Custom Field subtracting average area of "blank"
Options
I'm in need of some assistance in perfecting some custom field calculations. The situation is as follows.
I was hoping that I could have multiple "summarize custom fields" and reference which blank I want to use, but so far that hasn't worked either.
Any advice that you all could provide would be greatly appreciated.
Thanks!
-Adam
- I have an extraction solution being injected as a blank prior to every sample. Each blank/sample are injected in triplicate so I am taking the average area of the blank and subtracting it from the average area of the sample.
- Blanks are labeled EXS1 through EXSn, where n is the total number of blanks.
- I currently have samples labeled as S1 through Sn, again where n is the total number of samples.
I was hoping that I could have multiple "summarize custom fields" and reference which blank I want to use, but so far that hasn't worked either.
Any advice that you all could provide would be greatly appreciated.
Thanks!
-Adam
0
Best Answer
-
Can you try using the "Summarize Custom Fields Incrementally" function and insert it across various lines of the sample set? So if you have the first sequence of EXS1, followed by S1, then insert the function Summarize Custom Fields Incrementally on the next line and for the Label Reference box type in EXS1 and S1. Same for EXS2 and S2 etc. I dont know if this would work as the incremental function typically only displays a "rolling" review of the samples up to that point, independent of label reference entries, but it would be worth trying. Otherwise, you need a separate set of CFs for each possible set of entries, which would be unsustainable.
Another way around it is to create a peak, enumerated CF with use as Field selected, which ties in all the labels and instructs Empower to subtract the average area of each set of blanks from each set of samples. That could be a very long formula though if you run lots of samples!1
Answers
-
I was able to get this to work using the peak enumerated CF with the Use As Field selected. I did stumble upon this before you answered on Friday but it wasn't quite working. It wasn't until I saw in the help that I had to have the (fc) at the end of each entry in the translation table. Once I put that in there, I was able to find success!
I'm going to use Excel and the concatenation formula to build up the main CF formula as well as the translations. That should eliminate any typos.
Thanks so much for helping with this!
1 -
Now that I have this squared away, at least three of the adjusted area's will be used as calibrators. Is there a way to force Empower3 to use the new area counts? I was assuming that this is another custom field (maybe using the replace function), but I'm now thinking it has to be set-up in the processing method itself.Empower2019 said:
Thanks!0 -
I *think* I have figured this out. I created a real peak CF set for Standards Only, and the only thing in the calculation will be the output for the adjusted area from my enumeration calculation above. Then I can use that real peak CF number in the processing method (instead of the default "Amount").FizzyChemist said:
Now that I have this squared away, at least three of the adjusted area's will be used as calibrators. Is there a way to force Empower3 to use the new area counts? I was assuming that this is another custom field (maybe using the replace function), but I'm now thinking it has to be set-up in the processing method itself.Empower2019 said:
Thanks!0 -
Well, that was a bust. Apparently, you can't use the output of a field enum as a value. I tried everything possible but have had to resign changing the protocol to have only one injection each.
Now it's a simple real peak CF with a formula of "Area-EXS%..(Area)". And I can designate that CF as the Y-value in the processing method.
I still don't like not being able to take the average of the three and use that to subtract...but if this is the only way I can make it work, I'll work with what I can.
Edited: fixed grammar0 -
Thats the problem with use as field in Enum CFs- they cant be used for any further calculations. They work well when the formula output you require is the end of the road. In your case you could change the Enum CF to a Peak Real CF with the formula:
EQ(Label,"S1")*S1.%..AVE(Area)-EXS1.%..AVE(Area)+EQ(Label,"S2")*S2.%..AVE(Area)-EXS2.%..AVE(Area)+EQ(Label,"Sn")..etc I wouldn't recommend this if you have more than 10 Samples as the formula may crash.
Using that formula you will get a different value at S1..up to Sn..which is what you are looking for. And you are free to do more with it since its a peak real number instead of the Use As Field (fc) which is essentially a "photograph" of the answer and in a syntax unrecognisable for further manipulation by Empower.
For the 3 adjusted areas required as calibrators, are these used to calibrate the remainder of the Sn...samples or a different set of samples? You could always create a second method set for these calibrators with a processing method set up as this peak real cf in your Y Value then run the samples using your normal method set for all the non-calibrators then for the 3 or more calibrators select this new method set in the sample set Method Set option.
It sounds like you have it sorted though so maybe its moot at this point.0 -
This should be the last follow-up that I have. *Hopefully* 
I noticed that when I had my sample set running as
Extraction Solution
Calibrator
and repeating until everything was finished, I was able to get it to fully calibrate, but it is giving me 7 different calibration curves. At the end of the day, it really won't matter since it will only be using the last one to quantitate any controls/samples. However, when I am trying to print our Calibration plot based, it prints all 7 plots along with their corresponding calibration tables. It's fully explainable...but I'd prefer it not to be like that.
I've narrowed the "issue" coming from the fact that I have injected samples (aka the extraction solution) in between each of the calibrators. When I Altered the sample set to only process the calibrators (using the Area without doing the subtraction of the blank), I got a single calibration curve with the 7 points.
Is there any setting in the report editor to have it ignore the 6 previous curves and only use the latest set? I believe that there's a similar setting for Tables (I'd uncheck the "Include Duplicate Results" box), but I don't see that in the Calibration plot parameters (unless it really is there and I'm just missing it).
Edited to add: The same processing method was used for all three of the trials above, with the only exception being that the first one had "Area" set for the Y-Value. The other two had my CF Real Peak value set for the Y-Value.
Edited to add: I know 7 points on the calibration curve is a lot. For production purposes it will be three points, but for the validation stages, I went with 7 points.
Edited to add:
I *can* get it to print only the last curve by only preview/publishing the last calibrator. It shows that all of 7 points are based off 7 individual cal curve IDs, but again, all of the controls and sample data will be processed using the last cal curve Id (2352 in this instance).
Still, if I could somehow force it to only build a single curve with a single cal curve Id, I'd be much happier.0 -
This sounds like a report method data filtering issue. I see what you are saying about only wanting the last calibration curve displayed, can you filter by Cal Curve ID = 2352? If that is too awkward because the final Cal Curve ID would have to constantly be changed to display the filter as you want it, consider filtering it by another CF value in the sample set at the same line as your final calibrator.
For example, a simple Sample, Enum CF in your sample set and call it Calibrator and you could have 2 options from the enumerated list such as N/A and Calibrator or something similar. Then in the sample set, select N/A for all entries except the final calibrator which you can set as Calibrator in sample set and then simply filter by Calibrator = Calibrator.
Maybe im missing something in your post but it seems you have a lot of options to only display the final cal curve such as filter by Level, Cal Curve ID, samplename etc.
0