cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 
Choose Language Hide Translation Bar
aliegner1
Level IV

how to plot trace overlays when some steps are not the same length?

Looking for advice or help on how I can get this to plot correctly. I'm trying to do some sensor trace overlays, but the problem is that some of the steps have a dynamic end point, meaning the time length can change. If I try to overlay traces from time=0 at start of the recipe, each wafer trace gets slightly offset. OK, so I can get a column for time=0 at the start of each step, that resets back to zero at each step.

 

How can I get this to plot, so that the chunks of the trace all line up at the start of each step? I can't seem to get anything working in graph builder. I currently have the data as a TALL table

 

 

example: how it currently looks....slightly offset

aliegner1_3-1649773618957.png

 

Ideal state, lining up by step#

aliegner1_2-1649773066665.png

 

 

4 REPLIES 4
jthi
Super User

Re: how to plot trace overlays when some steps are not the same length?

Col Max(), Col Min() and Col Cumulative Sum() (or some else Col statistical functions) with correct by variables should get you quite close to what you are looking for.

-Jarmo
aliegner1
Level IV

Re: how to plot trace overlays when some steps are not the same length?

I'm not sure I follow what you are meaning.

Re: how to plot trace overlays when some steps are not the same length?

Do you have JMP Pro? If so, you might try using the Functional Data Explorer. I offers a lot of tools for dealing with anomalies in the data. For example, dynamic time warping would align all the curves for you.

Re: how to plot trace overlays when some steps are not the same length?

Hi @aliegner1 I am happy to help you out.  Functional data explorer is most likely the easiest solution and using the aligned steps in that platform.  We actually have a talk on FDE coming up on Friday in Mastering JMP.  https://www.jmp.com/en_us/events/live-webinars/mastering-jmp/understanding-and-modeling-response-cur... You should check that out and if you have any questions don't hesitate to reach out to me.