There have been a few questions about the set-up for problem 2,
which says:
X is an (k x n) matrix of x_{ji} entries indexed by j = 1, ...,
k and i = 1, ..., n. (I chose this so that ji would correspond
to kn -- that is, the row subscript would come first in both the
matrix and scalar notation.)
A few perceptive individuals noted that the derivation from
section relied on X being an n x k matrix. Although it's
possible to get the correct answer using the original set up, it
is probably easier to use the set up from section and Gary's
lecture notes:
X is an n x k matrix consisting of x_{ij} entries, where j
corresponds to columns 1, ..., k and i corresponds to
observations 1, ..., n, which means that x_i1 is 1 for all i and
draw x_i2 from the Unif(0,2)
You can do the problem either way. Hope this helps. Olivia
Show replies by date