Question: Non-linear fit problem

Hello,

I want to fit Data with a model that calculates the values by solving an ODE numerically.

solL:=dsolve({Lpde,ICS}, numeric, method=rkf45, parameters=[lambda[0],T[m]]);

Next, I created a procedure.

ans:=proc(p1,p2) solL(parameters=[p1,p2]); rhs(solL(3600)[2]); end proc;

This is giving a length L(t) depending on parameters lambda[0] and T[m]. I would like to fit this model with my data. These data are given for a fixed time t=3600 and are pairs of different length L depending on different T[m]. So for the fit T[m] should be my independent variable and not t anymore. Following this topic I tried this:

LVector:=Vector([179e-9,200e-9,221e-9,233e-9,228e-9,222e-9,204e-9,202e-9]);
TVector:=Vector([623,643,683,703,723,743,763,783]);
FitParams:=Statistics:-NonlinearFit(ans(lambda[0],TT), TVector, LVector, initialvalues=<150e-9>, 
output=parametervalues);

Unfortunately, the ODE is solved with respect to t but the fitting should be done with respect to a parameter in the dsolve command. This gives an error:

Error, (in dsolve/numeric/process_parameters) 'parameters' must be specified as a list of numeric values or equations of the form <parameter>=<value>

What can i do here?

 

Please Wait...