Theta rhythm in rat hippocampus during running

LFPs collected from 3 64-channel arrays implanted in CA1 of a rat running on a 3-arm maze to collect water rewards. This data is used to evaluate algorithms for decoding the rat's position from the LFP. Note that the values are 1-indexed, as is convention in matlab. For using in python, which is 0-indexed, I recommend subtracting 1 from lapID and trainInds.<b>X</b>: The multi-electrode LFP recorded in CA1, already downsampled to 25 Hz, highpass filtered at 2 Hz, and made complex-valued using the Hilbert transform. Shape = 3808 timesteps x 192 electrodes<b>lapID</b>: An array of shape (3808, 4) where columns = [trial number, cross validation fold, maze arm, time step in original file]. The last column is useful because the data is truncated to only include periods where the rat is running towards the reward port. <b>trainInds</b>: Indices for the training data, containing no overlap with lapID cross validation column indices. Shape = (2265, 5), with each column specifying which indices to use for training in one of 5 folds. This variable was created because we found that it was useful to have the same number of training samples from each maze arm (ie the decoder would assign maze arm without bias).<b>pos</b>: An array of shape (3808,1) indicating the position of the rat along the corresponding track, in mm.