In this lab, each student is to develop code to calculate a nonlinear regression t. Note you may not call a single high-level function in MATLAB or another package to do the tting; you must implement the root nding method discussed in class. The code can be developed in MATLAB, C, or any high level language.
The tting should be run on the following three data les log-data-A, log-data-B and log-data-C, found at: http://www.cecas.clemson.edu/~ahoover/ece854/labs/. In each le the two columns of the data are xi, yi.
The data follows a problem outlined in class and in the lecture notes. Fit a function of the form y = ln(ax) where a is the unknown.
The report should derive the math required for the nonlinear regression. It should show plots of the data for each le. It should also report the initial value of a used to start the regression, nal value of a found at the end of the regression, and number of iterations needed to converge. You are encouraged to try di erent initial values of a to observe its e ect on convergence.
The lab due date is given at the class web site. You must submit your code (as an attachment) and report (as an attachment) to ece_assign@clemson.edu. Use as subject header ECE8540-1,#2. This email is due by midnight of the due date.
1
