Macroundle10.doc

************************************* * * * MacroBundle * * * ************************************* (c) R. de Levie v 10, July 2010 TABLE OF CONTENTS List of macros Purpose Warranty Copyright Installation Updates The macros LIST OF MACROS (including short descriptions) Error analysis macro: Propagation computes the propagation of uncertainty for a single function of one or more independent input parameters with known standard deviations, or for mutually dependent parameters with a known covariance matrix. Linear least squares macros: LS is a general least squares fitting routine for linear, polynomial, and multivariate fitting, assuming one dependent variable. LS0 forces the fit through the origin, LS1 does not. The output provides the parameter values, their standard deviations, the standard deviation of the fit to the function, the covariance matrix, and (optionally) the matrix of linear correlation coefficients. ELS provides least squares smoothing and differentiation for an equidistant (in the independent variable) but otherwise arbitrary function using a Savitzky-Golay moving polynomial fit. ELSfixed uses a fixed-order polynomial, ELSauto self-optimizes the order of the fitting polynomial as it moves along the function. WLS is the equivalent of LS with the inclusion of user- assignable weights. GradeBySf computes the standard deviations of the fit Sf for all possible combinations and permutations of unweighted least squares fits of a user-specified multivariate expression of up to six terms. LSPoly applies LS to polynomial fitting to a polynomial of gradually increasing order (up to 14). LSMulti applies LS to an increasing number of terms of a multivariate least squares analysis. LSPerm