the matrix Amn.  in the pwscf package, we can generate Mmn and Amn. there is a term in the source code pw2wannier90.f90 around line 1090 , amn = ZDOTC(npw,evc(1,ibnd),1,sgf(1,iw),1). Can you please explain more about what kind of role do evc and sgf play in this statement.  From the user guide, I would say it's the inner product of Bloch states and trail functions. so, is sgf the trail function? ( I hope this question is not beyond the scope of the website.). 
       I hope that you can tell me how to generate wannier functions from the wannier package which you provided. 
      have a good day.


杜絕網路駭客，保障帳號安全 - 馬上設定 Yahoo!奇摩安全圖章！ --0-1740024784-1179843012=:46490-- From degironc@sissa.it Wed May 23 10:10:34 2007 From: degironc@sissa.it (Stefano de Gironcoli) Date: Wed, 23 May 2007 11:10:34 +0200 (CEST) Subject: [Wannier] about Amn In-Reply-To: <846698.46490.qm@web72301.mail.tp2.yahoo.com> References: <846698.46490.qm@web72301.mail.tp2.yahoo.com> Message-ID: This message is in MIME format. The first part should be readable text, while the remaining parts are likely unreadable without MIME-aware tools. --2056485387-1901079391-1179911434=:2226 Content-Type: TEXT/PLAIN; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 8BIT in PWscf evc is the array containing the eigenvectors (the complex k+G Fourier coefficients of them); it's a global array used all around the code. sgf is a local array created in the tool that builds the Amn matrix that contains the S matrix (for US pseudopotential, the identity matrix for NCPP) times the guiding functions (hence the name). hope it helps. Stefano de Gironcoli On Tue, 22 May 2007, joseph99077@yahoo.com.tw wrote: > hello > I got a question about the matrix Amn. in the pwscf package, we can generate Mmn and Amn. there is a term in the source code pw2wannier90.f90 around line 1090 , amn = ZDOTC(npw,evc(1,ibnd),1,sgf(1,iw),1). Can you please explain more about what kind of role do evc and sgf play in this statement. From the user guide, I would say it's the inner product of Bloch states and trail functions. so, is sgf the trail function? ( I hope this question is not beyond the scope of the website.). > I hope that you can tell me how to generate wannier functions from the wannier package which you provided. > have a good day. > > > > 杜絕網路駭客，保障帳號安全 - 馬上設定 Yahoo!奇摩安全圖章！ --2056485387-1901079391-1179911434=:2226-- From brad.malone@gmail.com Fri May 25 20:58:31 2007 From: brad.malone@gmail.com (Brad Malone) Date: Fri, 25 May 2007 12:58:31 -0700 Subject: [Wannier] Too many projections defined error Message-ID: ------=_Part_139432_23760487.1180123111998 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Content-Disposition: inline Hi, I'm having an issue that I'm very confused about. I'm running a calculation on silicon with 12 atoms per unit cell. There is a gap in my system (pwSCF tells me that 7.3 is the highest occupied state and 8.4 is the lowest unoccupied state). I just care about the valence bands and so I set my upper energy window to 7.8 eV and set num_wann=24 (num_bands=32). My frozen energy window is the same as my larger window and it covers all 24 valence bands. However, when I run the wannier code to generate the .nnkp file it tells me the following error: param_get_projection: too many projections defined Any help in understanding why this is would be greatly appreciated! Best, Brad ------=_Part_139432_23760487.1180123111998 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline Hi, I'm having an issue that I'm very confused about.  I'm running a calculation on silicon with 12 atoms per unit cell.  There is a gap in my system (pwSCF tells me that 7.3 is the highest occupied state and 8.4 is the lowest unoccupied state).  I just care about the valence bands and so I set my upper energy window to 7.8 eV and set num_wann=24 (num_bands=32).  My frozen energy window is the same as my larger window and it covers all 24 valence bands.  However, when I run the wannier code to generate the .nnkp file it tells me the following error:

 param_get_projection: too many projections defined


Any help in understanding why this is would be greatly appreciated!

Best,
Brad


------=_Part_139432_23760487.1180123111998-- From mostofi@MIT.EDU Fri May 25 21:22:28 2007 From: mostofi@MIT.EDU (Arash A Mostofi) Date: Fri, 25 May 2007 16:22:28 -0400 (EDT) Subject: [Wannier] Too many projections defined error In-Reply-To: References: Message-ID: Dear Brad I think the problem is in the "projections" block of the input file, not the energy windows. As you say, a 12 atom unit cell of silicon will need 24 maximally localised Wannier functions (MLWFs) to span the valence manifold and this is the number of projections you should define in the projections block. What we've found is that the generation of MLWFs for the valence manifold is very robust to the initial guess, so even specifying randomly centred s-orbital projections is good enough. This can be done by setting Begin Projections random End Projections in the input file. Cheers Arash ----------------------------------------- Dr Arash A Mostofi Dept of Materials Science and Engineering 13-4084 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA Tel: 617.2536026 ----------------------------------------- On Fri, 25 May 2007, Brad Malone wrote: > Hi, I'm having an issue that I'm very confused about. I'm running a > calculation on silicon with 12 atoms per unit cell. There is a gap in my > system (pwSCF tells me that 7.3 is the highest occupied state and 8.4 is the > lowest unoccupied state). I just care about the valence bands and so I set > my upper energy window to 7.8 eV and set num_wann=24 (num_bands=32). My > frozen energy window is the same as my larger window and it covers all 24 > valence bands. However, when I run the wannier code to generate the .nnkp > file it tells me the following error: > > param_get_projection: too many projections defined > > > > Any help in understanding why this is would be greatly appreciated! > > Best, > Brad > From brad.malone@gmail.com Sat May 26 19:57:30 2007 From: brad.malone@gmail.com (Brad Malone) Date: Sat, 26 May 2007 11:57:30 -0700 Subject: [Wannier] Re: Too many projections defined error Message-ID: ------=_Part_147304_20622608.1180205850595 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Content-Disposition: inline Thanks for the response Arash. Setting the projections to random (they were set to sp3) did indeed fix that issue. I ran the wannier code and generated the .nnkp file. However, in the nnkp file it lists my excluded bands as being 25-32 (which I had specified in the .win input file; perhaps unnecessary because my window excludes them anyway) but also it lists band #8 as being excluded, and I did not want or specify that. Is there a reason for this? I'm worried that it may be causing me an error when I run the Wannier code again because I get an error of "param_read: mismatch in silicon.eig". Inspecting the silicon.eig file it looks like it should be okay, unless the code thinks that there are 23 bands (having excluded band 8), as there are 24 bands (for my 64 points) listed in the silicon.eig file. I'm also interested in the bottom of the conduction band. Do you have a feel for how robust this is to the initial guess for the projections? And is it possible for me to specify sp3 projections without having 48 wannier bands? I'd like to use fewer bands if possible because it seems that I'm running into memory allocation issues that prevent me from starting with a denser k-grid. Thanks again for the help. Brad > Message: 1 > Date: Fri, 25 May 2007 12:58:31 -0700 > From: "Brad Malone" > To: wannier@quantum-espresso.org > Subject: [Wannier] Too many projections defined error > > ------=_Part_139432_23760487.1180123111998 > Content-Type: text/plain; charset=ISO-8859-1; format=flowed > Content-Transfer-Encoding: 7bit > Content-Disposition: inline > > Hi, I'm having an issue that I'm very confused about. I'm running a > calculation on silicon with 12 atoms per unit cell. There is a gap in my > system (pwSCF tells me that 7.3 is the highest occupied state and 8.4 is > the > lowest unoccupied state). I just care about the valence bands and so I > set > my upper energy window to 7.8 eV and set num_wann=24 (num_bands=32). My > frozen energy window is the same as my larger window and it covers all 24 > valence bands. However, when I run the wannier code to generate the .nnkp > file it tells me the following error: > > param_get_projection: too many projections defined > > > > Any help in understanding why this is would be greatly appreciated! > > Best, > Brad > > ------=_Part_139432_23760487.1180123111998 > Content-Type: text/html; charset=ISO-8859-1 > Content-Transfer-Encoding: 7bit > Content-Disposition: inline > > Hi, I'm having an issue that I'm very confused about.  > I'm running a calculation on silicon with 12 atoms per unit cell.  > There is a gap in my system (pwSCF tells me that 7.3 is the highest > occupied state and 8.4 > is the lowest unoccupied state).  I just care about the valence bands > and so I set my upper energy window to 7.8 eV and set num_wann=24 > (num_bands=32).  My frozen energy window is the same as my larger > window and it covers all 24 valence bands.  However, when I run the > wannier code to generate the .nnkp file it tells me the following error: >

class="gmail_quote"> param_get_projection: too many projections > defined


Any help in understanding why this is would > be greatly appreciated! >

Best,
Brad


> > ------=_Part_139432_23760487.1180123111998-- > > --__--__-- > > Message: 2 > Date: Fri, 25 May 2007 16:22:28 -0400 (EDT) > From: Arash A Mostofi > To: Wannier List > Subject: Re: [Wannier] Too many projections defined error > > Dear Brad > > I think the problem is in the "projections" block of the input file, not > the energy windows. As you say, a 12 atom unit cell of silicon will need > 24 > maximally localised Wannier functions (MLWFs) to span the valence manifold > and this is the number of projections you should define in the projections > block. > > What we've found is that the generation of MLWFs for the valence manifold > is very robust to the initial guess, so even specifying randomly centred > s-orbital projections is good enough. This can be done by setting > > Begin Projections > random > End Projections > > in the input file. > > Cheers > > Arash > > ----------------------------------------- > Dr Arash A Mostofi > Dept of Materials Science and Engineering > 13-4084 MIT > 77 Massachusetts Avenue > Cambridge MA 02139-4307 USA > Tel: 617.2536026 > ----------------------------------------- > > > ------=_Part_147304_20622608.1180205850595 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline Thanks for the response Arash.  Setting the projections to random (they were set to sp3) did indeed fix that issue.  I ran the wannier code and generated the .nnkp file.  However, in the nnkp file it lists my excluded bands as being 25-32 (which I had specified in the .win input file; perhaps unnecessary because my window excludes them anyway) but also it lists band #8 as being excluded, and I did not want or specify that.  Is there a reason for this?  I'm worried that it may be causing me an error when I run the Wannier code again because I get an error of "param_read: mismatch in silicon.eig".  Inspecting the silicon.eig file it looks like it should be okay, unless the code thinks that there are 23 bands (having excluded band 8), as there are 24 bands (for my 64 points) listed in the silicon.eig file.

I'm also interested in the bottom of the conduction band.  Do you have a feel for how robust this is to the initial guess for the projections?  And is it possible for me to specify sp3 projections without having 48 wannier bands?  I'd like to use fewer bands if possible because it seems that I'm running into memory allocation issues that prevent me from starting with a denser k-grid. Thanks again for the help.

Brad



Message: 1
Date: Fri, 25 May 2007 12:58:31 -0700
From: "Brad Malone" <brad.malone@gmail.com>
To: wannier@quantum-espresso.org
Subject: [Wannier] Too many projections defined error

------=_Part_139432_23760487.1180123111998
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

Hi, I'm having an issue that I'm very confused about.  I'm running a
calculation on silicon with 12 atoms per unit cell.  There is a gap in my
system (pwSCF tells me that 7.3 is the highest occupied state and 8.4 is the
lowest unoccupied state).  I just care about the valence bands and so I set
my upper energy window to 7.8 eV and set num_wann=24 (num_bands=32).  My
frozen energy window is the same as my larger window and it covers all 24
valence bands.  However, when I run the wannier code to generate the .nnkp
file it tells me the following error:

param_get_projection: too many projections defined



Any help in understanding why this is would be greatly appreciated!

Best,
Brad

------=_Part_139432_23760487.1180123111998
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

Hi, I&#39;m having an issue that I&#39;m very confused about.&nbsp; I&#39;m running a calculation on silicon with 12 atoms per unit cell.&nbsp; There is a gap in my system (pwSCF tells me that 7.3 is the highest occupied state and 8.4
is the lowest unoccupied state).&nbsp; I just care about the valence bands and so I set my upper energy window to 7.8 eV and set num_wann=24 (num_bands=32).&nbsp; My frozen energy window is the same as my larger window and it covers all 24 valence bands.&nbsp; However, when I run the wannier code to generate the .nnkp file it tells me the following error:
<br><br><blockquote style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;" class="gmail_quote">&nbsp;param_get_projection: too many projections defined</blockquote><div><br><br>Any help in understanding why this is would be greatly appreciated!
<br><br>Best,<br>Brad<br></div><br><br>

------=_Part_139432_23760487.1180123111998--

--__--__--

Message: 2
Date: Fri, 25 May 2007 16:22:28 -0400 (EDT)
From: Arash A Mostofi <mostofi@MIT.EDU>
To: Wannier List <wannier@quantum-espresso.org>
Subject: Re: [Wannier] Too many projections defined error

Dear Brad

I think the problem is in the "projections" block of the input file, not
the energy windows. As you say, a 12 atom unit cell of silicon will need 24
maximally localised Wannier functions (MLWFs) to span the valence manifold
and this is the number of projections you should define in the projections
block.

What we've found is that the generation of MLWFs for the valence manifold
is very robust to the initial guess, so even specifying randomly centred
s-orbital projections is good enough. This can be done by setting

Begin Projections
random
End Projections

in the input file.

Cheers

Arash

-----------------------------------------
Dr Arash A Mostofi
Dept of Materials Science and Engineering
13-4084 MIT
77 Massachusetts Avenue
Cambridge MA 02139-4307 USA
Tel: 617.2536026
-----------------------------------------



------=_Part_147304_20622608.1180205850595-- From brad.malone@gmail.com Sun May 27 01:34:57 2007 From: brad.malone@gmail.com (Brad Malone) Date: Sat, 26 May 2007 17:34:57 -0700 Subject: [Wannier] Re: Too many projections defined error Message-ID: ------=_Part_149621_19952660.1180226097234 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Content-Disposition: inline Aha, I see now that it simply first lists the number of excluded bands before listing the specific numbers. But I'm still wondering about why I'm getting the error in the .eig file. Thanks! ------=_Part_149621_19952660.1180226097234 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline Aha, I see now that it simply first lists the number of excluded bands before listing the specific numbers.  But I'm still wondering about why I'm getting the error in the .eig file.  Thanks!
------=_Part_149621_19952660.1180226097234-- From brad.malone@gmail.com Sun May 27 20:04:27 2007 From: brad.malone@gmail.com (Brad Malone) Date: Sun, 27 May 2007 12:04:27 -0700 Subject: [Wannier] Re: Too many projections defined error Message-ID: ------=_Part_128012_21952737.1180292667582 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Content-Disposition: inline Omitting the 'exclude bands' portion in my input file apparently fixes my issue with the .eig file. I no longer get the error. I don't know if what I described before is the reason for that or not, but it works now. Best, Brad ------=_Part_128012_21952737.1180292667582 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline Omitting the 'exclude bands' portion in my input file apparently fixes my issue with the .eig file.  I no longer get the error.  I don't know if what I described before is the reason for that or not, but it works now.

Best,
Brad
------=_Part_128012_21952737.1180292667582-- From lanhaiping@gmail.com Tue May 29 10:08:57 2007 From: lanhaiping@gmail.com (lan haiping) Date: Tue, 29 May 2007 17:08:57 +0800 Subject: [Wannier] about WFs' display Message-ID: ------=_Part_184085_8651585.1180429737382 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Content-Disposition: inline Dear All, I am trying to display WFs with XCRSDEN, while there are some problems i cannot solve . Attached is one of my results, the wf displayed is not full. Would you please tell me how to obtain a full WFs display ? Are there some tricks to generate WFs ? For my calculation , the wout file gave that Im/Re==0, So, i confirmed that there is no problem related with wannierise procedure. In addition , the interpolated bands are also consistent with PW results. By the way, is there any way to generate multi-WFs in a single xsf file ? Regards, H.P -- Hai-Ping Lan Department of Electronics , Peking University , Bejing, 100871 lanhaiping@gmail.com, hplan@pku.edu.cn ------=_Part_184085_8651585.1180429737382 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline Dear All,

   I am trying to display WFs with XCRSDEN, while there are some problems i cannot solve .
Attached is one of my results,  the wf displayed is not full.  Would you please tell me how to obtain
a full WFs display ? Are there some tricks to generate WFs ?  For my calculation , the wout file gave that  Im/Re==0,
So, i confirmed that  there is no problem related with wannierise procedure. In addition , the interpolated bands are also
consistent with PW results.

By the way, is there any way to generate multi-WFs in a single xsf file ?

Regards,

H.P


--
Hai-Ping Lan
Department of Electronics ,
Peking University , Bejing, 100871
lanhaiping@gmail.com, hplan@pku.edu.cn ------=_Part_184085_8651585.1180429737382-- From mostofi@MIT.EDU Tue May 29 16:26:06 2007 From: mostofi@MIT.EDU (Arash A Mostofi) Date: Tue, 29 May 2007 11:26:06 -0400 (EDT) Subject: [Wannier] about WFs' display In-Reply-To: References: Message-ID: Dear Ian I don't think your attachment made it through to my inbox, but the problem you describe may be resolved by setting wannier_plot_supercell = 3 in the Wannier90 input file. Cheers Arash ----------------------------------------- Dr Arash A Mostofi Dept of Materials Science and Engineering 13-4084 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA Tel: 617.2536026 ----------------------------------------- On Tue, 29 May 2007, lan haiping wrote: > Dear All, > > I am trying to display WFs with XCRSDEN, while there are some problems i > cannot solve . > Attached is one of my results, the wf displayed is not full. Would you > please tell me how to obtain > a full WFs display ? Are there some tricks to generate WFs ? For my > calculation , the wout file gave that Im/Re==0, > So, i confirmed that there is no problem related with wannierise procedure. > In addition , the interpolated bands are also > consistent with PW results. > > By the way, is there any way to generate multi-WFs in a single xsf file ? > > Regards, > > H.P > > > -- > Hai-Ping Lan > Department of Electronics , > Peking University , Bejing, 100871 > lanhaiping@gmail.com, hplan@pku.edu.cn > From jry20@cam.ac.uk Tue May 29 19:13:28 2007 From: jry20@cam.ac.uk (Jonathan Yates) Date: Tue, 29 May 2007 19:13:28 +0100 (BST) Subject: [Wannier] about WFs' display In-Reply-To: References: Message-ID: On Tue, 29 May 2007, Arash A Mostofi wrote: > I don't think your attachment made it through to my inbox, but the problem > you describe may be resolved by setting > > wannier_plot_supercell = 3 This is something of a FAQ, so maybe I can expand on the reasons behind Arash's answer. Defns : WF=Wannier Function : BZ=Brillouin zone A WF is not periodic in the unit cell. Therefore it is not generally sufficient to plot it in a single unit cell. wannier_plot_supercell controls the number of units cell in which we construct the WF. For a value of 3 this means the home unit cell, and the 26 neighbouring unit cells. This is normally sufficient to give a good picture of a WF. Plotting can be an expensive operation. You may wish to plot only a subset of the WF. You can control which WF are plotted using the wannier_plot_list keyword. The plotting of WF could be much improved. The next version of wannier90 (due over the summer) will have improvements in this area. Another, related technical issue is that when the WF are computed from a finite sampling of the BZ they will be periodic in a supercell of the unit cell. For example a WF computed from a 4x4x4 sampling of the BZ will 'live' in 4x4x4 supercell of the unit cell. This is apparent in Example 1 of Wannier90. If you plot WF in GaAs from a 2x2x2 BZ sampling using wannier_plot_supercell = 3 you will see the periodic images of the WF. >> By the way, is there any way to generate multi-WFs in a single xsf file ? No. It would be trivial to make this modification. However, I don't think it would be very useful. As I understand it Xcrysden is able to plot a single isosurface, which maybe a weighted sum of several scalar fields, but not multiple isosurfaces (ie several WF at once). So putting all the WF in a single file would make Xcrysden very slow to start (it parses the entire xsf file on startup) to no real advantage. Jonathan -- Dr Jonathan Yates | Theory of Condensed Matter Group Corpus Christi College | Cavendish Laboratory Cambridge, CB2 1RH, UK | Cambridge, CB3 OHE, UK email jry20@cam.ac.uk | Tel +44 (0)1223 337461