| T O P I C R E V I E W |
| kamitengo |
Posted - 01/24/2013 : 02:15:02 AM
Origin Ver. and Service Release: 8.5
Operating System:windows 7
I have a set of data going to be fit in function y=a*x^b
as shown in fig.1
So I selected the built in function allometric1 as the fitting
function. As a result of proper initialization, we had well fitting
curve immediately without clicking iteration.
However, after I chosen iteration, the results were getting worse.
The fitting curve will be converged as the following figure.
I cannot figure out how'd this happen.
Does anyone can give me some suggestions?
Many thanks in advance.
|
| 1 L A T E S T R E P L I E S (Newest First) |
| Hideo Fujii |
Posted - 01/24/2013 : 10:55:15 AM
Hi kamitengo,
I suspect that visually strange, but numerically it probably makes sense because your Y scale is Log10-base.
Data points near low boundary contribute little to the Chi^2. I see the Chi^2 value is huge, and immediately
iteration has stopped one step after initial parameters. (You can change the Y scale to linear to see the
"real" fitting situation.)
I suggest you convert your Y values to their log values, then apply to the appropriate function.
Hope this helps.
--Hideo Fujii
OriginLab |
|
|
