polyfit_roots

Similar to polyfit, but allows higher degrees by returning the roots of the polynomial.

POLYFIT_ROOTS Least-squares polynomial fit to data.
[R,K] = POLYFIT_ROOTS(X, Y, N, TOL) finds the roots R and constant K so that the polynomial P(s) = K*(s-R(1))*(s-R(2))* ... *(s-R(N)) is the best least-squares fit to the data Y at points X.

Argument TOL bounds the accuracy of the fit, and if ommited is taken to be 1e-14. A polynomial of degree n < N is returned if it fits the data with error less than TOL.

See polyfit_roots_drv.m for examples.