| T O P I C R E V I E W |
| JinsungKim |
Posted - 05/31/2012 : 02:53:23 AM
Origin Ver. and Service Release (Select Help-->About Origin): Origin 8 SR 4.0
Operating System: MS Windows 7
Origin Ver. and Service Release (Select Help-->About Origin): 8.0 SR4
Operating System: MS Windows 7
Hi guys,
I am now dealing with multiple linear regression.
Though I can manage how to run mulutiple linear regression in Origin, I wonder whether there are any optional functions which work to boundary conditions among coefficients of independent variables.
I.e., consider multiple linear regression below.
Y = a + bx + cz + e where Y is dependent variable, x and z are independent variables, a is constant, e is random error term, b and c are coefficient of interest.
Normally, Origin greatly compute best value of b and c but I would like to add one more condition to about b and c, such as b+c = 1.
Therefore, total algebraic system would be as following.
Y = a + bx + cz +e, b + c = 1.
Is Origin possible to find b and c with multiple linear regression method under certain condition (b+c=1)? Is Origin possible to lay certain condition in multiple linear regression dialogue window?
Thanks to your help!
Jinsung |
| 1 L A T E S T R E P L I E S (Newest First) |
| Hideo Fujii |
Posted - 05/31/2012 : 12:03:52 PM
Hi Jinsung,
Though the multiple regression tool doesn't have linear constraint option, instead, you can
define a user-defined fitting function with multiple independent variables; and you can use
the non-linear fitting tool, NLFit. In NLFit, you can set linear constraints like yours.
As for the way to set the linear constraints, you can refer to the following tutorial page:
http://wiki.originlab.com/~originla/howto/index.php?title=Tutorial:Fit_Curve_Through_Certain_Points#Use_Linear_Constraint
Hope this helps.
--Hideo Fujii
OriginLab
P.S. There is a video tutorial, "Origin: Using Linear Constraints When Fitting":
http://www.youtube.com/watch?v=Fh7broKA8AU |
|
|
