September 2006 - Ifeffit - millenia.cars.aps.anl.gov

Installing Tk/Horae on Fedora Core 5
by Bhoopesh Mishra 12 Sep '06

12 Sep '06

Dear All, I am in the process of installing Ifeffit 1.2.9 with latest Horae package (-062) on my a new machine on which I have put Fedora core 5. I followed the instructions for installing the entire package, and had partial success in it. Ifeffit and PGPLOT seem to be working fine. I believe Tk-804.027 is also working fine, I just ran "widget" and it works. However when I tried to ./horae_update, I got error message and it aborted without compilation. To look into the problem, I manually downloaded a previous version of horae (horae-061) from another computer on which everything is working fine (Tk-800.024 with horae-061 on a Fedora Core 2). In the manual installation of horae_061, perl Makefile.PL works fine. But "make" does not complete. It terminates with the same error message as the one with which ./horae_update terminated. (attached as a screenshot) I have attached the screenshots of the error message for horae_update. I have also attached screenshots of "make install" and "make test" for Tk-804-027 just in case you would like to see that. Since Ifeffit and PGPLOT are working, I am inclined to believe that it has something to do with Fedora Core 5 and its compatibility with Tk-804.027 and horae-062. Funny part is that I tried to downgrade the Tk to Tk-800.024, but it failed to install, which indicates some issues of Tk with FC 5. I also tried the combination of Tk-804.027 with horae-060. Has anyone gone through this combination of installing Tk-804.027 with horae-062 on a Fedora Core 5? Any insight would be extremely helpful. Thanks a lot in advance. Please let me know what more information would you need if the given information is insufficient. PS: I have been successfully using all these programs on FC 2 and FC 3 for last few years but some very generous people always helped me with the installation and am now trying to be independent. Thanks again, Bhoopesh

3 4

Re: X11 on 10.4.7
by Bruce Ravel 11 Sep '06

11 Sep '06

On Monday 11 September 2006 15:34, Kjell Bolmgren wrote: KB> KB> Bruce, KB> KB> Did you ever find out how to install X11 on 10.4.7. KB> KB> I have tried to install it from the cd and to download it but neither KB> works so far. Kjell, I am CCing this to response to the mailing list in hopes that someone will be able to lend you a helping hand. I'll restate now what I say every time I make a post about anything Apple: I don't own a Mac, I don't have a Mac at work, and my experience on the Mac in any other context is extremely limited. That Athena and Artemis work on the Mac is thanks to my choice to use cross-platform programming tools (like perl and Tk) and on the kind contributions of people like Paul, who went the final mile to make easy-to-install packages. B -- Bruce Ravel ---------------------------------------------- bravel(a)anl.gov Molecular Environmental Science Group, Building 203, Room E-165 MRCAT, Sector 10, Advanced Photon Source, Building 433, Room B007 Argonne National Laboratory phone and voice mail: (1) 630 252 5033 Argonne IL 60439, USA fax: (1) 630 252 9793 My homepage: http://cars9.uchicago.edu/~ravel EXAFS software: http://cars9.uchicago.edu/~ravel/software/

3 2

Re: [Ifeffit] Artemis error message
by Matt Newville 11 Sep '06

11 Sep '06

Stefano, Feff6 is an executable. When it runs it needs a file named feff.inp to be in the current working directory, otherwise it will say: Feff 6L.02 Fatal Error: cannot open file "feff.inp" in module rdinp which is exactly what you got when you typed 'feff6' in a directory without a feff.inp file. That doesn't tell us a lot, but it does mean that this version of feff6 is at least not corrupted, so the Installer is probably fine. The first error message from Artemis that you sent: "The error is probably due to having Artemis configured...." and gave an "Uh-Oh" in the Echo Buffer is definitely what Artemis says when it cannot find feff6. You said you got this when the feff_executable was set to 'feff6'. Did you try setting feff_executable to '/Applications/Ifeffit/bin/feff6'? Note that spelling counts. You could open a Terminal, type 'which feff6' and paste that into the feff_executable field. Note to Bruce: How about having a "browse" button for selecting the feff_executable? > I interpret this message with the conclusion that I can just type 'feff6' in > the Artemis preferences, as the command is actually seen and of course it is > missing the feff.inp file if I run the command from the terminal. Well, it *should* work that way, but using /Applications/Ifeffit/bin/feff6 will definitely work. It does work for several of us. Hope that helps, --Matt

1 0

Re: [Ifeffit] Demeter: my new software project
by Bruce Ravel 11 Sep '06

11 Sep '06

On Saturday 09 September 2006 11:02, Mauro Rovezzi wrote: > I really like your new project! I think it's a great idea because > Athena&Artemis are good softwares but clicking on them all the day it's > quite boring. Exactly. To my mind, the biggest problem is that it is difficult, time consuming, and error prone to encode a complex fitting model using Artemis' GUI. I forsee that one of my common uses of Demeter will be to encode a complex fitting model in a one-off script and have that script write out an Artemis project file. Then I'll get the good parts of the GUI while using something automated to deal with the tedious parts. > On the other hand, I find the Ifeffit wrapper too > low-level. In fairness, Ifeffit is great at doing its low-level thing. It's just that low-level tools tend not to scale up gracefully. > Well, Demeter seems to me the best solution as a middle-level > interface to Ifeffit, that permits to fork to user-oriented projects > written in Perl. I also appreciate the choice to move this project to a > version control system (Subversion), that is the future possibility for > users to directly contribute to the code. Thanks for all the kind words, Mauro. I am excited that you get the point of middleware and that you get the point of using Subversion. Even if you use and don't contribute, Subversion lets you always use up-to-the-minute code. B -- Bruce Ravel ---------------------------------------------- bravel(a)anl.gov Molecular Environmental Science Group, Building 203, Room E-165 MRCAT, Sector 10, Advanced Photon Source, Building 433, Room B007 Argonne National Laboratory phone and voice mail: (1) 630 252 5033 Argonne IL 60439, USA fax: (1) 630 252 9793 My homepage: http://cars9.uchicago.edu/~ravel EXAFS software: http://cars9.uchicago.edu/~ravel/software/exafs/

1 0

Demeter: my new software project
by Bruce Ravel 08 Sep '06

08 Sep '06

Hi! I think that some of you will be interested in my latest XAS software project, which I am calling Demeter. Demeter is an object-oriented programming interface to Ifeffit written in Perl. It is intended to encompass most of the current functionality of Athena and Artemis while at the same time being more generalized, lower-level, and much more robust. In short, Demeter is not a software product. Rather, it is the thing that software products can be made from. Demeter is targeted at a wide variety of XAS data analysis chores, such as * GUI's like Athena and Artemis. In fact, I intend to rewrite Artemis using Demeter once I have incorporated enough functionality into Demeter. * A parser for feffit.inp files allowing you to translate your old analysis using feffit into ifeffit scripts and Artemis projects. * The analytic component of a high-throughput system such as those presented at the XAFS13 conference by the groups of Wolfram Meyer-Klaucke and Sven Schroeder. * The backend to specialized XAS software, like Shelly Kelly's mkfit program (http://www.mesg.anl.gov/exafs.html) or for ideosyncratic analysis like that described in Physical Review B, 73, p. 184121 (2006) There is still A LOT to do, but even at this early stage Demeter is very useful. In fact, Demeter can already be used for real data processing and analysis. Demeter is available on an anonymous Subversion server, so anyone with an SVN client can check out a copy of Demeter at any time. Information about Demeter and the Subversion server is at http://cars9.uchicago.edu/~ravel/software/demeter.html You can browse the Demeter source code with your web browser at http://cars9.uchicago.edu/svn/demeter/ If this seems interesting to you, take a look at some of the short example programs demonstrating Demeter's capabilities and economy of language: http://cars9.uchicago.edu/svn/demeter/examples B -- Bruce Ravel ---------------------------------------------- bravel(a)anl.gov Molecular Environmental Science Group, Building 203, Room E-165 MRCAT, Sector 10, Advanced Photon Source, Building 433, Room B007 Argonne National Laboratory phone and voice mail: (1) 630 252 5033 Argonne IL 60439, USA fax: (1) 630 252 9793 My homepage: http://cars9.uchicago.edu/~ravel EXAFS software: http://cars9.uchicago.edu/~ravel/software/exafs/

2 1

silly me!
by Bruce Ravel 06 Sep '06

06 Sep '06

Hi folks, It just occured to me that I didn't mention the web page where I document all the changes to my codes in the most recent release. Athena especially has a number of interesting new features. You might want to take a look at this http://cars9.uchicago.edu/iffwiki/HoraeVersion062 B -- Bruce Ravel ---------------------------------------------- bravel(a)anl.gov Molecular Environmental Science Group, Building 203, Room E-165 MRCAT, Sector 10, Advanced Photon Source, Building 433, Room B007 Argonne National Laboratory phone and voice mail: (1) 630 252 5033 Argonne IL 60439, USA fax: (1) 630 252 9793 My homepage: http://cars9.uchicago.edu/~ravel EXAFS software: http://cars9.uchicago.edu/~ravel/software/exafs/

2 1

Compiling horae/iffefit on Mac OS 10.4.7 Intel
by Todd Hufnagel 02 Sep '06

02 Sep '06

Greetings: I'm trying to get ifeffit and the horae packages running on an Intel-based Mac running OS 10.4.7. As someone else noted previously, the binary downloads don't work on Intel Macs. (As an aside to the person who asked about the console log, in my case the programs quit immediately after launching and no messages are generated in either console.log or system.log. I do have X11 installed; this behavior is the same with either X11 or Aquaterm chosen for output.) So I decided to compile from the source. Unfortunately, I ran into a problem when compiling horae-062a. Using the Makefile.PL for Mac OS (http://cars9.uchicago.edu/~ravel/misc/horae_on_mac/Makefile.PL ) seems to work fine, and after "make;make install" I get no errors and executables for Artemis et al. are produced. Unfortunately, attempting to run any of the executables produces an error similar to the following: >tchmac:/Applications/ifeffit/bin hufnagel$ ./artemis >dyld: lazy symbol binding failed: Symbol not found: _iff_exec > Referenced from: >/Library/Perl/5.8.6/darwin-thread-multi-2level/auto/Ifeffit/Ifeffit.bundle > Expected in: dynamic lookup > >dyld: Symbol not found: _iff_exec > Referenced from: >/Library/Perl/5.8.6/darwin-thread-multi-2level/auto/Ifeffit/Ifeffit.bundle > Expected in: dynamic lookup > >Trace/BPT trap A little digging around on the net suggested that this error comes from compiling with gcc-4.0 rather than gcc-3.3. Switching to that compiler, I can again run Makefile.PL successfully, but then when I run 'make' I get a new and exciting error: >[...much successful stuff deleted...] >cp Zlib.pm ../../blib/lib/Compress/Zlib.pm >AutoSplitting ../../blib/lib/Compress/Zlib.pm >(../../blib/lib/auto/Compress/Zlib) >/usr/bin/perl /System/Library/Perl/5.8.6/ExtUtils/xsubpp -typemap >/System/Library/Perl/5.8.6/ExtUtils/typemap -typemap typemap >Zlib.xs > Zlib.xsc && mv Zlib.xsc Zlib.c >cc -c -I./zlib-src -g -pipe -fno-common -DPERL_DARWIN >-no-cpp-precomp -fno-strict-aliasing -I/usr/local/include -O3 >-DVERSION=\"1.41\" -DXS_VERSION=\"1.41\" >"-I/System/Library/Perl/5.8.6/darwin-thread-multi-2level/CORE" >Zlib.c >cc: installation problem, cannot exec `cc1': No such file or directory >make[2]: *** [Zlib.o] Error 1 >make[1]: *** [subdirs] Error 2 >make: *** [subdirs] Error 2 and obviously no executables are produced. A little more digging on the web revealed that the cc1 stuff should be in the /usr/libexec/gcc tree. Unfortunately, while there is a subdirectory there for gcc-4.0 for Intel, the only files for gcc-3.3 are for PowerPC. That is, while there is a /usr/libexec/gcc/darwin/ppc/3.3/cc1, there is no corresponding subdirectory for Intel processors. The reason the compilation worked for gcc-4.0 is that there is, in fact, a /usr/libexec/gcc/i686-apple-darwin8/4.0.1/cc1. At this point, I am stuck. I can compile under gcc-4.0 but get executables which don't run, and I can't compile at all under gcc-3.3. I also ran into trouble compiling ifeffit, but I haven't exhausted my own abilities in chasing down those problems (yet). Any assistance/advice/condolences would be appreciated. Thanks, Todd Hufnagel -- Todd C. Hufnagel Professor Department of Materials Science and Engineering Johns Hopkins University 102 Maryland Hall 3400 North Charles Street Baltimore, MD 21218-2681 Voice: (410) 516-6277 Fax: (410) 516-5293 Email: hufnagel(a)jhu.edu http://www.jhu.edu/~matsci/people/faculty/hufnagel/hufnagel.html

4 4

Alloy EXAFS data fitting
by XQ WANG 01 Sep '06

01 Sep '06

Dear all, I am trying to understand some EXAFS data collected on an alloy system (sorry I can't say it since these are not my data). It is a two-element alloy system, name AB. Their fitting results showed that A-B coordination number and A-A coordination number did not follow the similar trend with the temperature increase. I am just wondering if this case is possible? I had thought if A-A and A-B are from the same compound, they should follow the same trend. Maybe I am wrong. Another question is whether it is OK to combine coordination of A-A in A compound with CN of A-A in AB alloy if these two A-A bond distances are very close? If this case is OK, then combination of A-A CN and CN of A-B could follow differnt trend (I guess). But I thought we could not combine CN from different materials. Maybe I am wrong again?? Please let me know any suggestions concerning this situation. Thanks! Xianqin _________________________________________________________________ Check the weather nationwide with MSN Search: Try it now! http://search.msn.com/results.aspx?q=weather&FORM=WLMTAG

3 4