This question concerns equations in the comment statements in the spm_spm.m
file. In the documentation of the structure xVI, the field xX.V is listed
as an optional non-sphericity matrix. First, is V part of Vi or part of X?
Is that comment incorrect? The equationi is Cov(e) = sigma^2*V.
Farther down in the comment section, there is a paragraph that starts "In
some instances ..." The end of the paragraph reads:
W is stored in xX.W and cov(K*W*e) in xX.V. I do not see how the two
covariance expressions fit together.
Thank you for any information / explanations.
Sincerely,
Linda Seltzer
|