| Author |
 Topic  |
|
|
muller
France
3 Posts |
 Posted - 04/10/2003 : 12:26:46 PM
|
Hello,
I have to search 3 parameters a, b and c to fit a function F(x).
F(x)= a*f(x)+b*g(x)+c*h(x).
F(x), f(x), g(x) and h(x) are a set of data. what i want to do is for every x, i have to found the same parameters a, b and c to minimize the quantity : F(x)-(a*f(x)+b*g(x)+c*h(x)).
I've tried to do it in several ways. But, I do somthing wrong.
I need some help.
thamks |
|
|
Hideo Fujii
USA
1582 Posts |
Posted - 04/14/2003 : 12:04:42 PM
|
Taking a simpler sample such that with given data:
Data1: A(X)=1,2,3 B(Y)=0,1,2
Data2: A(X)=1,2,3 B(Y)=1,1,1
Data3(to be fit): A(X)=1,2,3 B(Y)=2.1, 3.1, 4.1
Try:
1) Open NLSF
2) Create a new user-defined function: y=P1*data1_b(x)+P2*data2_b(x); here P1 and P2 are parameters.
3) Fit Data3 as usual.
Note that if the x range of Data3 is out of Data1 or Data2, liner extrapolation will be done.
Does it make sense?
--Hideo
Edited by - Hideo Fujii on 04/14/2003 12:07:27 PM |
 |
|
| |
Topic |
|
|
|
Fit a dataset with a combine of other datasets