You're right. The jatom definition (sec. 8.3.2 of 12.1 UG) indicates that only jatom 0 should be used unless "x lapw2 -qtl -band" is ran for jatom >0. Thanks for pointing out my user's mistake. Sorry, the code is fine (expect if you consider that the program produces an output that probably shouldn't exist).
On 9/13/2012 1:57 PM, Zhu, Jianxin wrote: > Once jatom is set to 1, it implies to do the heavy band plotting. Then > I don't see the point that one can avoid running "x lapw2 --qtl -band". > This difference is mentioned in UG 3.11.4 and 3.11.5 (I am still > looking at the UG 9-18-2008 version). > > JX > > From: Gavin Abo <gsabo at crimson.ua.edu <mailto:gsabo at crimson.ua.edu>> > Reply-To: A Mailing list for WIEN2k users > <wien at zeus.theochem.tuwien.ac.at <mailto:wien at > zeus.theochem.tuwien.ac.at>> > Date: Thu, 13 Sep 2012 12:46:19 -0600 > To: A Mailing list for WIEN2k users <wien at zeus.theochem.tuwien.ac.at > <mailto:wien at zeus.theochem.tuwien.ac.at>> > Subject: Re: [Wien] Small agr bug in spag.f > > The issue is not with grace. The problem only happens in a certain > case. To reproduce, try the following. > > 1. In w2web, converge the scf cycle for the TiC example > - Create TiC.struct with StructGen using the lattice and > atomic position parameters in the UG > - Click through the "initialize calc." steps using default > values, except use no spin-polarization and 1000 k-points > - Run the scf cycle with charge convergence of 0.0001 > 2. After it converges in the 12 cycle, click Bandstructure under Tasks > - Click create TiC.klist_band with fcc > - Click x lapw1 -band > - Click edit TiC.insp, copy/paste the fermi value on the screen > into the file and change jatom value to 1 (there is no error if > jatom is 0) > - Click x spaghetti > - In a terminal, run: xmgrace TiC.bands.agr > - At least on my system, the error appears in the terminal > because spag.f wrote ^@ characters in the agr file (without the > fix) since xmlabel(:) is undefined. > > Note: Make sure that you don't do "x lapw2 -band -qtl" for a > band character plot, you will not get an error because > xmlabel1(:)=label1(:) is reached inside of the if statement. > > On 9/13/2012 11:37 AM, Zhu, Jianxin wrote: >> Hi Gavin, >> >> I couldn't reproduce the error you have. >> Is this the issue with your grace? >> Also I don't see any change with this part of generating the agr >> file when version 12.1 is compared with version 11.1. >> >> Cheers, >> >> Jianxin >> >> From: Gavin Abo <gsabo at crimson.ua.edu <mailto:gsabo at >> crimson.ua.edu>> >> Reply-To: A Mailing list for WIEN2k users >> <wien at zeus.theochem.tuwien.ac.at >> <mailto:wien at zeus.theochem.tuwien.ac.at>> >> Date: Wed, 12 Sep 2012 13:32:51 -0600 >> To: A Mailing list for WIEN2k users >> <wien at zeus.theochem.tuwien.ac.at >> <mailto:wien at zeus.theochem.tuwien.ac.at>> >> Subject: [Wien] Small agr bug in spag.f >> >> Dear Prof. Blaha, >> >> This is to report a small bug in Wien2k 12.1. >> >> An error like the following might occur: >> >> xmgrace TiC.bands.agr >> >> Nonterminating string: TITLE "TiC atom 1 # k >> ene character >> syntax error: TITLE "TiC atom 1 # k >> ene character >> >> The xmlabel is set for the case shown in the code below. For >> other cases, xmlabel is uninitialized and garbage out is >> written in the agr file. >> >> SRC_spaghetti/spag.f >> -------------------- >> if(jatom.eq.0) goto 206 >> 207 read(9,'(a80)',end=206) aline >> if(aline(2:5).eq.'JATO') then >> ... >> xmlabel1(:)=label1(:) !line 430 >> ... >> endif >> endif >> ... >> write(40,123) xmtitle, jatom_list(1), xmlabel1 #line 588 >> -------------------- >> >> Therefore, the following change is suggested in spag.f: >> >> xmlabel1 ='' !Add at line 406 >> char0=0.01d0 >> >> Best Regards, >> >> Gavin >> -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20120913/2c8bce9c/attachment.htm>