<t>I want to do principle component analysis on the residual vector, to attribute the un-explained variance to a number of "missing" factors from the regression.Now to do this in Matlab you need matrix (M*N where N>1), so my idea was to construct a matrix from the residual vector. This would produce...