Closed hgomersall closed 10 years ago
Not a very clear initial comment! It seems the current confidence function doesn't actually return what is described in the paper the code is ostensibly representing. This might be intentional, but if not, (something like) code as pushed.
Also, added epsilon as an explicit parameter and included the confidence in the computation of qtilde.
Merged! Thanks :).
... 2012, eqn 14 - this prob needs checking. Also, actually including the confidence paramter that was omitted from qtildematrices.