Hi guys, I've been trying to get the Nonlinear Optimization module (Levenberg Marquardt) in Eigen to work and I'm running into a little trouble.

The examples listed in the tests all have preentered data hardcoded in the objective functions. How do you pass in a data set into the function e.g. at runtime?

For instance, lets say I was making a program that fits a curve to a set of points that I read in from a file (defined at user prompt). Much appreciated, thanks!

When you create your functor class with operator() and df(), you can simply define another function in that class, say initDataPoints(...) which will set all the required data; for your case, these are the data points you read from a file.

Thanks, I ended up adding member objects to do the job

struct Objective_functor : Functor<double>
{
Objective_functor(int n_par, int m_dat, vtkPolyData *BdryElms, vtkPolyData *LRS) : Functor<double>(n_par,m_dat) ,myBdryElms(BdryElms), myLRS(LRS) {}
//Put data objects here
vtkPolyData *myBdryElms;
vtkPolyData *myLRS;
int operator()(const VectorXd &x, VectorXd &fvec) const
//fvec evaulation {....
..}
}

//Calling the instance
Objective_functor functor(n_par,m_dat,BdryElms,LRS_post);
...