Compute the scaled condition number of a rectangular matrix.
Usage
scaled.condition(x, scales = FALSE)
Value
The columns of a rectangular matrix x are equilibrated (but not centered),
then the scaled condition number is computed following the guidelines of Belsley (1990).
If requested, the column scalings are returned as the attribute 'scales'.
Arguments
x
a numeric rectangular matrix.
scales
a logical value indicating whether the scaling factors that allow balancing
the columns of x should be returned by the function.
References
Belsley, D.A. (1990).
Conditioning Diagnostics: Collinearity and Weak Data in Regression.
Wiley, New York.