Hi SPMers,
I was trying to use LCMV beamformer on the Henson's data downloaded from website. After converting(with epoching by trial definition), merging, and averaging, I received data with two conditions. I did analysis in each condition seperately. As I needed the inverse of the sensor covariance matrix, so I computed sensor covariance after input it from D by using spm_eeg_load. The result matrix is a 306 X 306 matrix, but after I check the rank, I found it only has rank around 70. How could the rank drops so much? Does it mean the sensors are dependant on each other?
Many thanks
Chao
|