Inconsistent definition of quad_form

[daltonhance]

In the bookdown document you have that Q* = RQR', with R being m x q and Q being q x q. This is consistent with Durbin and Koopman. But Stan defines quad_form as B' * A * B, rather than B * A * B'.

Pretty minor stuff but thought I'd point it out.

Really great work here. Thanks!