The Origin Forum
File Exchange
Try Origin for Free
The Origin Forum
Home | Profile | Register | Active Topics | Members | Search | FAQ | Send File to Tech support
 All Forums
 Origin Forum
 Origin Forum
 Problems fitting nonlinear data

Note: You must be registered in order to post a reply.
To register, click here. Registration is FREE!

Screensize:
UserName:
Password:
Anti-Spam Code:
Format Mode:
Format: BoldItalicizedUnderlineStrikethrough Align LeftCenteredAlign Right Horizontal Rule Insert HyperlinkUpload FileInsert Image Insert CodeInsert QuoteInsert List
   
Message:

* HTML is OFF
* Forum Code is ON
Smilies
Smile [:)] Big Smile [:D] Cool [8D] Blush [:I]
Tongue [:P] Evil [):] Wink [;)] Clown [:o)]
Black Eye [B)] Eight Ball [8] Frown [:(] Shy [8)]
Shocked [:0] Angry [:(!] Dead [xx(] Sleepy [|)]
Kisses [:X] Approve [^] Disapprove [V] Question [?]

 
Check here to subscribe to this topic.
   

T O P I C    R E V I E W
lopes_santista Posted - 04/29/2006 : 1:42:14 PM
Dear friends.

I'm having some problems trying to fit a nonlinear curve in Origin. I've tryed to fit the "BoxLucas2" equation and I only get error messages from the program (sometimes it asks me to fix parameters until only one is able to vary; sometimes it says that the function is not generating any value).
I wonder if someone can help me. I'm copying the data I want to fit.
Thanks
Rafael

x
2,32E-06
2,71E-06
3,26E-06
3,95E-06
5,75E-06

y
1,722E-02
5,206E-02
4,809E-02
3,472E-02
0,000E+00
1   L A T E S T    R E P L I E S    (Newest First)
larry_lan Posted - 04/29/2006 : 11:43:21 PM
Hi friend,

The reason that ask to fix the parameter is that the initialized parameter value seems not so good. You can adjust the initialize value manually.

I have tried your data to fit the "BoxLucas2" equation, and using the system initialized parameters' value, it do generate the error message sometimes, so it indicates that the parameter value is bad, or the build-in BoxLucas2 model may not be a good evaluation for your data.

Look at the general form of a downwards open curve of Box Lucas2 model:

Noticed that, in most cases, the upper limit of y is 1, and your data upto 52. And the peak is located nearby zero, while the peak value of your data is 2.71E-4. Though it is a very small value, but it do means.

So in order to fit your data, you can use another model.

Or if you do want to fit the data by Box Lucas2 model, I think there may be two solution:

1. Since the model do not fit the raw data, you can adjust some of the points to make the plot more reasonable.

2. Adjust the model slightly.
Improve the model by horizontally move the curve to the right side a little, and add a coefficient for the curve. The original function of Box lucas2 model is:

y = (a1/(a1-a2))*(exp(-a2*x) - exp(-a1*x))

Then I defined a user-defined function as:

y = B*(a1/(a1-a2))*(exp(-a2*(x-xc)) - exp(-a1*(x-xc)))

where xc is the horizontal movement, and B is the amplitude.

Then the fitting will generate a better result.



Hope it helps.

Larry
OriginLab GZ Office

Edited by - larry_lan on 04/30/2006 12:20:36 AM

Edited by - larry_lan on 04/30/2006 12:42:23 AM

The Origin Forum © 2020 Originlab Corporation Go To Top Of Page
Snitz Forums 2000