Explanation of Waveclus Output Files and Variables

Data for one session is stored on DAG-server ..\Data\Sortcodes\

Data for each monkey and session are stored in separate folders: ..\Data\Sortcodes\[MonkeyName]_phys\[SessionDate]\WC\

1. concatenation_info.mat - this file contains info about channel-block-file correspondence
2. dataspikes_ch*_*.mat -
3. settings.mat - settings that waveclus used during clustering
4. [The last but not the least] JPG plots for spike waveforms, features, and timecourse

Our version of waveclus combines blocks recorded on the same electrode depth together and sets the same threshold for them. If electrode depth on a channel changes over the session, the data for different depths are stored in separate files (e.g. dataspikes_ch007_1.mat and dataspikes_ch007_2.mat). One needs additional information to figure out the correspondence between recorded blocks and waveclus files.

1. blocksamplesperchannel - a cell array of size 1 by number of channels. Each sub-cell contains a matrix of size (number of blocks) by 2. The 1st column contains the starting sample number for the corresponding block (0 is the recording starting point), the 2nd - the ending sample number, respectively.
2. channels_to_process - an array with channel numbers
3. sr - sampling rate of the broadband signal
4. whattofindwhere - a cell array 1 by the channel number; each sub-cell has size 1 by the number of depths used for this channel and stores lists of blocks existing for each channel-depth
5. wheretofindwhat - a cell array 1 by the number of blocks; each sub-cell has size 1 by the number of channels and lists the waveclus file number where to find data for the corresponding cell-block (e.g. 1 states for dataspikes_ch*_1.mat, 2 - for dataspikes_ch*_2.mat, etc.)

Use ph_figure_out_waveclus_file_by_channel_and_blocks.m from spike_analysis repo. The usage is explained in the function help.