From lanhaiping at gmail.com Thu Mar 1 10:17:40 2007 From: lanhaiping at gmail.com (lan haiping) Date: Thu, 1 Mar 2007 17:17:40 +0800 Subject: [Wannier] the site cannot be connected Message-ID: Dear All, I just wanna check the new version of wannier package, but i cannot connect the site .. Is it down ,or others ? Regards h.p -- Hai-Ping Lan Department of Electronics , Peking University , Bejing, 100871 -------------- next part -------------- An HTML attachment was scrubbed... URL: From jry20 at cam.ac.uk Thu Mar 1 12:01:57 2007 From: jry20 at cam.ac.uk (Jonathan Yates) Date: Thu, 1 Mar 2007 11:01:57 +0000 (GMT) Subject: [Wannier] the site cannot be connected In-Reply-To: References: Message-ID: On Thu, 1 Mar 2007, lan haiping wrote: > I just wanna check the new version of wannier package, but i cannot connect > the site .. > Is it down ,or others ? Hi, I can see it fine from here (UK 11hrs GMT), so if you still can't see it wait a few hours and try again - it probably is/was a temporary network issue. Jonathan [the machine that hosts the website is physically on US Eastern time.] -- Dr Jonathan Yates | Theory of Condensed Matter Group Corpus Christi College | Cavendish Laboratory Cambridge, CB2 1RH, UK | Cambridge, CB3 OHE, UK email jry20 at cam.ac.uk | Tel +44 (0)1223 337461 From brad.malone at gmail.com Tue Mar 20 18:49:40 2007 From: brad.malone at gmail.com (Brad Malone) Date: Tue, 20 Mar 2007 10:49:40 -0700 Subject: [Wannier] Outputting the eigenvalues for a given k-grid Message-ID: Hi, I am trying to use the wannier code in conjunction with pwSCF to get the eigenvalues on a very fine k-grid. Is there some way within wannier to get it to output the eigenvalues on a finer k-grid than what was used in the pwSCF run? Thanks for the help, I appreciate it. Best, Brad -------------- next part -------------- An HTML attachment was scrubbed... URL: From jry20 at cam.ac.uk Tue Mar 20 19:32:45 2007 From: jry20 at cam.ac.uk (Jonathan Yates) Date: Tue, 20 Mar 2007 18:32:45 +0000 (GMT) Subject: [Wannier] Outputting the eigenvalues for a given k-grid In-Reply-To: References: Message-ID: On Tue, 20 Mar 2007, Brad Malone wrote: > Hi, I am trying to use the wannier code in conjunction with pwSCF to get the > eigenvalues on a very fine k-grid. Is there some way within wannier to get > it to output the eigenvalues on a finer k-grid than what was used in the > pwSCF run? Brad, Indeed you can, albeit indirectly. The general procedure is called Wannier interpolation (see http://www.arxiv.org/abs/cond-mat/0702554 for more details) You can use the Fermi surface plotting routine to get the information you want. Set fermi_surface_plot : True This will write out the eigenvalues on a NxNxN grid in the reciprocal unit cell (where N is set by fermi_surface_num_points (default = 50)). This data, along with some headers is written into the *.bxsf file. If you want the data to higher precision, or in a different order, you will have to modify the code. This should be fairly simple. Look for the routine plot_fermi_surface in plot.F90. Any questions, just ask. Yours Jonathan -- Dr Jonathan Yates | Theory of Condensed Matter Group Corpus Christi College | Cavendish Laboratory Cambridge, CB2 1RH, UK | Cambridge, CB3 OHE, UK email jry20 at cam.ac.uk | Tel +44 (0)1223 337461 From brad.malone at gmail.com Wed Mar 21 00:48:01 2007 From: brad.malone at gmail.com (Brad Malone) Date: Tue, 20 Mar 2007 16:48:01 -0700 Subject: [Wannier] Re: Outputting the eigenvalues for a given k-grid In-Reply-To: References: Message-ID: Awesome! I should have read the user guide more carefully. Thanks again Jonathan. Brad On 3/20/07, Brad Malone wrote: > > Hi, I am trying to use the wannier code in conjunction with pwSCF to get > the eigenvalues on a very fine k-grid. Is there some way within wannier to > get it to output the eigenvalues on a finer k-grid than what was used in the > pwSCF run? > > Thanks for the help, I appreciate it. > > Best, > Brad > -------------- next part -------------- An HTML attachment was scrubbed... URL: From brad.malone at gmail.com Sat Mar 24 23:28:26 2007 From: brad.malone at gmail.com (Brad Malone) Date: Sat, 24 Mar 2007 15:28:26 -0700 Subject: [Wannier] Error in plotting fermi surface Message-ID: Hi, I'm getting an error while trying to plot the fermi surface. It says "Error in allocating eig_int in plot_fermi_surface". I'm trying to generate a 100x100x100 grid, and I didn't get this error for 50x50x50. Is this some issue with memory or something? And is there a way for me to get around it? Thanks for the help! Brad -------------- next part -------------- An HTML attachment was scrubbed... URL: From jry20 at cam.ac.uk Sat Mar 24 23:59:21 2007 From: jry20 at cam.ac.uk (Jonathan Yates) Date: Sat, 24 Mar 2007 22:59:21 +0000 (GMT) Subject: [Wannier] Error in plotting fermi surface In-Reply-To: References: Message-ID: On Sat, 24 Mar 2007, Brad Malone wrote: > Hi, I'm getting an error while trying to plot the fermi surface. It says > "Error in allocating eig_int in plot_fermi_surface". I'm trying to generate > a 100x100x100 grid, and I didn't get this error for 50x50x50. Is this some > issue with memory or something? And is there a way for me to get around it? Brad, Failure to allocate an array generally means you've run out of RAM. In this case the array eig_int will be rather large; you can estimate its size in Mb with fermi_surface_num_points**3 x num_wann x 8 / (1024^2) Clearly there are a lot of other arrays taking up memory, but if eig_int is comparable, or much larger than the RAM of the your machine, you're going to be in trouble. A trivial solution is to find a machine with more RAM! The more elegant solution is to modify the code. The reason why the routine tries to store all the eigenvalues in memory is just so it can re-arrange the data into a form XCrysden can read. I seem to remember you're interested in the eigenvalues, rather than getting nice Fermi surface plot. In that case you could modify the plot_fermi_surface routine to write the eigenvalues out to a file as you go along, rather than storing them. I think this should be a trivial modification to make. (send a mail to myself and Arash if you want to discuss this in more detail - I think the technical details will get a bit off-topic for the list). Yours Jonathan -- Dr Jonathan Yates | Theory of Condensed Matter Group Corpus Christi College | Cavendish Laboratory Cambridge, CB2 1RH, UK | Cambridge, CB3 OHE, UK email jry20 at cam.ac.uk | Tel +44 (0)1223 337461 From joseph99077 at yahoo.com.tw Mon Mar 26 05:01:50 2007 From: joseph99077 at yahoo.com.tw (joseph99077 at yahoo.com.tw) Date: Mon, 26 Mar 2007 11:01:50 +0800 (CST) Subject: [Wannier] a user of wannier90 Message-ID: <410320.6187.qm@web72315.mail.tp2.yahoo.com> Hello I am a new user of wannier 90 and I've got a few questions about the system.(hope that my questions are not very silly.) I followed the instruction manuel "tutorial.pdf" included in the package.I tried to run all the examples in wannier90, which goes quite ok (i guess). one of the examples (ex.2) gave us the code to calculate fermi surfaces. I tried to insert those code to other examples and compile them. unfortunately something wasn't right. the same situation happend to me with calculating the number of band which is in the manuel example 3, fail to work in other examples. I guess it's mainly my problem, but i hope you can help me. thanks. good day. Joseph ___________________________________________________ ??????? ? ???????????????? http://messenger.yahoo.com.tw/ From jry20 at cam.ac.uk Mon Mar 26 12:59:57 2007 From: jry20 at cam.ac.uk (Jonathan Yates) Date: Mon, 26 Mar 2007 11:59:57 +0100 (BST) Subject: [Wannier] a user of wannier90 In-Reply-To: <410320.6187.qm@web72315.mail.tp2.yahoo.com> References: <410320.6187.qm@web72315.mail.tp2.yahoo.com> Message-ID: On Mon, 26 Mar 2007, joseph99077 at yahoo.com.tw wrote: > I followed the instruction manuel "tutorial.pdf" > included in the package.I tried to run all the > examples in wannier90, which goes quite ok (i guess). > one of the examples (ex.2) gave us the code to > calculate fermi surfaces. I tried to insert those code > to other examples and compile them. unfortunately > something wasn't right. the same situation happend to > me with calculating the number of band which is in the > manuel example 3, fail to work in other examples. > I guess it's mainly my problem, but i hope you can > help me. thanks. good day. Hi Joseph, You're going to need to give us some more details before we can help you. Take one of the systems you're try to run. Explain to us: (a) what the system is and what results you're expecting (b) exactly what you did (c) any error messages from wannier90. (look in the *.werr file) It will probably help us if you enclose the .win file. Yours Jonathan -- Dr Jonathan Yates | Theory of Condensed Matter Group Corpus Christi College | Cavendish Laboratory Cambridge, CB2 1RH, UK | Cambridge, CB3 OHE, UK email jry20 at cam.ac.uk | Tel +44 (0)1223 337461 From joseph99077 at yahoo.com.tw Thu Mar 29 14:53:41 2007 From: joseph99077 at yahoo.com.tw (joseph99077 at yahoo.com.tw) Date: Thu, 29 Mar 2007 20:53:41 +0800 (CST) Subject: [Wannier] bands problem in example5 "diamond" Message-ID: <647740.69525.qm@web72309.mail.tp2.yahoo.com> ___________________________________________________ ??????? ? ???????????????? http://messenger.yahoo.com.tw/ -------------- next part -------------- An embedded message was scrubbed... From: Subject: bands problem in example5 "diamond" Date: Mon, 26 Mar 2007 20:51:44 +0800 (CST) Size: 1951 URL: From marzari at MIT.EDU Thu Mar 29 15:29:20 2007 From: marzari at MIT.EDU (Nicola Marzari) Date: Thu, 29 Mar 2007 09:29:20 -0400 Subject: [Wannier] bands problem in example5 "diamond" In-Reply-To: <647740.69525.qm@web72309.mail.tp2.yahoo.com> References: <647740.69525.qm@web72309.mail.tp2.yahoo.com> Message-ID: <460BBF30.6020209@mit.edu> > another case, > I added > wannier_plot=true > in the example 7. and got the following message at the > bottom of "silane.wout" > > Exiting....... > plot_wannier: file UNK00001.1 not found > > In this case I guess I should make one UNK00001.1, but > I don't know how to make one for this case. Is there > any thing in the wannier90 which can produce that > file? > Hi Joseph, it's explained in the user guide, 7.15 . UNK... are file generated by your own electronic structure code, containing the |unk> orbitals. Arash, Jonathan - I presume that even if one uses PWSCF, one has to write a small script that produces the UNK... files, correct ? nicola --------------------------------------------------------------------- Prof Nicola Marzari Department of Materials Science and Engineering 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu From mostofi at MIT.EDU Thu Mar 29 15:55:55 2007 From: mostofi at MIT.EDU (Arash A Mostofi) Date: Thu, 29 Mar 2007 09:55:55 -0400 (EDT) Subject: [Wannier] bands problem in example5 "diamond" In-Reply-To: <460BBF30.6020209@mit.edu> References: <647740.69525.qm@web72309.mail.tp2.yahoo.com> <460BBF30.6020209@mit.edu> Message-ID: Dear Joseph As Nicola mentions, the UNK file is obtained from the underlying electronic structure calculation. In the case of PWscf, the post-processing code pw2wannier90.x is able to write this file. You first need to do a self-consistent calculation, then run the post-processing code pw2wannier90.x (with the "write_unk" flag set to "true" in its input file). If you are able to download the latest cvs version of PWscf, the examples directory contains a subdirectory called WAN90_example which shows how PWscf and Wannier90 are interfaced. Cheers Arash On Thu, 29 Mar 2007, Nicola Marzari wrote: > >> another case, I added wannier_plot=true >> in the example 7. and got the following message at the >> bottom of "silane.wout" >> Exiting....... >> plot_wannier: file UNK00001.1 not found >> >> In this case I guess I should make one UNK00001.1, but >> I don't know how to make one for this case. Is there >> any thing in the wannier90 which can produce that >> file? >> > > > Hi Joseph, > > it's explained in the user guide, 7.15 . UNK... are file generated > by your own electronic structure code, containing the |unk> orbitals. > > Arash, Jonathan - I presume that even if one uses PWSCF, one has to > write a small script that produces the UNK... files, correct ? > > nicola > > --------------------------------------------------------------------- > Prof Nicola Marzari Department of Materials Science and Engineering > 13-5066 MIT 77 Massachusetts Avenue Cambridge MA 02139-4307 USA > tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu > _______________________________________________ > Wannier mailing list > Wannier at quantum-espresso.org > http://www.democritos.it/mailman/listinfo/wannier >