[FieldTrip] Notification for users with different channel ordering between participants

[Herring, J.D. (Jim)]

Dear Fieldtrip users,

 

It has come to our attention by Philip Ruhnau that when averaging across
participants using ft_XXXgrandaverage, due to a bug, no care was taken
when the channel-order was different between participants (see
http://bugzilla.fcdonders.nl/show_bug.cgi?id=1986). This could occur, for
example, when after interpolating channels, the interpolated channels are
added to the end of the data and thus changing the channel ordering.
Particularly for EEG data this bug could be a problem because often
different channels are interpolated for different subjects. No problems
should have occurred when merely removing channels.

 

For MEG data, the same problem occurred when using ft_combineplanar after
interpolating, for example, vertical or horizontal channels separately. In
this instance the wrong channel pairs were combined. However, if channels
were interpolated after using ft_combineplanar, or if pairs of horizontal
and vertical channels were removed you should have no problems given that
the same channels were affected across all participants. 

 

The bug has been fixed as of rev. 7519. We strongly advise users that have
data containing different channel ordering between participants to
download the newest version of Fieldtrip and re-check grand averaged data.

 

Best,

 

Jim

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20130221/919dd238/attachment-0001.html>