[pymvpa] OneWayAnova for feature selection: how it works?

[Vadim Axel]

Hi,

I somewhat struggle to understand the logic and how it works. Here what is
written in the header of the file: "F-scores are computed for each feature
as the standard fraction of between and within group variances. Groups are
defined by samples with unique labels." (measures/anova.py)

Suppose I have 100 voxels, 2 conditions, 20 trials per condition. Suppose I
want to reduce the number of my voxels (features) to 50.  I can run one-way
ANOVAs for each voxel, where conditionID is the factor (two levels). But
this does not seem to be different from paired t-test. Also, this would be
clearly univariate selection.  In one-way ANOVA with data analysis we might
have subjects as between-group factor and treatment as within-group factor.
But treating voxels as "subjects" does not help me because there is no
grouping for voxels (grouping is for condition).

Any ideas? Any links for explanation would be also appreciated!

Thanks,
Vadim
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.alioth.debian.org/pipermail/pkg-exppsy-pymvpa/attachments/20140505/a397dcd1/attachment.html>