|
Hideo Fujii
USA
1582 Posts |
 Posted - 02/11/2013 : 2:56:29 PM
|
Hi Chirality899,
Say, you have a linear fitting function: y=a+b*x , the x intercept (=x0) would be x0 = -(a/b),
which you can regard it as a derived parameter. Although you can calculate the
standard error of this derived parameter by the error propagation rule, maybe a convenient
and easy way may be that you define the above fit function with a derived parameter, x0
for the non-linear fitter. So,
1) Using Fitting Function Builder, define y=a+b*x with a derived parameter, x0=-(a/b) .
2) Run NLFit using this function. The report sheet shows the standard error of x0.
(See the screenshot below.)
Hope this helps.
--Hideo Fujii
OriginLab
|
 |
|
Calculating standard error of x intercept
