topic Re: How to automatically recognize the curvature difference? in Discussions

How to automatically recognize the curvature difference?

Fruit325 — Mon, 15 Apr 2024 11:48:43 GMT

I have the datalogs where tempeature readings are collected continuously as time goes by for each wafer. Temperature profile looks like a mountain.

I collected the temperature readings from 10 wafers and I want to compare the temperatura profiles together by illustrating the temperature readings along the same (relative) time axis. I want to check if some of them are normal or not. Some of the mountains are overlapped and some of them are not. Some have different peaks or starting points. Is there any way to differentiate between the shape difference (mountains) instead of differentiating them by naked eye visualization...? Thanks for any suggestions!

Re: How to automatically recognize the curvature difference?

jthi — Mon, 15 Apr 2024 12:59:51 GMT

Which JMP are you using (standard or pro) and which version? If you have JMP Pro check out Functional Data Explorer

Re: How to automatically recognize the curvature difference?

MikeD_Anderson — Fri, 19 Apr 2024 13:58:12 GMT

@Fruit325 - there are several ways to handle this... is there any way you can post an anonymized screenshot of a graph so we can help you determine the right one?

Re: How to automatically recognize the curvature difference?

Fruit325 — Mon, 22 Apr 2024 15:21:39 GMT

Hi Mike,

Thanks! Here is an example:

We may differentiate the two curves from the peak.

Is there any way we could differentiate the two directly? Can we calculate the peak area? or locate the peak?

There are a lot of curves in our case and some of them may be overlapped or not, or have the same peak values.

Thank you!

Re: How to automatically recognize the curvature difference?

Fruit325 — Mon, 22 Apr 2024 15:23:11 GMT

Unfortunately, I'm using JMP Pro.

Thank you so much, I will check the data explorer you shared.

Re: How to automatically recognize the curvature difference?

MikeD_Anderson — Tue, 23 Apr 2024 19:31:50 GMT

Hi, @Fruit325 !

Thanks for that. You've got a few options on how to handle this. I actually did a presentation on this topic at the JMP Discovery Summit last year and there is a recording. You can find it here: https://community.jmp.com/t5/Discovery-Summit-Americas-2023/Spectris-Spooky-Spectroscopy-2023-US-30MP-1472/ta-p/651680

Have a look at that and reach out if there are any questions!

Best,