Author Topic: CALCZAF Batch Compute  (Read 6043 times)

Ashley Norris

  • Graduate
  • **
  • Posts: 7
CALCZAF Batch Compute
« on: April 20, 2016, 09:39:07 AM »
Hi everyone, I want to reprocess some old EPMA datasets of mine using different matrix correction options. To get things moving, I have generated a CALCZAF data file using my own software and can load this using CALCZAF without any difficulty.

My question is whether there is a way to calculate all of the entries in the data file in one go?

At the moment I can go to "Analytical"->"Display Current Sample Elements" and click "Calculate" and then "Load Next Dataset from Input File" and this works.  However, my input file contains 200 entries (with more to come, I want to check about 1000 entries all up)  and this seems awfully tedious.

Am I missing something fundamental? Or is it a simple matter of clicking on the correct button? I have tried to find it, and have read through the forum, as well as the help file, but maybe I missed it?

My final thought was whether this is a question of philosophy? It is clear to me that CALCZAF is intended for careful inspection of all aspects of the matrix correction routine, so maybe batch processing has been omitted intentionally?

Thank you in advance, Ash

Probeman

  • Emeritus
  • *****
  • Posts: 2860
  • Never sleeps...
    • John Donovan
Re: CALCZAF Batch Compute
« Reply #1 on: April 20, 2016, 09:48:36 AM »
Hi everyone, I want to reprocess some old EPMA datasets of mine using different matrix correction options. To get things moving, I have generated a CALCZAF data file using my own software and can load this using CALCZAF without any difficulty.

My question is whether there is a way to calculate all of the entries in the data file in one go?

At the moment I can go to "Analytical"->"Display Current Sample Elements" and click "Calculate" and then "Load Next Dataset from Input File" and this works.  However, my input file contains 200 entries (with more to come, I want to check about 1000 entries all up)  and this seems awfully tedious.

Am I missing something fundamental? Or is it a simple matter of clicking on the correct button? I have tried to find it, and have read through the forum, as well as the help file, but maybe I missed it?

My final thought was whether this is a question of philosophy? It is clear to me that CALCZAF is intended for careful inspection of all aspects of the matrix correction routine, so maybe batch processing has been omitted intentionally?

Thank you in advance, Ash

Hi Ashley,
Welcome back!    :)

Something fundamental I think:     :D

« Last Edit: April 14, 2020, 12:26:48 PM by John Donovan »
The only stupid question is the one not asked!

Ashley Norris

  • Graduate
  • **
  • Posts: 7
Re: CALCZAF Batch Compute
« Reply #2 on: April 20, 2016, 01:24:30 PM »
Now that's embarrassing!  :-[

Sorry to waste your time with such a simple question. I was in-and-out of the "File" menu all day long, and just didn't see that option. You know what they say: More speed less haste...indeed.

What I'm doing is recalculating a large set of trace probe analyses collected on an old Jeol using the SAMx software. To make this slightly less painful I've put together a web-page that allows me to upload the SAMx sample and standard files, and get back CalcZAF files ready for processing. The idea is to be as slothful as possible and let the scripting catch the file handling mistakes (1000s of points, dozens of sessions) I am bound to make.

I don't know if such a thing will be of use to anyone else, but on the off-chance it is, I'll post the link here once it's been debugged and ready for others to use.

Thanks again,

Ash

John Donovan

  • Administrator
  • Emeritus
  • *****
  • Posts: 3304
  • Other duties as assigned...
    • Probe Software
Re: CALCZAF Batch Compute
« Reply #3 on: April 20, 2016, 01:35:13 PM »
What I'm doing is recalculating a large set of trace probe analyses collected on an old Jeol using the SAMx software. To make this slightly less painful I've put together a web-page that allows me to upload the SAMx sample and standard files, and get back CalcZAF files ready for processing. The idea is to be as slothful as possible and let the scripting catch the file handling mistakes (1000s of points, dozens of sessions) I am bound to make.

I don't know if such a thing will be of use to anyone else, but on the off-chance it is, I'll post the link here once it's been debugged and ready for others to use.

No worries, as they say!  See probeman's "signature".   And I'm sure someone might find such a script useful so feel free to post a link.

When you say "old JEOL" which model do you mean exactly?   

Also note that CalcZAF doesn't have many of the quant features that Probe for EPMA has. For example, the iterated interference corrections and time dependent intensity corrections.  So while you can certainly run CalcZAF to get an idea of the variance from the 10 different matrix corrections (and 6 different MAC tables), one might have bigger accuracy errors to deal with.

On a completely unrelated topic, do you still have a link for creating periodic tables with different WDS crystals, like the one you did for me years ago?
« Last Edit: April 20, 2016, 01:37:40 PM by John Donovan »
John J. Donovan, Pres. 
(541) 343-3400

"Not Absolutely Certain, Yet Reliable"

Ashley Norris

  • Graduate
  • **
  • Posts: 7
Re: CALCZAF Batch Compute
« Reply #4 on: April 20, 2016, 03:07:47 PM »
We use a JXA-8600 running SAMx in a neighbouring department, but I'm not 100% certain which version it has because the system is not ours and is outside our control. It could be that the problem I'm up against is fixed in the latest SAMx version. Anyway, I'm doing some trace metals in iron, and for trace element analysis there is a rounding error in their quantification routine that quantizes the trace element output. I think the K ratio is being rounded to three decimal places, which is how it looks in the data files. If my measurements were below detection then this wouldn't matter, but I'm in a kind of limbo and seeing quantized values well above (5-times) the detection limit, so I believe the counts to be real.

Anyway, I've just run my first batch through CalcZAF and the results appear to be much better, with no apparent rounding errors, and my secondary standards are looking good: so it's a preliminary win for now! :) Tomorrow I'll evaluate the matrix corrections and see if there is much improvement to be had.

I absolutely do still have my online periodic table generator: http://norris.org.au/expet/ptab/

The generator has evolved a bit over the years, but is essentially still the same. If you have any trouble with the custom data upload (which is what I used back in the day to mark up the X-ray lines) just let me know, maybe send me the file (or I can dig out my old one) and I might add it as a dedicated option:  (x) Show Cameca X-Ray lines (for example).

Great to be back in touch, good to see PFW is going strong, and shoot me an email if the generator doesn't do as it's told and I'll give it a little kick to help it along,

Ash

PS: Also thanks for giving me a chance to post my little quantization problem here: perhaps someone will recognise it, or others may benefit from using my workaround.

« Last Edit: April 20, 2016, 03:18:05 PM by Ashley Norris »

John Donovan

  • Administrator
  • Emeritus
  • *****
  • Posts: 3304
  • Other duties as assigned...
    • Probe Software
Re: CALCZAF Batch Compute
« Reply #5 on: April 20, 2016, 03:28:28 PM »
Anyway, I've just run my first batch through CalcZAF and the results appear to be much better, with no apparent rounding errors, and my secondary standards are looking good: so it's a preliminary win for now! :) Tomorrow I'll evaluate the matrix corrections and see if there is much improvement to be had.

Cool.

Note that there are several topics discussing the various matrix corrections in CalcZAF/Probe for EPMA:

http://probesoftware.com/smf/index.php?topic=671.0

http://probesoftware.com/smf/index.php?topic=47.0

http://probesoftware.com/smf/index.php?topic=490.0

Is this instrument at Oxford?  I think there's a new Cameca(?) instrument there also... 

Also please feel free to post your periodic table generator description and link to the General EPMA board.
John J. Donovan, Pres. 
(541) 343-3400

"Not Absolutely Certain, Yet Reliable"

Ashley Norris

  • Graduate
  • **
  • Posts: 7
Re: CALCZAF Batch Compute
« Reply #6 on: April 22, 2016, 03:42:08 AM »
In case anyone has a similar need, I finished the SAMx to CalcZAF conversion tool today. It can be accessed here, and there is a comprehensive help page and some example files to get started:

http://norris.org.au/expet/calczaf/

This tool will save me about a week of number handling, and eliminate countless errors and mistakes along the way...