Mult : Fits linear data to equations using multiple
regression with regression
diagnostics and errors on all variables.
Mult solves linear least squares problems and allows errors in both y and x variables.
It also provides regression diagnostics to aid in identifying harmful data and outliers.
The method (General Least Squares) is as given in Ghilani & Wolf: Adjustment Computations
(4th Ed 2006 Wiley) but is extended to use singular value decomposition for stability.
It may also prove useful for mass balance problems in petrology.
Versions available:
These downloads hold executables, sample input files and documentation explaining the use of the program.