| Author |
 Topic  |
|
|
erickzhou
17 Posts |
 Posted - 07/20/2010 : 10:10:16 AM
|
Origin Ver.8 and Service Release (Select Help-->About Origin):
Operating System: WINXP
During the iteration process of Least square method for surface or curve fitting, absolute or relative error can be selected as the stop criteria. In the chapter of "goodness of fit"(http://www.originlab.com/www/helponline/Origin/en/UserGuide/Goodness_of_Fit.html), it seems that absolute error was selected.
But in the chapter of "Illustration of the Least Squares Method", it seems that relative error was choosed, i.e., the first equation to calculate RSS in "http://www.originlab.com/www/helponline/Origin/en/UserGuide/Illustration_of_the_Least_Squares_Method.html". But I don't know that what is the meaning of the parameters "wi", and how to calculate it?
Why choose different error in these two chapter, i.e. "goodness of fit" and "Illustration of the Least Squares Method"?
Wait for your early reply, thank you! |
|
|
larry_lan
China
Posts |
Posted - 07/20/2010 : 10:35:02 AM
|
Wi is the Weight. Please search our Help document to see how to include weighting during fitting.
We don't want to make the example complicated. So, usually, we suppose fitting with no weights when explaining some basic concept.
Thanks
Larry |
 |
|
|
erickzhou
17 Posts |
Posted - 07/20/2010 : 10:32:42 PM
|
Thank you. Is there any reference about how to include weighting during fitting?
|
|
|
| |
Topic |
|
|
|
Absolute or Relative Error in the fitting process?