[Eeglablist] Import larger ANT continuous file

[Min Bao]

Hi everyone,

I have a question on importing raw data by pop_loadeep(). Because of the memory limit of the WinXP system, I found I could only read at most 30 minutes (sampling rate: 512 Hz) long data. However, my experiment was longer than 40 minutes. So I tried using the "time1" and "time2" flags of pop_loadeep() to define which segment of the whole dataset would be imported.

Some of my codes are listed here. For simplicity I omited those lines about saving the data:

%===  first, I read the first 30 minutes ===
EEG = pop_loadeep([datapath,filename] , 'time1', 0, 'time2', 1800, 'triggerfile', 'on'); 

%=== then, I begin to read the remaining data by 5 minutes every time
% the loop will break as long as the end of the dataset is reached
ii = 1;
while 1
    ...
    EEG = pop_loadeep([datapath,filename] , 'time1', 1800+1+(ii-1)*300, 'time2', 1800+300+(ii-1)*300, 'triggerfile', 'on');
    ...
    ii = ii+1;
end

All these seemed to work well. And the epoch time displayed on the EEGLAB GUI looks correct for each segment, e.g. 0~1800, 1801~2400, etc. Then I concatenated all these segments. However, I found the number of events for each experimental condition was not correct. So I checked the first 40 events of each segment and took the latencies of all events for the target presentation which should be randomly distributed over time according to our experimental design. But I found they were all exactly the same, which means I think the "time1" and "time2" flags did not work in my code. I guess In the above loop, every time the code may just read the raw dataset from the very beginning rather than from the specified time point. Could anyone tell me if there was something wrong with my method? Thank you very much!

Min Bao



      
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20110324/8370384f/attachment.html>