Drexel dragonThe Math ForumDonate to the Math Forum


Math Forum
 Ask Dr. Math
 Discussions
 Internet Newsletter
 MathTools
 Problems of the Week
 Teacher2Teacher
 Teacher Exchange
 Workshops

Search All of the Math Forum:
Browse our Internet Mathematics Library

Views expressed in these public forums are not endorsed by Drexel University or The Math Forum.


Math Forum » Discussions » Software » comp.soft-sys.matlab

Topic: Fitting and correlation oddity
Replies: 1   Last Post: Aug 3, 2000 3:33 AM

Advanced Search
Reply to this Topic Reply to this Topic

Back to Topic List Back to Topic List Jump to Tree View Jump to Tree View   Messages: [ Previous | Next ]
Simon de Vet

Posts: 19
Registered: 12/7/04
Fitting and correlation oddity
Posted: Aug 2, 2000 12:05 PM
  Click to see the message monospaced in plain text Plain Text   Click to reply to this topic Reply

I've gotton my fitting working, but have found an oddity.

If I fit data of the form z=Ax+By, for sets of x and y values, I get one
set of coefficients (44 and 56).

However, if I fit the data in the form z=Ax+Bx, I get the coefficients
100 and 0, as expected, since both components are identical.

The weird part is that the correlation value for the x and y model is
0.99245. For the only x model, the value is 0.99473.

If having no y contribution is a better fit than having a y
contribution, why aren't the coefficients 100 and 0 in both cases?


Simon

Point your RSS reader here for a feed of the latest messages in this topic.

[Privacy Policy] [Terms of Use]

© Drexel University 1994-2013. All Rights Reserved.
The Math Forum is a research and educational enterprise of the Drexel University School of Education.