| Author |
 Topic  |
|
|
Karmar
United Kingdom
6 Posts |
 Posted - 10/20/2021 : 06:02:05 AM
|
Hi,
How to calculate standard error (SE) for y, where
y = 1/((a/b) + (c/d))
if I know SE of a, b, c, d, and a, b, c, d are obtained from the fitting data to the same equation (z = a*exp(x/b) + c*exp(x/d), dependent error).
Thanks in advance.
|
|
|
YimingChen
1609 Posts |
|
|
Karmar
United Kingdom
6 Posts |
Posted - 10/20/2021 : 09:26:32 AM
|
| Thank you very much James. It is a great feature I didn’t know. I will use it next time, but I have already obtained the parameters from different software, so I cannot use this feature for this data. Is there any other way I can solve the problem in Origin? |
 |
|
|
YimingChen
1609 Posts |
Posted - 10/20/2021 : 12:39:27 PM
|
Then you have to calculate manually. Maybe google "Propagation of Uncertain" to find the formula for your case.
James |
|
|
|
Karmar
United Kingdom
6 Posts |
Posted - 10/20/2021 : 3:16:25 PM
|
Thanks James, I have tried Build Fitting Function, but cannot see the derived parameter (as below).
 
What I am doing wrong? |
|
|
|
YimingChen
1609 Posts |
|
| |
Topic |
|
Error propagation with dependent variables