Chapter 3. Measured Data and Parameter Estimation


In addition to using IMD for modeling, you can also use IMD for parameter estimation (with confidence interval generation) using nonlinear, least-squares curve-fitting, based on the Chi^2 test of fit.

To illustrate, I will present an example showing how to determine the optical constants of a film from reflectance vs. incidence angle measurements.


Note: In IMD you can choose to use either the IDL CURVEFIT function (actually, a slightly modified version,) which is based on the Marquardt algorithm [14] described in reference [15], or the MPFIT function, an implementation of the MINPACK-1 Levenberg-Marquardt algorithm [19], developed by C. B. Markwardt <http://astrog.physics.wisc.edu/~craigm/idl/idl.html>


3.1 Using your measured data

3.2 Specifying fit parameters

3.3 Curve-fitting

3.4 Confidence intervals


Back | Contents | Next