ABSTRACT
In the present chapter we develop the general theory of linear models. The theory will be applied to various special cases in the subsequent chapters.
Our development of linear models depends on taking a geometric approach, thinking of the vector Y = (Yi, . . ., Yn)T of observations as a point in the vector space R n , and similarly fi = ( / / i , . . . , ^ n ) T , the mean vector for Y, as a vector in R n . We use a column vector notation, so /i and Y are n x 1 vectors.
