[pymvpa] ROC

[Yaroslav Halchenko]

> OK, I took a look.  I don't think they should be the same.  An ROC
> assumes that if you shift your criterion all the way to one side you
> get 0 false and true positives and to the other side you always get
> true and false positives.
correct ;-)

> In other words, you have to pad the points.
not necessarily... like in my examples... for crit>0.5 you get 0 tp, and
0 fp, for crit<0.5, you get 1.0 for tp, and 1.0 for fp ;-) (I leave it
up to anyone to decideto which outcome you assign when crit=0.5)

>  Also, depending on the number of points you have in the middle, your
> estimate should will change.
correct... but in the given cases there is no points in the middle ;)

-- 
Yaroslav Halchenko
Research Assistant, Psychology Department, Rutgers-Newark
Student  Ph.D. @ CS Dept. NJIT
Office: (973) 353-5440x263 | FWD: 82823 | Fax: (973) 353-1171
        101 Warren Str, Smith Hall, Rm 4-105, Newark NJ 07102
WWW:     http://www.linkedin.com/in/yarik