All analysis routines included draw upon a standard data format per cell.
Only trials in which the monkey correctly identified the tracked targets
are included for analysis in this data format.  Trials are carefully matched
so that across attention conditions the same set of stimulus movies brought
stimuli into and out of the receptive field (matched sensory conditions)

See analysis routines individually for how they extract access data structure.

%********** data - all relevent data fields for a single isolated neuron
%******
%****** data.label(nChannels) = {'SU1' ; 'MU1'; 'LFP1'; 'EYE1'};
%******            SU1 - single unit spike times (1ms precision)
%******            MU1 - multi-unit spike times (1ms precision)
%******            LFP1 - local field potentials (filtered 3-88 Hz 2nd-order Butterworth)
%******                   on the same electrode from which SU1 was recorded
%******            LFP2-N - if present, recordings from other electrodes, sorted
%******                    with LFP2 having highest LFP-LFP coherence with same electrode
%******                    (mean coherence over 3-88Hz) and LFPN having the lowest coherence
%******            EYE1 - the X position of the eyes, in visual degress (1ms sampling)
%******            EYE2 - the Y position of the eyes, in visual degrees (1ms sampling)
%******
%****** data.trials{ntrials} = [[Su1 data from 1:T];[Mu1 data from 1:T];...];
%******                     - each trials cell is a matrix with rows being 1-T ms precision
%******                        and the row corresponding to the label data type (SU1,MU1...)
%******                        from the list of labels above (if 'LFP1' is 3rd label, that is 
%******                        going to have its data in row 3 of each trials{i} cell.
%******
%****** data.spontaneous{ntrials} = [Su1 data 1:250ms]; %spike data only, only SU1 row then
%******                % data from interval when fixation first occurs
%******                % and before the stimuli first onset
%******
%****** data.time{ntrials} = [1:T]  % time precision in ms during trial
%****** data.attend(ntrials) = 1 if attended in RF during 2 of 4 tracking
%******                        2 if distracter in RF during 2 of 4 tracking,
%******                        3 if distracter in RF during 1 of 4 tracking
%****** data.exactspikes{ntrials} = exact single unit spike times per
%******                              per trial to 0.025 ms precision
%******                              each cell is a list of exact spike times
%****** data.fsample = 1000;        %samples per second
%****** data.sustained = XA:XB     % to mark the 800 ms sustained period in our task
%****** data.morepause = -250 to +250 after 1000ms pause, 1500 ms (the pause plus a bit more)
%****** data.pause = exact period of 1000ms pause in the task
%****** data.surround = -950 to -450 ms before pause, stimuli are outside
%******                   the RF (period of non-visual response?)
%****** data.waveform = [1:32];   % the spike waveform sampled at 25 us resolution
%****** data.isolation = 1 - well isolated single unit, 
%******                  2 - clear cluster but some (small) overlap with other cluster
%******                  3 - multi-unit, large overlap between clusters, impossible to select seperate
%******                  4 - well isolated, but lost midway in session (isolation not stable over time)
%******