[pymvpa] Q: Time-offset of category labels in Haxby 2001 dataset. Was this ever properly resolved?

[Michael Hanke]

On Tue, Mar 08, 2011 at 12:12:01PM -0500, Yaroslav Halchenko wrote:
> and first initial rest block in each chunk having 6 volumes (I think).  So, if
> we forget for a moment that we got 121 volumes instead of 120, then initial
> rest block should have indeed been allocated  5 volumes (although once again
> not exactly, due to 12/2.5 = 4.8) instead of 6 volumes.  Thus you could
> probably consider indeed that all labels in the dataset are shifted by 2
> volumes, and I should have not added 1 volume on both sides of a stimuli block,
> but rather 2 on the front, if I was looking for "precise" timing in terms of
> volumes.   But then you would obtain 7 rest volumes at the end of each run
> (121st volume?) That would be my speculation atm

To add another tiny bit of evidence for a shift you could take a look at
the peristimulus plot at http://www.pymvpa.org/tutorial_eventrelated.html

The block onset as encoded in the label vector is the t=2 volume/timepoint in
the plot. You'll see that this volume already reflects a deviation from
baseline for both plotted conditions. Therefore, I'd say that, on
average, the labels are shifted by one volume -- considering all
temporal uncertainty introduced by the timing of the study that Yarik
mentioned before.

Michael

-- 
Michael Hanke
http://mih.voxindeserto.de