Sorry, datafile attached. Also sorry that I sent the original email twice - I got an automatic email from the Argonne labs virus checker that I assumed was from the mailing list, but I now have half a dozen of them so I assume they were actually sent when the email was delivered to individual recipients. Thanks, Ian ---- Ian Godfrey PhD Student, UCL/JAIST Programme, Industrial Doctorate Centre in Molecular Modelling and Materials Science, Department of Chemistry, University College London. i.godfrey@ucl.ac.uk 02076 794864 ________________________________________ From: ifeffit-bounces@millenia.cars.aps.anl.gov on behalf of Bruce Ravel Sent: 28 February 2014 14:02 To: XAFS Analysis using Ifeffit Subject: Re: [Ifeffit] Saving and reopening project files in Artemis 0.9.18.2 on Windows 8.0 Pro x64 This is a confusing bug report. In your explanation, you dwell on the fact that you are importing a specific chi(k) file. You then give detailed instructions on how to reproduce the problem. However, you did not attach that file. The value of providing me with a recipe is so that I can do *exactly* what you did. That means you have to provide me with all the ingredients for the recipe. B On 02/28/2014 08:46 AM, Godfrey, Ian wrote:

Dear All,

I'm encountering a problem whenever I save a fitting project in Artemis (I have tried both the "Save Project" and "Save Current Fit" options from the file menu) and then try to reopen it . I have only encountered the issue with quick first shell fits, but I haven't tried anything else recently. Specifically the parameters in the paths list are reset to their default value (while those in the GDS window are not) when I reopen the file and when I try and fit the data again Artemis gives a message that:

/"This data file does not exist"/

The data file has not been moved, deleted or otherwise modified on my computer.

//

It may be relevant that I'm not importing an Athena project into Artemis, but I'm importing a chi(k) file (*.cho) exported from Viper (https://www.cells.es/Beamlines/CLAESS/software/viper.html) (export chi(k) on original grid).

I also frequently get a/"perl.exe has stopped working A problem caused the programme to stop working correctly. Windows will close the program and notify you if a solution is available" /error from Windows when exiting Artemis, which could be related.

I have attached Artemis log files from running the software to both save and open a project and I have attached an example problematic project file.

To reproduce:

* Start Artemis * File > Import > chi(k) data * Import a chi(k) data file * Set up a quick first shell fit * Customise some of the parameter names * Run a fit * File > Save Current Fit or File > Save Project (as) * Close Artemis * Reopen the saved project file * The parameter names in the paths list have been reset to their default (but those in GDS have not) * Run a fit * Artemis will complain that the data file is missing

If anyone has any suggestions for a workaround or knows whether this is a bug in Artemis or a problem at my end I'd be grateful to hear.

Thanks,

Ian

----

Ian Godfrey

PhD Student,

UCL/JAIST Programme,

Industrial Doctorate Centre in Molecular Modelling and Materials Science,

Department of Chemistry,

University College London.

i.godfrey@ucl.ac.uk mailto:i.godfrey@ucl.ac.uk

02076 794864

_______________________________________________ Ifeffit mailing list Ifeffit@millenia.cars.aps.anl.gov http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit

-- Bruce Ravel ------------------------------------ bravel@bnl.gov National Institute of Standards and Technology Synchrotron Science Group at NSLS --- Beamlines U7A, X24A, X23A2 Building 535A Upton NY, 11973 Homepage: http://xafs.org/BruceRavel Software: https://github.com/bruceravel _______________________________________________ Ifeffit mailing list Ifeffit@millenia.cars.aps.anl.gov http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit

I've changed my mind about this. Importing chi(k) data directly into Artemis does not seem to work, regardless of the content of the column data file. However, putting your .cho file into Athena then into Artemis works for me. Importing chi(k) data directly into Artemis should be considered broken in the current version of Artemis. B On 02/28/2014 10:11 AM, Bruce Ravel wrote:

On 02/28/2014 08:46 AM, Godfrey, Ian wrote:

...

I'm encountering a problem whenever I save a fitting project in Artemis (I have tried both the "Save Project" and "Save Current Fit" options from the file menu) and then try to reopen it . I have only encountered the issue with quick first shell fits, but I haven't tried anything else recently. Specifically the parameters in the paths list are reset to their default value (while those in the GDS window are not) when I reopen the file and when I try and fit the data again Artemis gives a message that:

/"This data file does not exist"/

The data file has not been moved, deleted or otherwise modified on my computer.

One of the central problems here is that the error message, "This data file does not exist", is vague and unhelpful in this context.

I am having trouble finding my way through your recipe and the data you have provided. As a result, I am not 100% sure what the problem is. However, I have a strong hunch. You'll let me know, I hope, if the following pans out.

I was unable to properly import your foilref.cho file into Artemis, although I can import it into Athena just fine. Artemis just isn't as clever about interpreting data as Athena is. That is because, in the work flow that I recommend for my software, it doesn't have to be. It turns out that the option to import chi(k) data into Artemis is expecting your chi(k) data to be on a uniform grid that starts at 0.

Admittedly, that's not well documented.

Your .cho file is not in the form that Artemis expects.

I am kind of astonished that things worked as well as they did for you. I am suspicious that the irregular chi(k) data was saved in some way incorrectly. When you reopened the .fpj file, it was imported incorrectly. Examining the contents of your .fpj file confirms that something odd happened to the data. Unfortunately, Demeter didn't complain until you tried to run a new fit -- long after the problem began.

If you simply must use VIPER's output as the input to Artemis, here is my suggested work-around. Import the .cho file into Athena and save an Athena project file. This will interpolate your data onto an even k-grid with zero padding at the beginning. Save the Athena project file and import it into Artemis.

Try that and let me know if that gets you over the hurdle.

Of course, there are two other work-arounds. The first is to use Athena rather than VIPER to process your data. The second is to use VIPER rather than Artemis to analyze your data :)

I will put this issue with importing non-standard chi(k) data directly into Artemis on the to do list, but I see it as an edge case that may not get huge attention.

B

-- Bruce Ravel ------------------------------------ bravel@bnl.gov National Institute of Standards and Technology Synchrotron Science Group at NSLS --- Beamlines U7A, X24A, X23A2 Building 535A Upton NY, 11973 Homepage: http://xafs.org/BruceRavel Software: https://github.com/bruceravel