[FieldTrip] ft_timelockgrandaverage

[Zizlsperger Leopold]

Dear Fieldtrippers,

in a within-subjects design I use ft_timelockgrandaverage with cfg.keepindividual = 'yes' to average timelocked EEG data (10 subjects) of the form:

 avg: [29x5000 double]
          var: [29x5000 double]
         time: [1x5000 double]
          dof: [29x5000 double]
        label: {29x1 cell}
        trial: [172x29x5000 double]
       dimord: 'rpt_chan_time'
    trialinfo: [172x28 double]
          cfg: [1x1 struct]

I stick closely to the FieldTrip tutorial for cluster permutation statistics. I use the newest version of fieldtrip. After the grandaverage there is no more 'avg' in the structure, so I can not go on in the tutorial e.g. with: 
GA_XvsY = GA_X;
GA_XvsY.avg = GA_X.avg - GA_Y.avg;

New data is of structure:

label: {29x1 cell}
          time: [1x5000 double]
    individual: [10x29x5000 double]
        dimord: 'subj_chan_time'
           cfg: [1x1 struct]

Do I get the tutorial wrong or is it my data ?
Thanks in advance
Best 
Leopold