| T O P I C R E V I E W |
| MRI_aic |
Posted - 10/01/2012 : 12:14:37 PM
I am trying to fit two dependent curves like:
y = exp(-(a+b)*x) & z = b/d*exp(-a*x)*(1-exp(-d*x))
with parameters a,b & d but I have the following error message:
(11) Parameter Initialization was called.
(12) ----------Levenberg-Marquardt-----------
Reduced Chi-sqr = --
Iterations Performed = 0
Total Iterations in Session = 0
(13) Fit did not converge - zero or missing weight value is not allowed for chosen weight method.
Anybody any idea?
Thanks! |
| 2 L A T E S T R E P L I E S (Newest First) |
| easwar |
Posted - 10/01/2012 : 5:32:56 PM
Hi,
Looks like you have assigned Weights during fitting, and the data has zero or missing values.
Weighting methods such as Instrumental assign reciprocal weights such as 1/value^2, so if any of the values are 0 or very close to zero, or missing, fitting cannot proceed.
Easwar
OriginLab |
| Hideo Fujii |
Posted - 10/01/2012 : 4:18:43 PM
Hi MRI_aic,
I suspect that the initial values of the parameters, a, b and d are not close enough to lead the convergence, though not sure.
Maybe first you can try to define one-dependent-variable function separately, and make sure each works? Once they work surely, then you can come to the two-dependent-variable function.
--Hideo Fujii
OriginLab |
|
|
