Applying cumsum correctly to follow the visualization paper exactly
rdangovs opened this issue · 0 comments
rdangovs commented
Lines 114 to 122 in 49698ef
cumsum
function takes [theta_0, theta_1 - theta_0, ..., theta_n - theta_{n-1}] and returns [theta_0, theta_1, theta_2, ..., theta_{n-1}, theta_n]. We should subtract theta_n, no? I did that and this is the result I got. Also, the latest version of master doesn't apply cumsum
before computing the PCA directions.