| T O P I C R E V I E W |
| ladyfloflo |
Posted - 08/04/2010 : 3:47:16 PM
Operating System: 8.1 SR3 V8.1.34.90
I am trying to fit my data with this known equation but it says that the fit did not converge though the equation did compiled!
P= exp((t-A)/(B+G*(t-A)))/(exp((t-A)/(B+G*(t-A)))+(exp((-H*(t-A))/B)*((1-Ps)/Ps)))
Where P is the dependent variable and t is the independent variable
the parameters are
A,B,G,H, and Ps.
I also don't know where and how to write the script or code for initializing the parameters and/or the script before the fitting
I wish the equation above to have these criteria:
t> A-(B/G)
P=0 for t<=A-(B/G)
the range of P is
P>=0 for all t>=0
For H>0,
P approaches the limit of unity at t becomes very large
At time t=A P=Ps
I also put some initial values for the parameters which I have calculated and still does not converge! What can I do? Or what is the problem???
Any help would be greatly appreciate I can email some data points too.
florence.janvier@gmail.com |
| 1 L A T E S T R E P L I E S (Newest First) |
| larry_lan |
Posted - 08/04/2010 : 4:41:14 PM
Maybe these links can help:
The Fitting Process -- Convergence, Tolerance and Dependencies
Parameter Initialization
General Linear Constraints
Thanks
Larry |
|
|
