[pymvpa] Category specific accuracy

[Jiang, Zhiguo]

HI, Yaroslav

Thanks for your response! i also have a question about the compuation of
accuracy. in the tutorial the ACC was given by (TN+TP)/(N+P),  so given the
confusion matrix as
[[ 6  0  3  0  0  5  0  1]
 [ 0 10  0  0  0  0  0  0]
 [ 0  0  7  0  0  0  0  0]
 [ 0  2  0 12  0  0  0  0]
 [ 0  0  0  0 12  0  0  0]
 [ 2  0  1  0  0  6  0  0]
 [ 2  0  1  0  0  0 12  1]
 [ 2  0  0  0  0  1  0 10]]
P                12            12            12            12            12
           12            12            12
N                84            84            84            84            84
           84            84            84
TP               6             10            7             12            12
           6             12            10
TN               69            65            68            63            63
           69            63
N=84; P=12; TP for first column (bottle) is 6, FP=2+2+2=6, FN=3+5+1=9, so
what I really dont understand is how does the program give 69 for TN
(shouldn't it be 96-6-9-6=75?. any idea on how it was computed?

Tony

so how does the TN=


On Sat, Jan 18, 2014 at 12:18 AM, Yaroslav Halchenko
<debian at onerussian.com>wrote:

>
>
> or did I misunderstood the question: subject talks about category
> specific ones which could be get from diagonal of stats.matrix or from
> various .stats.stats (IIRC and that is not confusing!!! ;-)), and then
> question is "for each validation run": confusion matrices per each run
> could be found under cvte.ca.stats.matrices IIRC
>
> On Fri, 17 Jan 2014, Jiang, Zhiguo wrote:
>
> >    Hi, All
> >    I am totally new to PyMVPA tool. and forgive me this beginner's
> question.�
> >    I am trying to follow the tutorials
> >    :�[1]http://dev.pymvpa.org/tutorial_classifiers.html. is there a
> simply
> >    way to pull out the accuracy for each category (bottle, cat, chair,..
> >    respectively in the worked example) for each validation run? I know
> it is
> >    all in cvte.ca.stats.as_string(description=true) but i am sure there
> is a
> >    better and more elegant way to do so .�
> --
> Yaroslav O. Halchenko, Ph.D.
> http://neuro.debian.net http://www.pymvpa.org http://www.fail2ban.org
> Senior Research Associate,     Psychological and Brain Sciences Dept.
> Dartmouth College, 419 Moore Hall, Hinman Box 6207, Hanover, NH 03755
> Phone: +1 (603) 646-9834                       Fax: +1 (603) 646-1419
> WWW:   http://www.linkedin.com/in/yarik
>
> _______________________________________________
> Pkg-ExpPsy-PyMVPA mailing list
> Pkg-ExpPsy-PyMVPA at lists.alioth.debian.org
> http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/pkg-exppsy-pymvpa
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.alioth.debian.org/pipermail/pkg-exppsy-pymvpa/attachments/20140118/5dcb5ffa/attachment.html>