polyFit[n, {xdata, errX}, {ydata, errY}, #, opts] finds the polynomial expression of the form y = a + b x + c x^2 + ... that fit best with 'xdata' and 'ydata' (it's a least-squares fit) 'n' is the degree of the polynomial, 'errX' and 'errY' are the errors of 'xdata' and 'ydata' (they're optional) '#' is the output format (graphical or numerical) The chief advantages in confront of Fit[] are : - the graphical output - error of the parameters - chi-square test and linear-correlation coefficient - error propagation The function includes a complete and well tested management of input error and own messages. All the details displayed by ?polyFit (after evaluating the function)