[pymvpa] NiftiDataset mask: is it OK to use not discrete binary mask?

[Michael Hanke]

Hi,


On Mon, May 11, 2009 at 06:54:52PM +0200, Vadim Axel wrote:
> Hi,
> 
> My mask is based on exported SPM contrast definition, so instead of 0 and 1,
> it contains 0 values for exclude and range of values 0<value<1 for include.I
> expect that all non-zero voxels to be included. Indeed, based on number of
> selected features in my dataset I get what I expect. I just wanted to make
> sure that this is by design and I wasn't just lucky.

Yes, this is the way it is intended to work.

Michael


-- 
GPG key:  1024D/3144BE0F Michael Hanke
http://apsy.gse.uni-magdeburg.de/hanke
ICQ: 48230050