[Ifeffit] Conflicting FEFF calculation

Tue Apr 4 09:43:40 CDT 2017

Dear Robert and Bruce,

Thank you very much for your explanation and time; I am grateful for 
that. I will test the problem with new version and report back as soon 
as possible.

With regards,

Ashis

On 2017-04-04 16:07, Bruce Ravel wrote:
>
>
> Robert, thanks for doing so much work to verify what Ashis is seeing.
>
> That said, I am still not seeing this problem.  My working notion is 
> that it is a problem that somehow got fixed while I was doing some 
> other work a few months ago.
>
> This morning, I prepared and uploaded a new installer package as a 
> pre-release of the upcoming version.  Here's the direct link.
>
> https://s3.amazonaws.com/demeter4xas/Demeter_0.9.26_with_Strawberry_Perl_(64)_pre2.exe 
>
>
> If one or both of you could install this and let me know if the 
> problem persists, I would be grateful.
>
>
> I did uncover a bunch of of problems with Demeter's crystallography 
> bits while I was working on
>
>    https://github.com/bruceravel/WebAtoms,
>    http://millenia.cars.aps.anl.gov/webatoms
>
> I don't really see the value of sifting through all that work to 
> figure out what had an impact on Ashis' issue.  If the newer code 
> works -- that seems good enough to me.
>
> B
>
>
> On 04/03/2017 06:08 PM, Robert Gordon wrote:
>> Hi Ashis,
>>
>> I can reproduce what you are seeing.
>> I exported the atoms.inp's for your two structures.
>>
>> If I import TAA first and run atoms, it looks ok. I then import TPA and
>> run - ok.
>> I re-run Atoms on TAA, it changes.
>>
>> In fact, from a clean start of Artemis 0.9.25 on win10pro/AMD64, if I
>> import TAA's inp and run it, it looks ok.
>> I added it again, it still looks ok, even if I re-run Atoms. I add TPA
>> and run it, it looks ok, but if I rerun
>> either of the TAA, they have changed. Changes persist even if TPA 
>> discarded.
>>
>> From another clean start of Artemis, I import TAA and run Atoms - ok.
>> I import a different inp (ni metal) and run it - ok. Re-run TAA and it
>> is ok.
>>
>> From another clean start, I import Ni metal and run it...ok
>> Import TPA and run it...ok...Rerun Ni and Ni changes...definite symmetry
>> change.
>> This symmetry change persists even if I discard TPA and import another
>> copy of Ni.
>>
>> Clean start, import Ni and run, then import another Rhombohedral
>> structure (B phase)
>> and same problem...symmetry changes and I don't seem to be able to get
>> it to change back,
>> even when I deliberately change the space group for Ni and it is the
>> only calculation remaining.
>>
>> Clean start...import Ni...ok....add rhombohedral structure (B) but don't
>> run it..rerun Ni and Ni changes.
>>
>> Clean start, import hexagonal structure (in P6)...ok...import
>> Ni...ok....change hex to Rhomb indexed on hex
>> and Ni changes when rerun...
>>
>> So...problem seems to be related to the symmetry - stuck in Rhombohedral
>> once a R-structure
>> is added...or symmetry changed to R.
>>
>> Haven't tried on Linux yet.
>>
>> -R.
>>
>>
>>
>>
>>
>> On 4/3/2017 1:46 PM, Ashis Biswas wrote:
>>>
>>> Dear Bruce,
>>>
>>> Sorry for late reply. I wanted to test in another computer before
>>> replying. Unfortunately, this time also I got different results for
>>> TAA. Following is the click by click description of what I am doing to
>>> arrive in this state:
>>>
>>> Double click on “Atemis” shortcut on desktop => Click “add” in the
>>> Feff calculation tab => Open “TPP” Atom input file from desktop=>
>>> Click “Run Atom” => Click “Run Feff” => Click “add” in the Feff
>>> calculation tab => Open “TAA” CIF file from desktop=> Click “Run Atom”
>>> => Click “Run Feff”
>>>
>>> I am sorry for killing your time. Actually, I am very much new in this
>>> field and therefore I really want to be sure that what I am doing for
>>> importing object for feff calculation is correct or not. Thank you
>>> very much for your support.
>>>
>>> With regards,
>>>
>>> Ashis
>>>
>>>
>>> On 2017-03-28 16:13, Bruce Ravel wrote:
>>>> On 03/27/2017 11:02 AM, Ashis Biswas wrote:
>>>>> Thank you very much for your reply. Here I am attaching two Artemis
>>>>> projects files (TAA_TPA and TPA_TAA), named according to their 
>>>>> order of
>>>>> import. In my computer the paths for *TAA* are different in two 
>>>>> project
>>>>> files. For example, i) in TAA_TPA project the Reff for three single
>>>>> scattering paths of O (@O1.1; @O3.1, and @O5.1) are different to
>>>>> that in
>>>>> TPA_TAA project, ii) the single scattering path @C5.1 (path number
>>>>> 6) at
>>>>> Reff of 2.682 A0 in TAA_TPA project is missing in TPA_TAA project, 
>>>>> and
>>>>> so on.
>>>>>
>>>>> I am also attaching the papers from where the coordinates for TAA and
>>>>> TPA are obtained: TAA (Kamenar et al., 1993: tables 1 & 2; Fig. 1) 
>>>>> and
>>>>> TPA (Pappalardo et al., 1983: Table 1; Fig. 1).
>>>>>
>>>>
>>>> I agree that the TPA_TAA.fpj project file that you sent me has
>>>> incorrect data in the Feff tab for TAA.
>>>>
>>>> I have not, however, been able to reproduce that here.  And not for
>>>> want of trying -- I spent the last half hour trying to find a way to
>>>> replicate what you are showing in the TPA_TAA file. Regardless of
>>>> the order in which I import the two structures and regardless of
>>>> whether I have discarded and reimported Feff calculations, when I
>>>> click the "Run Atoms" button on the TAA cif file, I /always/ get the
>>>> structure from Table 4 in the Kamenar paper.  I cannot reproduce the
>>>> Feff tab that you show in your TPA_TAA project file.
>>>>
>>>> I have no idea how you got into the state you show in the TPA_TAA
>>>> project.  It is clear that something went awry, but I don't know if
>>>> it is an error of the program or of the user.  I can't fix a problem
>>>> I can't observe.
>>>>
>>>> I understand that this is not the answer you were looking for.  I am
>>>> simply not seeing the problem you are reporting.
>>>>
>>>> B
>>>>
>>>>
>>>>
>>>
>>>
>>>
>>> _______________________________________________
>>> Ifeffit mailing list
>>> Ifeffit at millenia.cars.aps.anl.gov
>>> http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit
>>> Unsubscribe: http://millenia.cars.aps.anl.gov/mailman/options/ifeffit
>>
>>
>>
>> _______________________________________________
>> Ifeffit mailing list
>> Ifeffit at millenia.cars.aps.anl.gov
>> http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit
>> Unsubscribe: http://millenia.cars.aps.anl.gov/mailman/options/ifeffit
>>
>
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://millenia.cars.aps.anl.gov/pipermail/ifeffit/attachments/20170404/ae7c4e8e/attachment-0001.html>