sum normalization error (x must be increasing if s > 0)

ramorri

Hi iolite team,

I am trying to use sum normalization (as internal std) for analyzing natural magnetite samples. I keep getting an error that I don't see an earlier post commenting on. The error says 'x must be increasing if s > 0'. If I look under my 'messages' it seems like something is wrong with my input value for sum normalization (I used 100). I will post a screenshot of the error. Any suggestions on what I might be doing wrong?

Thanks!
Rebecca

ramorri

Ok, I think I figured out what I was doing. Because I drilled spots, I assumed a consistent fractionation effect for each ablation (whoops), and had selected a 'fractionation' (spline) for each element analysed. I have since removed the 'fractionation correction' on each element, and am now going through and seeing if the correction needs to be applied (most of them do not). Basically working through what the 3D trace element webinar goes over around 1:05 (1 hr 5 mins onward), and will leave this post up in case others run into this same issue in the future.

thx, R

Bence

Hi @ramorri

Thanks for bringing this to our attention. If you wouldn't mind sharing your experiment with us, we'll use it to make a more user-friendly message that explains when downhole fractionation cannot be applied. We won't share the experiment with anyone... it'll be just to improve the user experience.

Thanks again,
Bence

ramorri

Yes of course, I will email it to you. I also just noticed that it also only gives this error when I am treating reference materials the same way as the unknowns with sum normalization as an internal std. However if I tell the reference materials to use a known internal std value (say Ca42) and only use sum normalization for unknowns, then I do not get the error. Weird (?). Maybe that's the better question I should be asking. Can I use different internal std techniques (i.e., sum normalization for unknowns, and known IS values for Ref Materials)? I would think so given I tried that and the data looks good!

Cheers, Rebecca