Least Squares with linear and non-linear parameters

In two unrelated research project I have recently needed to do some non-linear model fitting. After some research I concluded that either a) nobody bothered to develop working equations for least-squares optimization with nonlinear parameters and without linearization in the Newton-Raphson procedure, or b) that it would be easier to derive these formulas than to excavate them from some forgotten references. It may well be that Newton-Raphson scheme for non-linear least squares is inefficient without the above-mentioned linearizations. I will probably look to it in the future. Nevertheless, deriving these equations was a lot of fun. Hope you’ll enjoy the write-up. Continue reading “Least Squares with linear and non-linear parameters”


Hello Universe!

Dear Universe,

I guess you’re not yet well mixed – let’s then have some fun before it happens.

This is gonna be (I guess, but really – who knows?) a blog about theoretical chemistry, with an occasional tidbit of physics, mathematics or computers. It is also bound to have a bunch of uncategorizable, silly stuff, such as cat images or reviews of sci-fi books.

Finally, a test of the LaTeX capabilities of this free WordPress account:

\huge S = k_B \ln \Omega