Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.

- JMP User Community
- :
- Discussions
- :
- Duration of Segment

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page

Highlighted

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Mar 31, 2020 5:35 AM
(2852 views)

I have a data set with many profiles like the image below. Each profile varies some on the slope and duration of the peak. My goal is to find a way to measure the duration of the peak portion of the profile. Can someone help me with this?

1 ACCEPTED SOLUTION

Accepted Solutions

Highlighted
I ran a summary of the table with the range of time and grouped by part number and peak or slope. Then I joined the summary table with the original table.

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

7 REPLIES 7

Highlighted
##

I would need some more information. First, are you trying to predict the duration of the peak, when it begins, or both? Second, what kind of data do you have to predict this with (i.e., how many variables, what types, etc. - a simple example would be good to see). Or, are you just wanting to measure the peak period in data that you have, without predicting anything? Third, once the peak occurs, does it stay at that level or does it decline at some point? Relatedly, are there any interim peaks, where the slope begins again?

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Re: Duration of Segment

Highlighted
##

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Re: Duration of Segment

1) I am trying to measure the duration of the peaks so that I can then calculate the area under the curve for the different profile.

2) I attached a portion of my data table if that is helpful. I also attached a few more images of my data to show some variations in the profile heating for different parts.

3) There is some variation in the peak value. That has been one of the difficulties with finding the duration.

Highlighted
##

I can't help you as much as I'd like but perhaps someone else can - I think I can get you/them started, however. I'm not very good with time formatting, so my meager attempts have not been successful. But I do think you can use the Col min and max functions to get the starting and ending times for the peaks (once you get the time stamp formats correct - which I was not able to figure out). Then you can use the difference between these times to measure the length of the peak. I wasn't sure whether you wanted to do this by Part Number or some other characteristic, but you can use that as a By variable in the column min and column max functions, and column min and column max do work for time values (once you get them formatted the way you want). Then, it looks to me like you can estimate the area under the curve (your Bivariate Fit graph) using a linear approximation to the slope portion, along with the rectangle under the peak portion.

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Re: Duration of Segment

Highlighted
##

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Re: Duration of Segment

I will let other members who are smarter than me answer your specific question. I have a question for you. What will you do with the answer? Will the time be a response or a predictor / factor in another analysis or model? What will that analysis be?

I ask because a lot of people have functional data. They try to extract features for further analysis. There is a JMP Pro feature called the Functional Data Explorer that can make extraction unnecessary. It works with the entire function / profile / curve. The unique characteristics of each function are captured as functional principle components. The essential FPCs can now be profiled or used as a response or factor in another JMP platform such as ANOVA or regression.

Learn it once, use it forever!

Highlighted
##

I am going to calculate the are under the curve. Is there a way to get area under the curve with FDE?

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Re: Duration of Segment

Highlighted
##

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Re: Duration of Segment

No, not directly.

Learn it once, use it forever!

Highlighted
I ran a summary of the table with the range of time and grouped by part number and peak or slope. Then I joined the summary table with the original table.

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Article Labels

There are no labels assigned to this post.