[Mne_analysis] filter type, comparing conditions

[Eliezer Kanal]

Thanks for the quick answer about the filter.

Regarding the comparisons, I had thought that the ">" notation meant  
it was comparing the signal intensity to baseline, with the baseline  
being the average (hence "average") value of the magnetic field (from  
the time interval specified in "scales"). In my case, I have two  
conditions, and I want to compare the intensity of the two signals.  
Instead of loading each and then visually comparing the two, or  
subtracting the timeseries in matlab and making an fif of the  
difference, I want to know if I can load two timeseries and have MNE  
compute the difference between the two, so I can get an idea about the  
differential activity between the conditions. Thanks -

Eliezer


On Sep 24, 2008, at 1:37 PM, Matti Hamalainen wrote:

>
> Hi,
>
> On Sep 24, 2008, at 1:30 PM, Eliezer Kanal wrote:
>
>> Hello folks -
>>
>> Two questions:
>>
>> 1) When I specify a highpass or lowpass filter in mne_process_raw
>> during averaging, what filter is it using (butterworth, chebychev,
>> etc.)?
>
> The filters are realized in the frequency domain by weighting the
> frequency response. The bandpass window transition regions have a
> cos^2 ahspe with a width as specified in the filter dialog.
>
>> 2) When loading a dataset in mne_analyze, under "available data
>> sets" it always lists conditions as " [condition1] > average ",
>> " [condition2] > average ", etc. Is there any way to load a dataset
>> as " [condition1] > [condition2] "?
>
> The notation, inherited from Neuromag software, labels the condition
> and the kind of quantity (average) computed. You can also compute
> standard errors of mean with mne_browse_raw/mne_process_raw, hence
> the ">average" specifier.
>
> What would you like to accomplish with " [condition1] > [condition2]"?
>
> - Matti
>