Given a data set in a two-dimensional feature space, suppose e_(1) is the first principal component for the data set which has the largest variance of the projected data. What is the second principal component?