[FieldTrip] ft_combineplanar on Neuromagdata

[Frédéric Roux]

Dear fieldtrip users,

sorry to bother you with this really trivial question.
I am running into an issue using ft_combineplanar on Neuromag data.

The code I am using is as follows:

cfg = [];
cfg.channel = {'MEGGRAD'};

grad_data = ft_selectdata(meg_data);

%after this step there are only planar-gradients left

cfg = [];
cfg.method = 'mtmfft';
cfg.output = 'pow';
cfg.taper = 'hanning';
cfg.foi = 0:100;
cfg.keeptrials = 'no';

spectrum1 = ft_freqanalysis(cfg,grad_data);

% returns the FFT power spectrum

cfg = [];

spectrum2 = ft_combineplanar(cfg,spectrum);

% this step should combine horizontal and vertical gradients into 
% one single gradient aka reduce the number of channels

However, spectrum does not change.
This can be seen by

isequal(spectrum1.powspctrm,spectrum2.powspctrm) == 1

Also the number of channels (n = 204) is not reduced after
ft_combineplanar when in fact there should only be n = 102 
channels left.

Is this related to the fact that ft_combineplanar is designed to take
only time-frequency maps as input or am I doing something wrong here?

Any advice would be highly appreciated.

Fred