ShihChengChangeLog.txt

Change Log for matlab functions used to analyze cat data

Shih-Cheng Yen

January 7, 2004
Imported project to CVS.

January 8, 2004
@refractory/plot: Modified help comments to indicate which optional
	input arguments are valid for ISI histogram versus recovery 
	function. Now correctly uses varargin{:} instead of varargin when
	passing varargin to other function.
@refractory/refractory: Added optional input arguments to help comments.
	Now correctly uses varargin{:} instead of varargin when	passing 
	varargin to other function.

January 9, 2004
getFreeFiringRate: New function to compute free firing rate, q(t).
getFunctionValues: New function to compute recovery function w(t).

January 13, 2004
spikeRecovery: Removed function argument in first line of help comments.
	Changed histstep to isibinsize to be more informative.
@refractory/plot: Added option to plot qt, Wt, and rt functions. Changed
	histstep to isibinsize. Now adds trailing 0 before plotting qt, Wt,
	and rt instead of in getFreeFiringRate.
@refractory/refractory: Changed histstep to isibinsize. Added optional
	arguments qtbinsize, duration and binsize. Changed object data 
	structure so it is always accessible by specifying obj.data. Added 
	complete description of object fields. Added code to compute qt,
	Wt, and rt. Now prints progress to screen. Added repetitions and 
	wt fields.
@refractory/subsref: Changed 'rf' to 'data'.
getFreeFiringRate: Removed trailing 0 in rt, qt, and Wt.
getSurrogateSpikes: New function to generate surrogate spikes using
	relative refractory period.

January 14, 2004
getSurrogateSpikes: Updated help comments to include cell number
	argument.
@refractory/plot: Added titles, xlabels and yabels to plots.
@refractory/generateSurrogates: New class method that calls 
	getSurrogateSpikes.

January 15, 2004
getFreeFiringRate: Now converts spike times to column vector using
	new function vecc before continuing.
@refractory/refractory: Added help comments to specify fields needed
	in input data structure. Fixed paragraph wrapping.
mlqsub: New tcsh script to check MATLAB features before submiting 
	job to queue.

January 16, 2004
mlqsub: Rearranged input arguments and made numloops optional.
CatSurrogates: Script to use with qsub to generate surrogates.

January 19, 2004
mlqsub: Modified to work properly with qsub. Now correctly prints number
	of available licenses. Now will use matlab script as is instead of
	using sed to replace occurrences of @ if loops is 1.
@refractory/generateSurrogates: Added code to generate sets of 
	surrogates and store them in a big cell array.
CatSurrogates: Modified to run in 4 jobs. Modified to use new 
	generateSurrogates function.
@refractory/plot: Removed commented code that used to call 
	getFunctionValues. 
@refractory/refractory: Removed curve fit object from data structure
	to allow ogier to load refractory object although it does not have
	a license for the curve fitting toolbox. Now passes curve fit 
	object to getFunctionValues.
getFunctionValues: Modified to take curve fit object as input argument.
getSurrogateSpikes: Added details of input data structure in help
	comments.

January 20, 2004
getSurrogateSpikes: Added cstep optional input argument.

January 21, 2004
getSurrogateSpikes: Fixed bug that was not moving on to the next
	repetition if spike wasn't found at the end of a repetiton.

January 23, 2004
CatSurrogates: Modified to use fewer cells in each group.
gSS.m: New function modified from getSurrogateSpikes.m to work better
	with Matlab compiler. Optional input arguments were removed and
	numeric inputs are now explicitly converted to numbers. Added code
	to loop over sets in order to extend run time of job in queue in 
	order to make use of more processors at a time. Will also reduce
	the number of queue jobs needed to do 1000 sets of surrogates.
gSSwrapper: New tcsh shell script to submit gSS jobs to queue. Now
	passes the number of sets per run to the new gSS.
mlscript.m: Simple matlab script for testing with batch scripts.

January 27, 2004
getRefreshReversals.m: New function to generate list of refreshes 
	between direction reversals in the jittering grating stimulus.

February 2, 2004
getDirectionReversals.m: New function to generate list of refreshes
	between direction reversals in the Buracas-like stimulus.
getGrating1OverF.m: New function to generate 1/f^beta velocity
	distribution for gratings.
phaserand3d.m: New function to scramble FFT phases for movies.

February 3, 2004
getDirectionReversals: Added code to print comments in output file.
getGrating1OverF: Added code to print comments in output file. Now
	also prints out number of phases as first number in file. Modified
	to take framerate as optional input argument instead of refresh 
	rate. Now labels axes of plots and prints out min and max temporal
	frequency as title of first plot. Now takes two optional input
	arguments 'showplot' and 'showgrating' (with the alias 'showall'
	referring to both) to display just the plot or the grating or both.

February 4, 2004
showGratingSeq.m: New function to animate grating using given colormap
	sequence.
getGrating1OverF: Modified to use showGratingSeq. Added help comments.
	Now takes minphase into account when computing temporal frequency.
getDirectionReversals: Modified to use frame rate instead of refresh
	rate. Capitalized optional argument names to facilitate use of 
	abbreviations. Added code to plot sequence as well as to animate
	grating.
getGratingReversals.m: New function to produce Buracas like grating
	reversals using output file like those produced by getGrating1OverF.
	getDirectionReversals is now obsolete.

February 5, 2004
getGrating1OverF: Modified so that sequence starts from 0. Added polar
	plot. Now runs fft on unmodified sequence so that we don't have high
	frequencies due to the wrapping.

February 10, 2004
phaserand.c: New function to perform phase randomization of a movie 
	sequence using the 3D FFT from the FFTW libraries. This version 
	includes threading and was developed for 64-bit SGI IRIX. The 
	inverted FFT is normalized by the total number of pixels in the 
	movie sequence and the value of 128 (assume that is the mean
	grayscale of the movie sequence. Now normalizes output images
	to have similar mean grayscale to input images.
phaserandf.c: Single precision version of phaserand.c to run on
	Mac OS X. Uses FFTW-fma libraries.
Makefile: For phaserand.c and phaserandf.c.
phaserand3d.m: Now leaves phase of DC component intact while scrambling
	the rest.

February 11, 2004
phaserand.c: Moved index arithmetric outside for loop to reduce compute
	time.
showGratingSeq: Now returns the sequence of graylevels for each distinct
	pixel. Removed obsolete code that used palette sequences.
getGrating1OverF: Changed optional input argument MinPhase to MinTF. 
	Now generates a phase vector where the phase of the negative 
	frequencies are the complex conjugate of the positive frequencies.
	Added another variable called lastphase instead of subtracting
	1 from nphases and having to add 1 to it a couple of times. Now 
	plots the phase sequence including negative degrees to avoid 
	discontinuities in the sequence. Added 2nd axis on phase plot to 
	indicate palette indices. Improved PlotFFT code to consolidate 
	multiple operations into fewer lines. Removed phase plot. Plots
	the FFT of the graylevels returned from showGratingSeq.
getGratingReversals: Fixed length of r used during initialization. Added
	lastphase instead of subtracting 1 from nphases. Removed polar plot.
	Added plots of the FFT of the graylevels returned from 
	showGratingSeq.

Februaru 12, 2004
showFFTPhaseDiffs.m: New function to illustrate the differences when
	inverting a frequency domain signal with phase vectors that are 
	totally random and phase vectors where the second half is the 
	complex conjugate of the first half.

February 15, 2004
phaserand.c: Fixed bug that was not reading the input files correctly.
	Added code to compute min and max of output so that we can make 
	sure that after shifting the mean to be the same as the input, the
	range of gray levels is between 0 and 255. Added for-loop to
	recompute phase randomized images if scale is smaller than 0.1
	due to the occasional large negative number in the output (another
	bug?).
	
February 16, 2004
phaserand.c: Added code to print out mag and phase of minimum element.

Februaru 17, 2004
phaserand.c: Fixed bug in code that scaled values so they were not 
	negative and had the same mean gray level as the input. Removed
	code that printed out magnitude and phase of minimum element as
	the minimum value is not related to single magnitude and phase.

April 5, 2004
gSS.m: Replaced code to save output in mat file with code to save data
	using fwrite to reduce disk space by 50%.

April 8, 2004
getSurrogateSpikes: Added code to plot intermediate calculations used
	to create spike trains. Added qscale input argument which changes 
	qt to probability instead of firing rate, the latter which divided
	the probability by the bin size.
gSS: Removed code associated with dbflag and displayflag to optimize 
	performance.
plotRasters: New function used to plot rasters of data as well as 
	surrogates.
readSurrogateBin: New function used to read binary files created by
	gSS function.

April 12, 2004
@FMagnitude: New class that computes the FFT magnitude of spike trains.
@FMagnitude/FMagnitude: Constructor function.
@FMagnitude/plot: Plots FFT magnitude along with standard deviation 
	without DC.

April 13, 2004
@FMagnitude/FMagnitude: Added code to make sure empty matrices are not
	passed to nptSpikeTimesFFT.
gSS: Added rep info in printf statements.

April 16, 2004
getFunctionValues: Simplified to take fewer mandatory arguments so
	this function can be called more easily by getFreeFiringRate2.
@refractory/refractory: Added code to parse new data structure and
	call appropriate new functions (i.e. getFreeFiringRate2).
getFreeFiringRate2: New function that is modified from getFreeFiringRate
	and works with the new data structure.

April 19, 2004
getFreeFiringRate2: Added rec output argument in case when we don't 
	really know the qt bin size. Updated help comments to change REC 
	input argument from mandatory to optional. Added 'frameavg'
	optional input argument to average qt within a frame.
@refractory/refractory: Now stores qtbinsize, rtEdges, and repetitions
	for each cell. Fixed typo in help comments for data.edges. Changed 
	qt, Wt, rt, rtEdges to cell arrays since they might be of different
	length. Added wt field to help comments. Now initializes data 
	structure fields so there is no need to resize memory on the fly.
getSurrogateSpikes: Removed qscale optional input argument. Now uses
	qtbinsize for each cell in input data structure. Modified code to
	check length of qt for number of bins rather dividing duration by
	qtbinsize since that might not be accurate. Updated to access
	rtEdges as cell array.
@refractory/plot: Updated to access rtEdges as cell array.

April 21, 2004
@refractory/refractory: Specified units of qtbinsize in help comments.
	Modified to work with new input data structure that stores 
	original_spiketrain in cell_info field. Modified to work with 
	nptdata	functions. Added cellname field to identify cells.
getFreeFiringRate2: Modified to use new input data structure that 
	stores adjusted_spiketrain in cell_info field. Modified help 
	comments to specify units for optional input arguments. Fixed bug 
	in determining number of reps. Now takes into account that rtbinsize 
	is in units of seconds.
getSurrogateSpikes: Now takes into account that rtbinsize is in units 
	of seconds. Now clears the current figure if 'displayflag' is on.
@refractory/plot: Added description of other three kinds of plots in
	help comments. Uncommented code that set the x-limits for the ISI
	plot. Now labels plots with the name of the cell instead of the 
	number.
@catcell: New class that encapsulates Jonathan's data structure.
@catcell/catcell: Constructor function that also computes mean and 
	standard deviation of spike	counts for each frame.
@catcell/plot: Plots the rasters as well as the variance versus mean
	spike count.

April 22, 2004
catCompute: New function that loads data refractory object and generates
	surrogate spike trains. Also computes and saves spike count mean
	and standard deviation for each frame of the surrogates.
readGratingSequence: New function that reads phase sequence files 
	created by getGrating1OverF.m and getGratingReversals.m.
getFreeFiringRate2: Now returns number of bins in each frame.
@refractory/refractory: Now stores the number of bins in each frame
	in data structure.
gSS: Fixed typo in fprintf call (used $i instead of %i).

April 25, 2004
@fano: New class to display Fano data.
@fano/fano: Constructor function for FANO class.
@catcell/get: New function that overloads get function of nptdata class
	to set appropriate number of events according to varargin.
@catcell/catcell: Fixed typo in calling createEmptyObject when called
	without arguments. Added scmean and scstd fields to data structure
	when creating empty objects.
@catcell/plot: Added code to plot surrogate spike trains. Added code to
	plot spike count mean and variance for surrogate data.
treCompute: New function to compute entropy for TRES/TREZ measurements.

April 26, 2004
treCompute: Fixed bug that was saving output files as filename.mat 
	instead of using the input argument.

April 27, 2004
@trentropy: New class to display spike time reliability data.
@trentropy/trentropy: Constructor function for TREntropy class.
getTREntropy: New function that computes entropy of histc output.
@catcell/subsref: New funtion that facilitates access to catcell 
	objects' data structure.
@fano/fano: Completed constructor function.
@fano/name2index: Copied from @nptdata/name2index with references to
	nptdata changed to fano.
@fano/plot: Plot function for Fano objects.
@fano/plus: Overloaded plus operator for Fano objects.
@fano/ProcessSession: Overloaded ProcessSession function for Fano 
	objects.
@fano/subsref: New funtion that facilitates access to fano 
	objects' data structure.
treCompute: Fixed typo in comment.
@refractory/subsref: Modified to allow easy access to internal data 
	structure.

April 28, 2004
@fano/plot: Changed 'Cells' in title string to 'cells'.
@trentropy/plot: Changed 'Cells' in title string to 'cells'.

May 4, 2004
@catcell/get: Replaced surrogate sets constant with default optional
	input argument 'SurrogateSets'.
@catcell/plot: Added plots to check spike count mean, variance and 
	fano factor for data versus surrogates. Replaced surrogate sets 
	per file constant and surrogate file basename with default optional 
	input arguments	'SurrogateSetsPerFile' and 'SurrogateBasename'. 
	Updated help comments to describe new plots and new optional input
	arguments.
@fano/plot: Plots are now labelled as Fano instead of Entropy.
@refractory/plot: Added ability to plot 'rt', 'qt', and 'Wt' together.
@trentropy/trentropy: Wrapped long lines.

May 5, 2004
@catcell/catcell: Modified to compute spike latency with respect to
	frame onset.
@catcell/plot: Modified to plot spike latency data. Will also compute
	data if the requested bin sizes are different from what was 
	computed.

May 6, 2004
eventCompute: New function that computes event intervals of the
	surrogate spike trains.

May 7, 2004
eventCompute: Added code to compute percentage of reps that have spikes
	in both events making up a sub-stimulus interval. Modified to use
	Jonathan's new code which gets rid of long output arguments.
@FMagnitude/FMagnitude: Added code to generate data automatically, load
	and save objects as needed. Now works with new mat files of the cat
	data.
@FMagnitude/plot: Added title to plots.

May 10, 2004
@FMagnitude/FMagnitude: Added help comments. Fixed bug in processing
	data when data structure is passed as input argument.
@FMagnitude/plot: Added parsing of input arugments so this function can
	be called without passing an extraneous numeric argument. Added
	option to show DC amplitude.
@fano/ProcessCell: New function that replaces ProcessSession that is 
	necessary as a result of changes in the nptdata class.
@trentropy/ProcessCell: New function that replaces ProcessSession that 
	is necessary as a result of changes in the nptdata class.
@catcell/plot: Added some examples of InspectGUI usage in help comments.

May 12, 2004
@FMagnitude/FMagnitude: Removed 'keepNumeric' option from getOptArgs
	since it is now the default. Now correctly saves the varargin 
	returned from getOptArgs.
@FMagnitude/plot: Removed 'keepNumeric' option from getOptArgs since it
	is now the default.
@catcell/catcell: Now correctly saves the varargin returned from 
	getOptArgs.
@catcell/plot: Removed 'keepNumeric' option from getOptArgs since it is 
	now the default.
@fano/fano: Now correctly saves the varargin returned from getOptArgs.
@fano/plot: Removed 'keepNumeric' option from getOptArgs since it is now
	the default.
@refractory/refractory: Now correctly saves the varargin returned from 
	getOptArgs.
@trentropy/trentropy: Now correctly saves the varargin returned from 
	getOptArgs.
@trentropy/plot: Removed 'keepNumeric' option from getOptArgs since it 
	is now the default.

May 13, 2004
eventCompute: Fixed bug that created 1000x1000 cell array instead of
	1000x1 cell array. Added input argument to specify whether event 
	upper thresholds should be taken from the data or computed for each 
	surrogate. Added input argument to specify output filename.
@catcell/plot: Added option to look at distribution of event intervals.

May 14, 2004
@catcell/plot: Added option to plot multi-taper power spectrum density
	of spike trains. Added lines to Latency plot to indicate refresh
	intervals.

May 26, 2004
@catcell/plot: Fixed bug that tried to plot line for 95th percentile
	of event intervals for the surrogates when the surrogate file is not
	present.
@variability: New class that combines the fano and trentropy objects
	so we can plot and compute correlations between fano data and
	trentropy data.
@variability/variability: Constructor for new class. Now keeps the 
	directory path so we can plot surrogate data if necessary.
@variability/name2index: Function that maps names to indices.
@variability/plot: Merged plot functions from fano and trentropy 
	classes. Added plots to look at Fano versus	trentropy z-scores and
	mean versus entropy.
@variability/plus: Function to add objects.
@variability/ProcessCell: Batch processing function.
@variability/subsref: Function to access data.

May 31, 2004
@variability/ProcessCell: Fixed bug in creating empty object - now
	correctly calls variability instead of fano.

June 14, 2004
@variability/variability: Added code to process grating data. Added code
	to search for data.mat files instead of	*g*c.mat. Added code to 
	figure out cell names for cells with data.mat files.
@catcell/catcell: Added code to search for data.mat files instead of
	*g*c.mat. Added code to figure out cell names for cells with
	data.mat files.


August 25, 2004
@sparsity/plot.m: Removed code that changed object data. Fixed histogram 
	edges so the edges used in the calculation is also used in the plot.
	Added 'histc' option to bar plot to align bars to align the edges 
	correctly.
@FiringRate/plot.m: Removed code that changed object data. Fixed histogram
	edges so the edges used in the calculation is also used in the plot.
	Added 'histc' option to bar plot to align bars to align the edges 
	correctly.
@responsiveness/plot.m: Removed code that changed object data. Fixed 
	histogram edges so the edges used in the calculation is also used in
	the plot. Added 'histc' option to bar plot to align bars to align 
	the edges correctly.
@events/plot: Removed code that changed object data. Fixed histogram 
	edges so the edges used in the calculation is also used in the plot.
	Added 'histc' option to bar plot to align bars to align the edges 
	correctly.
@responsiveness/private/CalcResponsiveness: Changed code back to return 
	p_dist instead of the distribution of R values.
@adjSpikes/get.m: Broke long command line into two for better 
	legibility.

August 27, 2004
@FiringRate/plot.m: Changed hist call to histcie to better control bin
	limits used to calculate the histogram and to make sure the
	histogram plot does not have a separate value at the maximum.
	Changed semi-log plots to use line plots instead of bar plots.
@adjSpikes/plot.m: Added FMarkersOff option to turn off frame markers
	when plotting rasters.
@responsiveness/plot.m: Changed hist call to histcie.
@sparsity/plot.m: Changed hist call to histcie.
@sparsity/plot.m: Changed hist call to histcie.
@FMagnitude/FMagnitude.m: Added code to use ispikes object to create
	object. Mean and std in data are now stored in columns to facilitate
	plotting data from multiple cells.
@FMagnitude/plot.m: Removed mandatory second input argument and added
	code to figure out if we are to plot individual data or aggregate
	data. Added code to plot aggregate data.
@FMagnitude/plus.m: New function to create aggregate FMagnitude object.
@FMagnitude/subsref.m: New function to access data in FMagnitude 
	objects.

August 30, 2004
@FMagnitude/plot.m: Added code to normalize all means before plotting so
	as to provide comparisons across data series.

September 8, 2004
@FMagnitude/FMagnitude.m: Now uses adjSpikes instead of ispikes so we
	can compute time-frequency data, which are averaged over 
	repetitions. Added code to store data for time-frequency plots. Now 
	passes varargin to adjSpikes so we can make use of the SaveLevels 
	and RedoLevels functionality. Added remove optional input argument 
	to getOptArgs to remove arguments specific to this function. Now
	passes varargin to nptSpikeTimesFFT so options for that function can
	be specified when the FMagnitude object is being instantiated. Code
	to store time-frequency data for old data structures might need
	revision.
@FMagnitude/plot.m: Added option and code to plot time-frequency data.
@FMagnitude/plus.m: Added code to add time-frequency data.
@adjSpikes/adjSpikes.m: Raster field in data now stores spike times in
	columns rather being a structure containing spike times in rows.
	Storing spike times in columns makes it easier to call
	nptSpikeTimesFFT and should be easier to plot as well. Removing the
	need for the structure simplifies working with the data. Added 
	rasterIndex field to keep track of which rasters go with which data
	set.
@adjSpikes/plus.m: Added code to add newly modified raster field and
	newly added rasterIndex field.
@adjSpikesXCorr: New class to compute cross-correlations along with
	shift-predictor.
@adjSpikesXCorr/adjSpikesXCorr.m: Constructor function for new 
	adjSpikesXCorr class.
@adjSpikesXCorr/get.m: Get function for new	adjSpikesXCorr class.
@adjSpikesXCorr/plot.m: Plot function for new	adjSpikesXCorr class.
@adjSpikesXCorr/plus.m: Plus function for new	adjSpikesXCorr class.
@adjSpikesXCorr/subsref.m: Subsref function for new	adjSpikesXCorr 
	class.

September 9, 2004
@FMagnitude/FMagnitude.m: Now uses different name for varargin returned 
	by getOptArgs since nargin is dependent on varargin so if we remove
	certain arguments from it using getOptArgs, the lengths will
	change. The function nargin is also called just once and the result
	stored instead of calling it multiple times.
@adjSpikes/adjSpikes.m: Now sets the number and holdaxis to 1 and 0
	respectively when creating nptdata object. This allows the adjSpikes
	objects to be added without a warning from nptdata/plus.
@adjSpikes/plus.m: Revised using the template in dirfiles. This was
	necessary to avoid errors caused by the check for empty objects
	being in the wrong place. Also uses new functionality in the
	concatenate function to correctly add the raster field together.
@adjSpikes/plot.m: Uncommented code that changed directory to the proper
	directory in order to plot data. Not sure why it was commented out
	in the last version. Now correctly selects only the relevant data
	in adjFramePoints and adjSpiketrain to pass to CreateRaster.

September 27, 2004
@adjSpikesXCorr/adjSpikesXCorr.m: Modified code to compute cross-
	correlations using the histogram of spike time differences ala
	Nirenberg & Latham (1998), which is very much faster than creating 
	spike trains and calling Matlab's xcorr. Added code to perform 
	Kolmogorov-Smirnov test on xcorr and shift-predictor to see	if the 
	two distributions are significantly different. The significance 
	level can be specified using the KSAlpha optional input	argument. 
	Changed default BinSize to 1. Added code to do Gaussian fit on bins 
	that are significant using binomial statistics. Added option to
	specify bins for computing excess correlated fraction since the new
	way of computing cross-correlations has bins on either side of 0
	instead of having a bin at 0. Now stores string specifying if
	distributions are not different, if center of Gaussian fit is not 
	within fit points (i.e. data inside fit points is a straight line),
	if no fit points were found, or the width and center of the Gaussian
	fit.
@adjSpikesXCorr/plot.m: Added option to plot differences between cross-
	correlation data and shift-predictor, as well as the Gaussian fit.
	Now creates text box with result of Gaussian fit, ECF, and KS stat.
@adjSpikesXCorr/plus.m: Now adds new fields kspvalue, ksstat, legendstr,
	fitgoodness, fresult, xcwidth, pciCounts.
@latency: New class that computes latency of spikes relative to stimulus
	frame onset.
@latency/latency.m: Constructor for new latency class. Code mostly from
	@catcell/catcell.
@latency/plot.m: Plot function for new latency class. Code mostly from
	@catcell/plot.
@latency/plus.m: Plus function for new latency class.
@latency/subsref.m: Subsref function for new latency class.

September 28, 2004
@FiringRate/private/CalcFiringRate.m: Modified to include the spike 
	counts from the last frame. Optimized code by replacing for loops 
	with single	call to histcie, followed by a reshape.

September 30, 2004
@adjSpikes/adjSpikes.m: Modified to pass on optional input arguments to
	objects that are instantiated. Modified call to stiminfo from
	stiminfo('auto',1,'save') to stiminfo('auto',vargin{:}) to remove
	redundant second argument and hardcoded save option. Added comment
	about not passing optional input arguments to ispikes since redo
	does not quite work with ispikes from the cluster directory yet.
@FiringRate/firingRate.m: Modified to pass on optional input arguments
	to object that are instantiated. Modified call to stiminfo from
	stiminfo('auto',1,'save') to stiminfo('auto',vargin{:}) to remove
	redundant second argument and hardcoded save option. Added 
	SessionDir to nptdata before inheriting	from it.
@catcell/plot.m: Modified MTTime optional input argument to MTBinSize
	to better describe what it represents.

October 1, 2004
@adjspikes: Changed adjspikes to lower case so objects saved on Windows 
	will work on non-Windows machines.
@adjspikes/adjspikes.m: 	
@adjspikes/display.m: Changed case of adjspikes.
@adjspikes/get.m: Changed case of adjspikes.
@adjspikes/plot.m: Changed case of adjspikes. Changed catcell references
	to adjspikes.
@adjspikes/plus.m: Changed case of adjspikes.
@adjspikes/subsref.m: Changed catcell references to adjspikes.
@jointevents/jointevents.m: Changed case of adjspikes.
@latency/latency.m: Changed case of adjspikes.
@responsiveness/responsiveness.m: Changed case of adjspikes. 
@sparsity/sparsity.m: Changed case of adjspikes.
createCatAVI.m: Changed case of adjspikes.
@adjspikespsth: Changed adjspikespsth to lower case so objects saved on
	Windows will work on non-Windows machines.
@adjspikespsth/adjspikespsth.m: Changed case of adjspikespsth. Changed
	case of adjspikes.
@adjspikespsth/get.m: Changed case of adjspikespsth.
@adjspikespsth/plot.m: Changed case of adjspikespsth.
@adjspikespsth/private/CreateRaster.m: Changed case of adjspikespsth.
@adjspikespsth/private/adjPSTH.m: Changed case of adjspikespsth.
@adjspikespsth/private/adjPSTHOLD.m: Changed case of adjspikespsth.
@adjspikesxcorr: Changed adjspikesxcorr to lower case so objects saved
	on Windows will work on non-Windows machines.
@adjspikesxcorr/adjspikesxcorr.m: Changed case of adjspikesxcorr. Added 
	check for empty xcdiplus before	trying to use it. If xcdiplus is 
	empty, fitgoodness, fresult and	xcwidth are initialized to empty 
	values. Changed case of adjspikes.
@adjspikesxcorr/get.m: Changed case of adjspikesxcorr.
@adjspikesxcorr/plot.m: Changed case of adjspikesxcorr.
@adjspikesxcorr/plus.m: Changed case of adjspikesxcorr.
@adjspikesxcorr/subsref.m: Changed case of adjspikesxcorr.
@firingrate: Changed firingrate to lower case so objects saved on
	Windows will work on non-Windows machines.
@firingrate/firingrate.m: Changed case of firingrate.
@firingrate/plot.m: Changed case of firingrate.
@firingrate/plus.m: Changed case of firingrate.
@firingrate/subsref.m: Changed case of firingrate.
@firingrate/private/CalcFiringRate.m: Changed case of firingrate.
@fmagnitude: Changed fmagnitude  to lower case so objects saved on 
	Windows will work on non-Windows machines.
@fmagnitude/fmagnitude.m: Changed case of fmagnitude.
@fmagnitude/plot.m: Changed case of fmagnitude.
@fmagnitude/plus.m: Changed case of fmagnitude.
@fmagnitude/subsref.m: Changed case of fmagnitude.

October 5, 2004
@firingrate/firingrate.m: Added 'CaseInsensitive' to nptDir call to 
	check for presence of previously saved object since the case
	of the filename might not change even if we re-save the object.
@adjspikespsth/adjspikespsth.m: Added 'CaseInsensitive' to nptDir call 
	to check for presence of previously saved object since the case
	of the filename might not change even if we re-save the object.
@adjspikes/adjspikes.m: Added 'CaseInsensitive' to nptDir call to check 
	for presence of previously saved object since the case of the 
	filename might not change even if we re-save the object.

October 11, 2004
@refractory/plot.m: Fixed typo in spelling of linear in help comments.
@refractory/refractory.m: getOptArgs now stores the modified varargin
	in varargin2.
getFreeFiringRate2.m: Removed unnecessary keepNumeric flag in call to 
	getOptArgs. Added code to remove spikes that occurred in the blank
	frame of the old data since it was causing problems with the
	insertion of the recovery function wherever there are spikes.
@refractory/plot.m: Removed mandatory second argument so that this
	function will work with nptdata/plot.
@refractory/refractory.m: Added code to compute refractory object using
	stiminfo, adjspikes, and isi objects.
@adjspikesxcorr/plot.m: Added code to plot significant bins in 
	difference between x-corr and shift predictor (not complete).

October 13, 2004
Removed directories @adjSpikes, @adjSpikesPSTH, @adjSpikesXCorr, 
	@FiringRate, @FMagnitude on CVS server by hand since SmartCVS
	on Windows is unable to distinguish between lowercase and 
	upppercase.

October 14, 2004
@firingrate/private/CalcFiringRate.m: Corrected typo (colon at the end
	of the line instead of semi-colon).
@adjspikespsth/adjspikespsth.m: Removed fields sessionname, groupname,
	cellname since they are no longer in adjspikes.

October 15, 2004
@latency/plot.m: Added title to plots.

October 19, 2004
@variability/variability.m: Now saves the modified varargin in 
	varargin2. Added code that use adjspikes and stiminfo objects in
	calculating variability. Now sets flag after checking to see if the
	session we are processing is a grating session. Added code to check
	for the presence of obj_type in the stimulus_info structure since 
	the field is not present in stiminfo objects. Now correctly appends
	'.mat' when checking for the presence of surrogate data files. Added
	code that does not assume name of variable saved in surrogate data
	files.

October 20, 2004
@variability/plot.m: Added mean vs Fano plot, including 2D histogram. 
	Added plot of raw entropy, including option to also plot entropy of 
	surrogate data. Re-organized optional input arguments so we can just 
	specify either 'Fano'/'Entropy' and 'SurrPercent'/'SurrZScores' to 
	reduce number of cases we need to check. Added inputs to specify 
	histogram min, max and step size separately for both x- and y-axes. 
	Added code to alter	plot colors, symbols and size for data and 
	surrogates using optional input arguments.

October 21, 2004
@adjspikes/plot.m: Now uses dirLevel to create shortname for the title.

October 22, 2004
@variability/plot.m: Added option to select z-score or percent threhold
	for all plots. Added different colors for points above and below the
	spike count threshold, and for points which exceed and do not exceed
	the specified threshold. Added flags plotAllData and plotOneData so
	we don't have to keep checking ~isempty(Args.NumericArguments. Moved
	code that checked for points above and below the spike count
	threshold to the top since it is common to all options. Added option
	to change the font size in axes tick lablels, x- and y-labels, and 
	title. Now uses line command instead of errorbar to plot surrogate
	data so the horizontal bars at the ends of errorbars won't overlap 
	with each other. Now uses parameter-value pairs to specify marker
	and color when using the plot command so we can use any color
	instead of being limited to the 7 colors in LineSpec. As a result,
	we now also have a LineStyle option but it is probably never going
	to be used as an optional input argument but rather a way of listing
	the constants used in the function. Added code to draw lines 
	separating points from different cells in population plots. Added
	option to use time of frame points from adjspikes object so we can
	use linkedzoom with the adjspikes object to inspect the rasters.
	Fixed bug in code that checked for points above and below the spike
	count threshold so it uses the threshold in Args instead of 1. Moved
	code that separated indices into signifcant/not-significant, above/
	below spike count threshold to the top so it is not redundantly in
	every option. Created concatenated matrix containing the different 
	sets of indices so we can call histcie just once. Moved code that
	changed directory to the cluster directory to load surrogate data or
	to get frame times to the top so it is not redundantly in several
	options. The variable xvals now initialized properly to either be in
	frame numbers or time for all options. Each option now sets strings
	for title, x- and y-labels and the actual label command is now at 
	the end	so it is easier to modify if we add options like 'FontSize'.
	Moved "axis tight" command to the end so it is not scattered 
	throughout the code.

October 23, 2004
@variability/plot.m: Now checks plotCellSep flag at the end to see if
	lines need to be drawn to indicate where the frames from one cell
	ends and the other begins. Fixed histogram code to properly plot all
	four types of bar plots using the stacked option.

October 26, 2004
plotRasters.m: Modified to use plot instead of line to draw rasters.
@refractory/plot.m: Now does not show plot legends unless the 
	ShowLegend option is specified. Added option to display surrogate 
	spike trains. Now uses dirLevel to shorten name for display in 
	title.
@variability/plot.m: Consolidated code to create histogram and other
	common plot types into section at the end. The plotType argument is
	now used to pick between the different types of plots. Now properly
	loads surrogate data when ShowSurrogateData is used with the default
	option, which is to plot mean vs variance.

January 5, 2005
@firingrate/firingrate.m: Fixed case when calling CalcFiringRate.
@frdiff: New class to compute differences in firing rates among clusters
	within a group.
@frdiff/frdiff.m: Constructor function for new frdiff class.
@frdiff/get.m: Get function for new frdiff class.
@frdiff/plot.m: Plot function for new frdiff class.
@frdiff/plus.m: Plus function for new frdiff class.
@frdiff/subsasgn.m: Subsasgn function for new frdiff class.
@frdiff/subsref.m: Subsref function for new frdiff class.

January 6, 2005
@frdiff/frdiff.m: Now computes and stores the sum of differences and 
	the magnitude of the vector sum. 
@frdiff/plus.m: Updated to add new fields.
@frdiff/plot.m: Added code to plot cummulative data.

January 7, 2005
@frdiff/frdiff.m: Now stores firing rates instead of the objects so we
	can run the vecsimilarity calculation more easily. Now calls
	vecsimilarity to compute vmag field. Added vang field to compute
	similarity index ala Vinje and Gallant, 2000. Added code to
	rearrange differences for 3 vectors so that the order is the same as
	the other cases. Now computes the mean of the sum of differences to
	normalize for differences in number of frames.
@frdiff/plus.m: Updated to add vang field.

January 12, 2005
@adjspikes/get.m: Added optional input argument to modify the default
	group plot property of this object. This allows the adjspikes
	rasters to be plotted in a stacked manner instead of overplotting.
@adjspikes/plot.m: Changed showTitle option to HideTitle since it wasn't
	being used. Removed linkedZoom option since it is not being used.
	Removed irrelevant optional input arguments in call to getOptArgs.
	Added option to not add a title to the plot so that when the rasters
	are plotted in a stacked fashion, the titles won't get in the way.
@frdiff/frdiff.m: Changed frdsum field to frdmean to better reflect 
	the calculations. Added frdmax field to save the maximum difference
	between firing rates. Now uses variable to save current directory 
	instead of calling pwd multiple times.
@frdiff/plot.m: Added optional input arguments to create plots of 
	different data. Added option to plot histograms either by specifying
	the number of bins or the bins to be used. Now appropriately labels
	axes and titles. Added option to create plots with 2 y axes.
@frdiff/plus.m: Updated to use frdmean field instead of frdsum and now
	adds the frdmax fields.
@firingrate/firingrate.m: Remove unused fields dlist and setIndex.
	Now correctly creates a firingRate field in the createEmptyObject 
	function. Now saves output of pwd in a variable so it does not have
	to be called multiple times.
@firingrate/plot.m: Changed showTitle option to HideTitle since it was
	not being used. The HideTitle prevents the title from being created
	which may be useful when creating composite plots with InspectGUI.
	Now plots time course of PSTH by default (although labels in the
	default case are not appropriate).
@firingrate/private/CalcFiringRate.m: Now correctly removes edge effects
	from the beginning and the end of the vector produced by 
	convolution.

January 18, 2005
@frdiff/frdiff.m: Added code to compute mean firing rate from a pair or
	more cells and then to compute the mean and max differences based on
	Poisson differences from the mean, i.e. assuming the standard deviation
	is the square root of the mean.
@frdiff/plot.m: Added options to specify bin size of histograms. Added
	comment to clarify computation of the mean difference between more 
	than 1 pair of cells. Added code to plot control data. Now creates the
	appropriate title for histograms.
@frdiff/plus.m: Added code to add fields for control data.
@variability/plot.m: Added option to remove lines separating data from 
	different cells. This makes the plot a little more legible when the 
	overall	size is small.

February 3, 2005
@adjspikes/plot.m: Now saves current directory instead of calling pwd 
	multiple times. Now returns to original directory so the user doesn't 
	suddenly find themselves in some unexpected directory.
@firingrate/firingrate.m: Removed unused fields dlist and setIndex. Added 
	meancounts, timebins, and binsize fields to save the mean rate, the 
	time bins used to compute the histogram and the binsize used to create 
	the	histogram bins. This makes it easier to create plots of firing 
	rates as a function of time.
@firingrate/plus.m: Now adds meancounts, timebins, binsize fields.
@firingrate/plot.m: Removed unused HideTitle option. Added PSTH option to 
	plot firing rate as a function of time. Removed extraneous characters 
	"fi" at	end of nptSubplot call. 
@firingrate/private/CalcFiringRate.m: Now returns meancounts and timebins 
	so we can easily display the psth as a function of time, which then 
	allows linkedzoom to work with objects like adjspikes. Reverted code 
	used to compute frame level spike counts with overlap to last working 
	version since the new version seems to be incorrect. Fixed bug in 
	Repetitions code that would generate bins that were not monotonically 
	increasing.
@latency/plot.m: Removed SubPlot option since that option is now implied 
	when the function is called without numeric arguments.
@refractory/plot.m: Added LabelsOff option to work with nptdata/plot. 
	Removed description of plot in title to make it better when used with 
	nptdata/plot to create subplots.

February 7, 2005
spikeRecovery.m: Now returns pfpts which may be changed in the process of 
	fitting using the PeakRatio and FitWindow criteria. Added option to do
	running fits to find best fit and to display and pause the fits. Added
	option to specify window to do fit (ala Berry & Meister, 98). Added 
	option to check the ratio between the value at the peak and the value 
	at expfitmax. Added code to check for invalid fits to prevent problems 
	with the plot function later.
@refractory/refractory.m: Now uses pfpts returned by the spikeRecovery 
	function.

February 14, 2005
@frdiff/plot.m: Added option to do plotyy plots with mean and max, vang and 
	vmag. Added code to alter the plot symbols when using plotyy so the 
	data points are apparent. Added the mean value in the title for 
	histograms. Now computes the histogram on the ctrldata only if it is 
	not empty.

February 16, 2005
spikeRecovery.m: Removed curve fit output argument since it was smoothing 
	out the absolute refractory period and thus possibily leading to 
	artificially short ISI's. Removed the PeakRatio option which was 
	introduced to differentiate it from the FitWindow option since the 
	PeakRatio also uses the FitWindow argument. Now computes the 
	Peak-to-End ratio in log coordinates to make it more consistent with 
	the rest of the calculations. Added code to locate the maximum point by 
	finding the first point that crosses a specified percentile of the 
	maximum. This helps to prevent cases of spurious maximums at large 
	latencies. Added code to do successive fits until the slope of the 
	exponential is negative if the Peak-to-End ratio is below threshold. 
	Now uses new npt function zerolog to avoid log of zero errors. Added 
	code to check for NaN's when grabbing the value at expfitmaxpts and 
	iteratively decreasing the data points until a valid value is found. 
	Moved code in fitWindow subfunction back to the FitWindow option since
	it is now only used in that case.
refractory.m: Now no longer uses the curve fit object that used to be 
	returned by the spikeRecovery function. 

February 17, 2005
refractory.m: Added the curve fit object back since the recovery function is 
	simply too noisy if derived from the raw histogram. Now passes the raw
	recovery points to getFreeFiringRate2 as well so that can be used to ensure
	the absolute refractory period is observed. Now also passes the isibinsize
	to getFreeFiringRate2 to make sure that the function values generated from 
	the curve fit object use the same bin sizes that were used to compute the 
	curve fit object.
spikeRecovery.m: Added the code to compute the curve fit object back since the
	recovery function is simply too noisy if derived from the raw histogram. 
	Fixed bug that was adding 1 to pfpts, which caused the refractory/plot 
	function to overshoot the end limit since it was using pfstart + pfpts to
	obtain pfend.
getFreeFiringRate2.m: Renamed qtbinsize to ISIBinSize, which is now passed to 
	this function by refractory.m to make sure that the bins size used in the 
	isi histogram computed in deriving the recovery function is similar to the
	one used to compute the free-firing rate. ISIBinSize is also now used to
	generate the function values from the curve fit object. The function values
	obtained from the curve fit object are combined with the data points of the
	recovery function to ensure that the absolute refractory period is observed.

February 24, 2005
@firingrate/plot.m: Added option to create heterogeneity plots of the peak
	and mean firing rates.
@firingrate/private/CalcFiringRate.m: Fixed bug that was not converting spike
	counts to rates for frame level calculations with no overlaps.
@latency/plot.m: Replaced the dummy optional argument with real optional input
	arguments to look at the auto-correlation of the latency histogram, the
	cross-correlation of the latency histogram with the refresh, and the FFT
	of the latency histogram.
@firingrate/heterogeneity.m: New function to return the organize data from the
	object into groups using nptdata/groupDirs.
@refractory/subsasgn.m: New function to modify object data.
getGrating1OverF.m: Function moved to npt/scripts/miscellaneous.
getGratingReversals.m: Function moved to npt/scripts/miscellaneous.
showGratingSeq.m: Function moved to npt/scripts/miscellaneous.

March 9, 2005
@frdiff/frdiff.m: Added code to compute the correlation coefficient between
	firing rates.
@frdiff/plus.m: Added code to add the fields used to store the result of the
	correlation coefficient calculation.
@adjspikesxcorr/adjspikesxcorr.m: Added code to compute sliding cross 
	correlation. Added code to generate surrogates to compute significance of
	sliding cross correlation.
@adjspikesxcorr/plus.m: Modified xc and sxc fields to be added together using
	the concatenate function instead of [] since the length of xc is variable
	if the sliding cross correlation is calculated.
@adjspikesxcorr/plus.m: Added code to plot data of sliding cross correlation.
	This function used to use the lags field to identify xcorr data but this is
	will no longer work so need to use some other field.
@adjspikesxcorr/private/asxcSurrSynchrony.m: New function to generate 
	surrogates and compute cross correlations for a frame.
@adjspikesxcorr/private/combineSurrSynchrony.m: New function to combine the
	data from all the surrogates generated to compute significance of cross
	correlations.
@adjspikesxcorr/private/computeSynchrony.m: New function to compute the cross
	correlations of two histogrammed spike train.

March 10, 2005
ccwrapper: Added shell script to create surrogates with relative refractory
	period and compute their spike count variability to CVS. Fixed code
	so it will correctly take a list of directories as input arguments.
trewrapper: Added shell script to compute spike time variability of surrogate
	spike trains with a relative refractory period to CVS. Fixed code
	so it will correctly take a list of directories as input arguments.
eventwrapper: Added shell script to compute events of surrogate spike trains
	with a relative refractory period to CVS.
gSSwrapper: Now creates symbolic link to rfd data.
@adjspikesxcorr/private/asxcSurrSynchrony.m: Added code to explicitly convert
	arguments to numbers so that we can compile this function into a standalone
	executable. Added code to check for frames with no spikes. Added code to
	compute for more than 1 frame since the jobs on ogier were taking too little
	time so not all the nodes were being used.
asxcwrapper: New shell script to submit queue jobs to compute surrogates for
	each frame. Now properly encapsulates the command that produces the 
	jobid with `. Fixed bug that was not incrementing while loop variable.
	Fixed missing quotes from the command to combine surrogate data. Modified to
	pass starting and ending frame numbers to asxcSurrSynchrony. Modified to use
	the beforeok depend syntax in qsub instead of the afterok depend syntax 
	since that required a string that was too long. Fixed bug that was using
	framen variable which has been replaced by startframe and endframe. Fixed
	bug that was missing a space between @ and endframe.
@adjspikesxcorr/private/combineSurrSynchrony.m: Modified to convert arguments
	explicitly to numbers.
@adjspikesxcorr/adjspikesxcorr.m: Added input argument for specifying number of
	frames per job when calling the shell script to generate the surrogates.
	Now stores the name of the shell script in Args. Added missing argument for
	name of final surrogate file. Fixed typo in spelling of WaitForSurrogates.
	Added code to run shell script in the background so we don't have to wait
	for the shell script to finish submitting all the jobs.
@adjspikesxcorr/plus.m: Added \n character in warning.

March 11, 2005
@adjspikesxcorr/adjspikesxcorr.m: Added help comments about how data fields are
	used when the Sliding option is used. Added code to compute surrogate data.
	Now correctly specifies full name of surrogate file so that ispresent is 
	able to find it. Added code to re-save object if surrogate data was added to 
	the	object. Fixed bug that was adding a field named ActualBinSize to the 
	data structure. 
@adjspikesxcorr/plot.m: Added Sliding flag to temporarily identify objects
	computed with the Sliding option. The object should be modified to retain
	the Args structure used to generate the object. I am holding off on this 
	since I still have some old adjspikesxcorr objects that I would like to 
	take a look at. Added code to create plot of surrogate histogram using 
	imagesc, with data values plotted on top. Added code to create plot of 
	z-score distributions using HistMin and HistStep arguments.

March-16-2005
@latency/plot.m: Added the Color input argument so that this function can be 
	used with the nptgroup/plot function to create group plots.Now uses 
	'ShortName' in the title in every case.
@adjspikesxcorr/adjspikesxcorr.m: Added code to use dirLevel(0,'GetClusterDirs') 
	to generate ClusterDirs from the current directory if none were passed in.
	Surrogate histogram bins now go all the way to the maximum of the data so 
	we can see the data along with the surrogate histograms when we plot them 
	on top of each other.
@adjspikes/plot.m: Added Interleaved, GroupPlots and GroupPlotIndex input 
	arguments to create plots where the rasters of a group of cells are 
	interleaved to increase visibility.
@firingrate/firingrate.m: Now uses the binsize returned from CalcFiringRate to 
	initialize data.binsize since there was a problem of data.binsize being 
	'frame' when the default arguments are used. This caused an error in the 
	plot function which tries to convert spike count values into rates using 
	the data.binsize value.
@adjspikesxcorr/plus.m: Commented out code that checked if lags vectors were 
	the same since they will not be the same when the Sliding option is used.
@adjspikesxcorr/plot.m: Fixed bug that failed to initialize the lagslength 
	variable.
@firingrate/private/CalcFiringRate.m: Now returns the binsize used in the 
	calculation so if the default window size, which is 'frame', is used, we 
	can store the actual window size in the constructor. Shifted the meancounts 
	= counts line in the frane level calculation with no overlap so that if the 
	variable counts is converted to rate, meancounts will also be converted to 
	rate.
@firingrate/plot.m: Fixed bug in the way the Color argument was used with the 
	stairs command.

March-31-2005
@adjspikesxcorr/plot.m: Added coded to plot significant excess synchronous 
	spikes in red for the sliding xcorr data. 
@adjspikesxcorr/adjspikesxcorr.m: Changed optional input arguments so it now 
	takes number of subwindows, number of bins in a subwindow, and the number 
	of bins to increment when evaluating overlapping windows. Added code so 
	that these values can be specified in terms of the frame duration by having 
	the arguments evaluated programmatically if the arguments are strings 
	containing the word "frame". Changed WaitForSurrogates to WaitForQsub since 
	you can now use qsub to compute data without surrogates. Also changed 
	SurrWaitTime and SurrProgram to QsubWaitTime and QsubProgram respectively. 
	Now uses the raster field in adjspikes instead of the continuous data to 
	compute synchrony. This way the data can be evaluated in the same way that 
	the surrogates are evaluated using the new xcorr algorithm that does not 
	require checking for refractory period violations in the surrogates. Now 
	calculates and preallocates many variables needed in the surrogate 
	generation program so they won't have to be regenerated over and over by 
	each run. 
CatCmds.m: Added commands to plot sliding xcorr data. 
@adjspikesxcorr/private/asxcSurrSynchrony.m: Revamped to compute xcorr for 
	surrogates without having to check for refractory period violations. 
	Renamed input argument from frame to window since it is now possible to do 
	the analysis on any sized window. Now computes complete xcorr on both the 
	data and the surrogates instead of just a few lags. Added optimization so 
	that the xcorr is computed only on relevant trials. Now loads most 
	variables from the mat file created by adjspikesxcorr so the variables 
	don't have to be recomputed and allocated again and again. 
asxcwrapper: Commented out code that invoked final job to combine the 
	surrogates since the matrices will be too big. 
@adjspikesxcorr/private/asxcSurrSynchrony.m: Now converts xc matrix to sparse 
	matrix before saving to conserve disk space. 
@adjspikesxcorr/private/asxcSurrSynchrony.m: Removed surrSets argument since 
	that is now loaded from the mat file created by adjspikesxcorr. 
asxcwrapper: Removed sets argument since asxcSurrSynchrony no longer needs that 
	as an argument. 
@adjspikesxcorr/adjspikesxcorr.m: Now no longer passes the number of surrogates 
	to asxcwrapper since that is now passed through the saved mat file. 
@adjspikesxcorr/adjspikesxcorr.m: Fixed bug in code that failed to change 
	JobFrames to JobWindows when the name of the input argument was changed. 

April-1-2005
@adjspikesxcorr/adjspikesxcorr.m: Added DataFile field to Args so we don't end 
	up using a constant string in the middle of the code. Added SurrFilePrefix 
	to Args so we can check for presence of surrogate files. Added code to 
	compute synchronous spikes from data file generated by the surrogate 
	generation program. Added code to check for presence of surrogates before 
	trying to generate surrogates. 
@adjspikesxcorr/adjspikesxcorr.m: Removed code to load surrogate data when 
	loading saved objects since that is now done by CreateObject. 
	
April-2-2005
@adjspikesxcorr/adjspikesxcorr.m: Added code to save the appropriate lags 
	vector. Fixed bug that was not saving the asxcdata file with the right 
	name. 

April-05-2005
CatCmds.m: Modified to use pre-generated ndinter nptdata object.
@adjspikesxcorr/adjspikesxcorr.m: Now passes datafile and surrprefix arguments 
	to asxcwrapper so we can generate surrogates with different parameters. 
asxcwrapper: Added datafile and surrprefix arguments that are passed to the 
	asxcSurrSynchrony program so we can generate surrogates with different 
	parameters. 
@adjspikesxcorr/private/asxcSurrSynchrony.m: Added datafile and surrprefix 
	arguments instead of hard-coding in the program so that we can generate 
	multiple surrogates with different parameters. 
@adjspikesxcorr/plot.m: Removed HistMin argument which defaulted to 0 so we can 
	see negative z-scores. Added LabelsOff argument so creating subplots using 
	nptdata/plot will work properly. Added code to plot running z-score. 
	Modified code to create time vector on x-axis to work with new lags field. 
	Added check for data with no significant z-scores so we don't get an error 
	when trying to plot the significant bins in red. 
@adjspikesxcorr/adjspikesxcorr.m: Now rounds SubWindowBins and OverlapIncrBins 
	to integers to make sure that they are indices into binidx. 
CatCmds.m: Added code to create ndintra and ndinter nptdata objects for 
	Charlie's machine. Added code to generate adjspikesxcorr objects. 
@latency/get.m: Get function for object to specify default GroupPlotProperties.  
@adjspikesxcorr/subsasgn.m: Function to assign values to object's structure.

April-06-2005
@adjspikesxcorr/adjspikesxcorr.m: Now saves nsbins as string instead of 
	converting it twice. Now rounds subwindow bins up so we make sure we don't 
	miss part of the frame. Fixed bug that was not sending the correct number 
	of jobs to the asxcwrapper program.

April-11-2005
@adjspikesxcorr/plot.m: Added option to plot mean and std of the synchronous 
	spikes of the surrogates. Now works with cummulative objects. Added hack to 
	work around old objects that don't have the pciCounts field set for 
	cummulative objects. 
@adjspikesxcorr/plus.m: Modified to correctly add objects created by Sliding 
	option. Added hack to work around older objects created without the 
	pciCounts field set correctly. 
@adjspikesxcorr/adjspikesxcorr.m: Added remark in help comments to indicate 
	that the field pciCounts is now used to keep track of the frames in sxc 
	that belong to each pair. 
@adjspikes/plot.m: Added Fast option to speed up plotting of rasters by 3 to 9 times. 
CatCmds.m: Now uses Fast option in adjspikes to plot rasters.

April-18-2005
@adjspikessync/private/asxcSurrSynchrony.m: New class that implements the 
	sliding xcorr calculation that used to be in the adjspikesxcorr class. 
@adjspikessync/private/computeSynchrony.m: New class that implements the 
	sliding xcorr calculation that used to be in the adjspikesxcorr class. 
@adjspikessync/subsref.m: New class that implements the sliding xcorr 
	calculation that used to be in the adjspikesxcorr class. 
@jointevents/jointevents.m: Replaced calls to dirLevel with getDataDirs. 
@adjspikesxcorr/plus.m: Replaced with code from version 1.3 since the 
	modifications to accomodate the sliding xcorr calculations are no longer 
	necessary now that there is a separate class to perform that calculation. 
@adjspikesxcorr/plot.m: Added option to turn off the significance plot if 
	desired. Now uses getDataDirs to shorten the pair name for the plot title. 
@variability/plot.m: Replaced calls to dirLevel with getDataDirs. 
@gsparsity/gsparsity.m: Replaced calls to dirLevel with getDataDirs. 
CatCmds.m: Added commands to plot several adjspikessync objects. Added commands 
	to plot several adjspikesxcorr objects. 
@adjspikesxcorr/adjspikesxcorr.m: Added AutoCorr option so that the 
	autocorrelation count is subtracted from the xcorr count so the plots are 
	clearer. The AutoCorr option also now makes it easier to create the 
	adjspikesxcorr object for single clusters using ProcessDirs. 
@adjspikes/adjspikes.m: Removed code to change directory before instantiating 
	the stiminfo object since the stiminfo object now correctly uses 
	getDataDirs to change to the appropriate directory. Removed unused dlist 
	and setIndex fields when creating the empty data structure. Added 
	adjSpiketrain, adjFramePoints, raster and rasterIndex fields to the empty 
	data structure. 
@adjspikessync/private/combineSurrSynchrony.m: New class that implements the 
	sliding xcorr calculation that used to be in the adjspikesxcorr class. 
@refractory/refractory.m: Removed code to change directories before 
	instantiating the stiminfo object since the stiminfo object now correctly 
	changes to the appropriate directory. 
@adjspikes/plot.m: Added slighty more separation between repetitions when using 
	the interleaved option. 
@responsiveness/heterogeneity.m: Replaced calls to dirLevel with getDataDirs. 
@frdiff/frdiff.m: Replaced calls to dirLevel with getDataDirs. 
@stcov/stcov.m: Replaced calls to dirLevel with getDataDirs. 
@latency/latency.m: Replaced calls to dirLevel with getDataDirs. 
@frdiff/plot.m: Replaced calls to dirLevel with getDataDirs. 
@adjspikessync/plot.m: New class that implements the sliding xcorr calculation 
	that used to be in the adjspikesxcorr class. 
@adjspikessync/plus.m: New class that implements the sliding xcorr calculation 
	that used to be in the adjspikesxcorr class. 
@latency/plot.m: Replaced calls to dirLevel with getDataDirs. 
@refractory/plot.m: Replaced calls to dirLevel with getDataDirs. 
@adjspikessync/get.m: New class that implements the sliding xcorr calculation 
	that used to be in the adjspikesxcorr class. 
@adjspikesxcorr/get.m: Added default group plot separation setting. 
@sparsity/heterogeneity.m: Replaced calls to dirLevel with getDataDirs. 
@adjspikessync/adjspikessync.m: New class that implements the sliding xcorr 
	calculation that used to be in the adjspikesxcorr class. 
@responsiveness/responsiveness.m: Removed code to change directories before 
	instantiating the stiminfo object since the stiminfo object now correctly 
	changes to the appropriate directory. 
@variability/variability.m: Replaced calls to dirLevel with getDataDirs. 
@adjspikessync/subsasgn.m: New class that implements the sliding xcorr 
	calculation that used to be in the adjspikesxcorr class. 

April-25-2005
@firingrate/firingrate.m: Removed code that changed directory to the session 
	directory before instantiating the stiminfo object since the stiminfo 
	object can do that on its own now. 
@frdiff/frdiff.m: Fixed bug that was not converting the p values into a vector 
	before taking the maximum which resulted in p vectors that were different 
	in size which caused problems with the plus function. 

April-27-2005
@adjspikessync/adjspikessync.m: Added option to generate asxcdata file without 
	calling the shell script to submit jobs to the queue so we can compute the 
	asxcdata file in parallel across multiple nodes. 
@adjspikessync/adjspikessync.m: Now checks if datafile is present so we can 
	just launch the shell script to submit jobs to the queue once the datafiles 
	are created using multiple nodes. 

May-03-2005
@shufflesync/private/shufflesyncsurr.m: Function to compute xcorr of 
	shuffled surrogates.
@shufflesync/subsasgn.m: Function for class to compute shuffled surrogate 
	synchrony.
@shufflesync/private/computeSurrData.m: Function to compute synchronous 
	spikes from all surrogates.
@shufflesync/shufflesync.m: Function for class to compute shuffled 
	surrogate synchrony.
@shufflesync/plot.m: Function for class to compute shuffled surrogate 
	synchrony.
@shufflesync/plus.m: Function for class to compute shuffled surrogate 
	synchrony.
@shufflesync/get.m: Function for class to compute shuffled surrogate 
	synchrony.
@shufflesync/private/shsywrapper: Shell script to submit pbs jobs to 
	compute shuffled surrogate synchrony.
@shufflesync/subsref.m: Function for class to compute shuffled surrogate 
	synchrony.
@shufflesync/private/computeSurrData.m: Now constructs the filename for 
	each surrogate instead of using the structure returned by nptDir since 
	that seems to be causing error when compiling with mcc.

May-04-2005
@shufflesync/private/shsywrapper: Modified to shorten the names of the 
	processes passed to qsub since there is a limit of 15 characters.
@shufflesync/shufflesync.m: Fixed bug that was not saving the xctemp 
	variable.
@shufflesync/shufflesync.m: Added missing space separating arguments in 
	syscmd.
@shufflesync/private/shufflesyncsurr.m: Added code to make sure the 
	histogram is taken for the data and for each of the surrogates 
	separately, especially when xcmat is a row vector.
@shufflesync/shufflesync.m: Fixed bug that did not declare NoQsub as an 
	input argument flag.
@shufflesync/shufflesync.m: Replaced call to syncshufflesurr with 
	shufflesyncsurr.
@shufflesync/shufflesync.m: Modified code in NoQsub option to correctly 
	call computeSurrData to produce the result file and then to create the 
	object using the result file.
@shufflesync/private/shufflesyncsurr.m: Fixed bug that was replacing xctemp 
	from the data file with a temporary variable which produced surrogate 
	files with incorrect sizes which was causing computeSurrData to fail.
@shufflesync/shufflesync.m: Modified to use nptDir instead of ispresent to 
	look for saved object file since ispresent is much slower when there 
	are a lot of files in the current directory.

May-06-2005
@shufflesync/private/shufflesyncsurr.m: Modified to shorten long lines.
@shufflesync/shufflesync.m: Added option to return empty object if no 
	result file is present instead of generating the surrogates.
@shufflesync/private/shufflesyncsurr.m: Modified to check for refractory 
	period violations for each surrogate to reduce memory usage.
@shufflesync/private/shufflesyncsurr.m: Now skips overlap checking for 
	repetitions with only 1 spike and repetitions with spikes that are 
	separated by more than the minimum required for overlap.
@shufflesync/shufflesync.m: Renamed numCombinations (which is no longer 
	used by shufflesyncsurr) to needrefcheck. 

May-07-2005
@shufflesync/shufflesync.m: Fixed the fields in the createEmptyObject 
	function to remove setNames and add Args. 
@shufflesync/plot.m: Added option to highlight synchronous spikes in the 
	rasters created by adjspikes.
@adjspikes/plot.m: Added option to highlight selected spikes in the 
	spiketrains. Removed references to stimInfoDir since the stiminfo 
	object knows how to change directories itself. 

May-10-2005
@shufflesync/plot.m: Moved code to find synchronous spikes to the 
	constructor. The Rasters option now just uses the MarkMatrix field in 
	the object. 
@shufflesync/plus.m: Added MarkMatrix field to store synchronous spikes 
	for the raster plot. 
@shufflesync/private/computeSurrData.m: Added calculation of the shift 
	predictor data. 
@shufflesync/shufflesync.m: Modified to save all information necessary to 
	create object in the data file so that we can first check for the 
	presence of the result file and create the object if the result file 
	is present using the result file and the data file. Modified to 
	include space for the shift predictor data in the variables created 
	for the surrogate generation. Addded variables used in the surrogate 
	generation for the psth calculation. Added code that used to be in the 
	plot function that saved synchronous spikes for highlighting in the 
	raster plot. Now calls a new makeObject subfunction to create an 
	object instead of repeating code to create object in three different 
	places. Added fields in the data structure of the object for the shift 
	predictor data and for marking synchronous spikes. 
@shufflesync/private/shufflesyncsurr.m: Added code to compute the shift 
	predictor for the data. Added code to compute the psth for the data 
	and the surrogates including repetitions that only have spikes in one 
	cell. 
@adjspikes/plot.m: Added option to highlight selected spikes in the raster 
	plot. 
@shufflesync/private/computeSurrData.m: Modified to return error only if 
	there are few surrogate files than expected since it is possible for 
	the surrogate file prefix to match more than the expected files (e.g. 
	a prefix of shsysurr will match shsysurr0001 and shsysurrA0001). 
@shufflesync/private/computeSurrData.m: Now correctly uses numwindows 
	instead of the number of files that match the surrogate file prefix 
	for all calculations. The old method was causing problems when there 
	were files in the directory like shsysurrA0001.mat when the prefix we 
	are looking for is shsysurr. 
@shufflesync/shufflesync.m: Added printf statements when waiting for qsub 
	to provide some visual feedback. 

May-11-2005
@shufflesync/shufflesync.m: Moved code that marked the synchronous spikes 
	in the raster to the makeObject function so that we don't have to find 
	ways to regenerate the data file without generating the surrogates in 
	order to mark the spikes. As a result, we no longer save markmat1 and 
	markmat2 in the data file. Renamed MarkMatrix field to 
	RasterMarkMatrices to reflect the use of this field to store both the 
	synchronous spikes in the data as well as the shift-predictor. 
@shufflesync/plot.m: Now uses logical indices to plot the significant and 
	non-significant z-scores without any overlap instead of plotting all 
	points first and then plotting the significant points in a different 
	color. Added plot of the shift-predictor values in the SurrHist plot 
	option. Modified to make use of the new rasterMarkMatrices field in 
	the object to plot synchronous spikes in the data and in the 
	shift-predictor. Now also passes input arguments to the nptgroup/plot 
	function. 
@shufflesync/plus.m: Modified to add the new RasterMarkMatrices field. 
@shufflesync/private/shufflesyncsurr.m: Now correctly initializes the xc1 
	to nan's instead of 0's which was causing spurious synchronous spikes 
	in the shift-predictor. 
@adjspikes/plot.m: Added option to highlight any number of sets of spikes. 
	Now uses the new function nptDefaultMarkers to get the markers for the 
	different sets of spikes. The nptDefaultMarkers function includes the 
	ability to specify the markers using an input argument. 
@shufflesync/plot.m: Now correctly changes directory to the combination 
	directory before trying to instantiate nptgroup object when creating 
	the raster plots. 
@shufflesync/plus.m: Now adds the ShiftPSyncSpikes field of the object. 
@shufflesync/shufflesync.m: Changed SkipNoresult option to SkipNoResult so 
	we can use aliases more easily. 

May-13-2005
@shufflesync/plot.m: The default stairs plot of the number of synchronous 
	spikes in each window now works properly on new objects. Added labels 
	to the default plot. 

May-24-2005
@shufflesync/private/shufflesyncsurr.m: Added use of constants to index 
	into the data or surrogate columns to make it easier to modify in the 
	future. Modified to save the data and surrogate spike times (but not 
	the shift-predictor since it will be identical to the data) instead of 
	the psth so we can do kstest2. Now also saves the repetition numbers 
	so we can reconstruct the surrogate spike trains if we wanted to. Now 
	uses a for-loop to generate surrogate spike trains for repetitions 
	without spikes in both cells instead of replicating the code. 
@shufflesync/private/computeSurrData.m: Added use of constants for indexing
	into the data or surrogate columns to make it easier to modify in the 
	future. Now computes the significant xcorr bins as well as the kstest2 
	between the psth distributions of the data versus the surrogates. 
@shufflesync/shufflesync.m: Added vectors of surrvec + 1 and surrvec + 2 to
	the data file since we no longer save the spike times of the 
	shift-predictor, which should be identical to the data. As a result, 
	when we save the spike times we are saving numSurrogates+1 columns of 
	data. On the other hand, when we save the xcorr, we have to add to add 
	the shift predictor so we have numSurrogates+2 columns of data. 
@shufflesync/plot.m: Made SurrHist the default plot type since it is 
	currently the most used option. Renamed previous default plot type to 
	the input argument SyncSpikes. Added PSTH option to display psth 
	distribution between the data and the surrogates (still in 
	development). Added XCorr option to display full xcorr instead of just 
	the synchronous spikes (still in development). 
@shufflesync/private/shufflesyncsurr.m: Fixed bug in placement of code 
	that saves the repetition numbers with spikes present. 
@shufflesync/private/shufflesyncsurr.m: Now correctly transposes repsnopair 
	before saving into repswspike2. 
@shufflesync/private/shufflesyncsurr.m: Now converts repsnopair to column 
	vector by using (:) since simply transposing did not work. 
@shufflesync/plot.m: Added option to specify threshold for p values 
	returned by kstest2 to test for differences in the PSTH. Moved code 
	used to compare xcorr plots for different shuffle values to PSTH 
	temporarily in order to finalize the code for the XCorr option. The 
	code in the XCorr option now checks the zscores to see if it exceeds 
	the z-score threshold, and checks the p-values returned by kstest2 to 
	make sure the PSTHs of both cells are not significantly different from 
	the surrogates. 
@shufflesync/private/computeSurrData.m: Now stores the mean, std, and 
	zscores of the xcorr values in the result file so we can use different 
	thresholds in the plot function to look at the results. Now correctly 
	renames the raw xcorr values of the data and the surrogates so it does 
	not conflict with existing variables. Added check to make sure there 
	are at least 4 spikes in each cell before attempting to use kstest2 
	since it is not accurate below 4. Now converts matrices with a large 
	number of zeros to sparse matrices to save disk space. 
@shufflesync/private/computeSurrData.m: Added more detailed comments on why 
	each cell has to have more than 4 spikes per cell in order for kstest2 
	to be accurate. 
@shufflesync/private/shufflesyncsurr.m: Removed output arguments since it 
	seems to be causing the computeSurrData job to be executed prematurely 
	in pbs. 

May-26-2005
@shufflesync/plot.m: Added NoXCorrData flag so that we can merge mutiple 
	XCorr plots into one. Added GroupPlots and GroupPlotIndex arguments so 
	we can differentiate multiple plots from each other. Updated 
	development code in the PSTH option to work with new data structures. 
	Now uses pcolor instead of imagesc so we can have unequal bin sizes 
	which allows the pcolor plot to correctly line up with respect to time 
	of the stimulus. Now marks the bins which the z-score threshold is 
	exceeded but the psth's are considered different in gray. 

June-01-2005
@shufflesync/plot.m: Added option named Try and moved development code into 
	that section so that we can test the PSTH code. Added NoPsthData option 
	similar to NoXCorrData so we can overplot multiple PSTH. Added YLim 
	option to limit the y-axis of the xcorr plot. Added code to plot PSTH 
	data for alternating windows with different symbols. Now uses start 
	time of bins to properly align with the rasters. 
@shufflesync/shufflesync.m: Changed default WindowBins to 'frame' since 
	that seems to be the most commonly used setting. Added RecalcData 
	option to just save the data file so we can add some fields to the data 
	file that can be used by the plot function or the computeSurrData 
	function without having to regenerate the surrogates. Now saves binsc 
	to the data file so we can plot the PSTH of the data easily. Now saves 
	subbins so that the PSTH can be plotted with the actual start time of 
	each bin. 

June-02-2005
@shufflesync/plot.m: Removed commented code. 
@shufflesync/private/shsywrapper: Now takes only the window numbers and the 
	name of the surrogate data file as input since shufflesyncsurr and 
	computeSurrData have been modified to take only the surrogate data file 
	as input. 
@shufflesync/private/shufflesyncsurr.m: Now takes the start and end window 
	numbers and the name of the surrogate data file as the input arguments. 
	The name of the data file is now available in the surrogate data file. 
@shufflesync/private/computeSurrData.m: Now takes only the surrogate data 
	file as input. The data file name stored in the surrogate data file is 
	used to load the other required variables. Moved code that computed the 
	PSTH for the surrogates from the plot function here. The result is 
	saved to the result file so the plot function just has to load it and 
	plot it. Moved the computation of RasterMarkMatrices from the 
	constructor to here to speed up creation of objects. 
@shufflesync/shufflesync.m: Now saves binsc in the data file so we can plot 
	the PSTH for the data easily. Now saves variables used in the surrogate 
	computation that depend on input arguments into a separate file named 
	Args.SurrData. This way, the data file can contain only variables that 
	are totally dependent on the data and so different runs with different 
	arguments can share the same data file. Now checks for the presence of 
	the data file and will only recompute the histogram if the data file is 
	not present or if the RedoDataFile argument is specified. Now calls 
	shsywrapper shell function with just window numbers and the surrogate 
	data file which now contains all the necessary information to invoke 
	shufflesyncsurr and computeSurrData functions.  Moved code that 
	computed the RasterMarkMatrices to the computeSurrData function to 
	speed up object creation since most objects are being created on the 
	fly in order to compare the results of using different shuffle values. 
@shufflesync/private/shsywrapper: Now echoes the job id for computeSurrData 
	as well as the number of jobs it is waiting for. 
@shufflesync/shufflesync.m: Renamed datafile variable saved in surrogate 
	data file to cellpairdatafile since there is a class called datafile 
	and it seems to be conflicting with the use of that variable name. 
@shufflesync/private/computeSurrData.m: Renamed datafile variable saved in 
	surrogate data file to cellpairdatafile since there is a class called 
	datafile and it seems to be conflicting with the use of that variable 
	name. 
@shufflesync/private/shufflesyncsurr.m: Renamed datafile variable saved in 
	surrogate data file to cellpairdatafile since there is a class called 
	datafile and it seems to be conflicting with the use of that variable 
	name. 

June-03-2005
@shufflesync/shufflesync.m: Now loads DataFile and ResultFile directly into 
	workspace instead of assigning to another variable to make the data 
	more easily accessible. Now converts mark matrices to logical matrices 
	since computeSurrData now saves the mark matrices as sparse matrices to 
	save disk space. Renamed object structure variable from data to objdata 
	to avoid conflict with data variable saved in the ResultFile. Updated 
	to use the variable name dzscore to access the z-scores in the 
	ResultFile. 
@shufflesync/private/computeSurrData.m: Now only computes number of 
	synchronous spikes for relevant windows to reduce redundant 
	calculations. Renamed zscore variable to dzscore to avoid conflict with 
	the zscore function. Now saves individual mark matrices as sparse 
	matrices since we can't create sparse matrices for cell arrays. Using 
	sparse matrices will allow us to cut down on disk usage. 
@shufflesync/plot.m: Now loads data from result file to plot PSTHs. Shifted 
	XCorr code above Try to keep it closer to PSTH code so it is easier to 
	edit. 
@shufflesync/plot.m: Now converts PSTH from spike counts to spike rates in 
	Hz. Now uses nanindex to plot psth matrices with nan values so the 
	number of points plotted is reduced. Moved development code in PSTH 
	section to Args.Try. 
@shufflesync/private/computeSurrData.m: Now creates matrices filled with 
	nan for computing PSTH so we can reduce number of plots plotted. 

June-08-2005
@shufflesync/private/computeSurrData.m: Added code to compute tests for 
	normality for each xcorr phase lag in each window. 
@shufflesync/private/shsywrapper: Added code to perform computations on the
	local disks of each ogier node. 
@shufflesync/plot.m: Added option to use the results of jbtest or 
	lillietest to determine which points should be highlighted. 
@shufflesync/private/computeSurrData.m: Now converts jbmat and limat to 
	sparse matrices before saving to reduce disk usage. 

June-09-2005
@shufflesync/private/computeSurrData.m: Removed hpsth variable (which was 
	used to store the h value from the psth test) since we don't want to 
	have to rerun this function to try another p value. This can be done by 
	the plot function by using the p values returned from the tests. Now 
	saves the p values returned by jbtest and lillietest instead of the h 
	value. As a result, in order not to confuse real p values of 0, the 
	matrices are initialized with nan instead of zeros and are no longer 
	converted to sparse matrices since the presence of the nan's will not 
	lead to much reduction in disk usage. 
@shufflesync/plot.m: Added optional argument to specify p value for the 
	test to see if the surrogate distribution for each xcorr phase lag is 
	normal. Added alias argument to specify p value for both psth and 
	surrogate distribution tests. Now uses the p-values computed by 
	computeSurrData and saved in the result file to determine which bins 
	have surrogate distributions that are normal. The bins that are above 
	the zscore threshold and have similar psth's but have surrogate 
	distributions that are not normal are now plotted in white. 
@shufflesync/private/computeSurrData.m: Now converts binidx to non-sparse 
	matrix so that the logical comparisons work. 
@shufflesync/plot.m: Temporarily added back code to use jbmat and limat 
	from old results files. Added code to Try section to display normal 
	and non-normal distributions. 
@shufflesync/private/computeSurrData.m: Removed conversion of sparse 
	matrix since logical comparisons work fine with sparse matrices. The 
	problem I had run into before was that cell arrays cannot be converted 
	to sparse matrices. 
@shufflesync/private/computeSurrData.m: Now explicitly converts sparse 
	matrices in l.sptrains to full matrices to make sure kstest2 does not 
	crash. Sparse matrices are possible now that binidx is saved in 
	shufflesync as a sparse matrix since the code in shufflesyncsurr 
	sometimes indexes into binidx directly and saves the result into 
	l.sptrains. 

June-20-2005
@shufflesync/private/shsywrapper: Added code to lauch computations on 
	subnodes using batch (commented out right now since we are still doing 
	this by hand for now). 
@shufflesync/plot.m: Now thresholds the p values returned by 
	computeSurrData for the normality test. Now plots the truly significant 
	bins last to improve visibility. Added code to Try section to plot the 
	distributions that are normal and those that are not, as well as their 
	p-values to see what the jbtest is doing. 

June-23-2005
@variability/plot.m: Updated call to getDataDirs to shorten names by adding 
	'DirString'. 
@shufflesync/plot.m: Updated call to getDataDirs to shorten names by adding 
	'DirString'. 
@latency/plot.m: Updated call to getDataDirs to shorten names by adding 
	'DirString'. 
@adjspikessync/plot.m: Updated call to getDataDirs to shorten names by 
	adding 'DirString'. 
@refractory/plot.m: Updated call to getDataDirs to shorten names by adding 
	'DirString'. 
@adjspikesxcorr/plot.m: Updated call to getDataDirs to shorten names by 
	adding 'DirString'. 
@frdiff/plot.m: Updated call to getDataDirs to shorten names by adding 
	'DirString'. 

July-13-2005
@sparsity/heterogeneity.m: Removed unused options IntraGroup and 
	InterGroup. Now calls groupDirs without the IntraGroup argument since 
	that is now the default. Removed calls to groupDirs with the InterGroup 
	argument since that option was never implemented in groupDirs. 
@responsiveness/heterogeneity.m: Removed unused options IntraGroup and 
	InterGroup. Now calls groupDirs without the IntraGroup argument since 
	that is now the default. Removed calls to groupDirs with the InterGroup 
	argument since that option was never implemented in groupDirs. 
@shufflesync/shufflesync.m: Now uses the shortcut AND operator (&&) to skip 
	checking the other condition if the first condition is false. 
@frdiff/frdiff.m: Now uses isempty(nptDir(...)) instead of ispresent since 
	ispresent is slower when there are a lot of files in the directory. Now 
	also uses the shortcut AND operator (&&) to skip checking the other 
	condition if the first condition is false. 
@frdiff/plot.m: Added option to create heterogeneity plots. 

July-19-2005
@frdiff/plot.m: Added option to plot PSTH correlation coefficients, 
	including differentiating between those with signicant and 
	non-significant values. Added option to plot PSTH correlation 
	coefficients against vector angle. Added code to mark values in the 
	heterogeneity plot of the PSTH correlation coefficients that are not 
	significant. 
@frdiff/frdiff.m: Modified comments to differentiate between signal and 
	noise correlation. Code for noise correlation still yet to be 
	completed. 
@frdiff/frdiff.m: Now calls getDataDirs to generate cluster directories 
	using the current directory so we can just use ProcessDirs to create 
	the objects instead of having to use ProcessDays. Now passes the 
	'NoiseCorr' flag to the firingrate constructor to make sure that the 
	difference between the spike counts on individual trials and the 
	average of all the trials is computed. Now computes and saves the noise 
	correlation coefficients and p-values, in addition to the signal 
	correlation values (but only for pairs). 
@firingrate/private/CalcFiringRate.m: Added 'NoiseCorr' option to compute 
	difference between spike counts on each trial and the average over all 
	trials. When that option is specified, the meancounts field is the 
	noise counts. 
@frdiff/plot.m: Added option to plot noise correlation data. Modified code 
	to mark insignificant correlation coefficients to include noise 
	correlation data. 
@frdiff/plus.m: Modified to add noise correlation data fields. 
@firingrate/firingrate.m: Added 'NoiseCorr' option to compute difference 
	between spike counts on each trial and the average over all trials. 
	When that option is specified, the meancounts field is the noise 
	counts. 

July-21-2005
@frdiff/plot.m: Added code to highlight non-significant correlation 
	coefficients in the histograms. This uses stacked bar plots with 
	modifieid bins in order to simulate the histc option to bar, which does 
	not work with stacked bar plots. For other histogram plots with control 
	data, the data is now plotted using the bar command with the histc 
	argument and the control data is plotted using the stairs command so as 
	to emphasize that it is a different distribution. 
@frdiff/plot.m: Added comments to explain why we do stacked bar plots for 
	histograms of correlation coefficients and bar/stairs combination for 
	other histograms. 

July-26-2005
@frdiff/plot.m: Now separates out significant and insignificant correlation 
	values only if we are plotting histograms otherwise the GroupedIntra 
	option will not work since it tries to index all the data. 
@firingrate/firingrate.m: Added spike_matrix field so that we can compute 
	the correlation between individual windows in frdiff. 
@adjspikespsth/plot.m: Removed code that changed directories before 
	instantiating stiminfo object since the stiminfo object now changes 
	directories automatically. 
@frdiff/frdiff.m: Added code to compute noise correlations for individual 
	windows. 
@firingrate/plot.m: Commented out conversion from spike counts to rate 
	since that is specified when calling the constructor. 
@firingrate/private/CalcFiringRate.m: Added spike_matrix output so that we 
	can compute the correlation between individual windows in frdiff. 
	Replaced code to compute firing rates in specified window sizes and 
	overlap with vectorized code that was previously moved to the end of 
	the function and not used. Modified code to deal with overlaping 
	windows. Now aligns overlapping windows to the start of the window 
	instead of the center of the window, which would change depending on 
	whether the window size used was an even or odd number. Added 
	'NoiseCorr' option to compute deviations from mean count across 
	repetitions for each window so we can compute noise correlations in the 
	frdiff object. 
@frdiff/plot.m: Added option to plot windowed noise correlation data. Now 
	plots data using time instead of data points so we can align the data 
	with plots from other objects. Now uses stairs instead of plot so it is 
	clearer that we are plotting windowed data. Now plots the control data 
	using dashed lines to improve visibility of the data under the control 
	data. 
@frdiff/plus.m: Now concatenates the new fields winnoisecorr, 
	winnoisecorrp, and timebins. 
@frdiff/frdiff.m: Now saves the timebins from the firing object so we can 
	plot the frdiff data in time. This way we will be able to align the 
	plots with other objects using linkedzoom. 

July-27-2005
@frdiff/plot.m: Added option to plot the average firing rate of two cells. 
	Added option to plot the 2D histogram of average rate versus noise 
	correlations. 
@frdiff/frdiff.m: Added option to threshold the firing rate of each cell 
	before computing the windowed noise correlation only on the bins that 
	are above threshold for both cells. 
@firingrate/private/CalcFiringRate.m: Now correctly converts meancounts to 
	rate as well when using frame delimited windows with no overlap so that 
	PSTH plots work properly. 

July-28-2005
@frdiff/plot.m: Moved code to threshold firing rates of individual cells 
	before computing the windowed noise correlation from the constructor to 
	the plot command so we don't have to keep regenerating the object in 
	order to try out different thresholds. 
@frdiff/frdiff.m: Moved code to threshold firing rates of individual cells 
	before computing the windowed noise correlation to the plot command so 
	we don't have to keep regenerating the object in order to try out 
	different thresholds. Now saves the individual cells firing rate so we 
	can threshold the rates in the plot function. 
@frdiff/plus.m: Now adds the frmat and frSetIndex fields added to allow the 
	plot function to threshold the firing rates of the individual cells 
	before computing the windowed noise correlation. 

July-30-2005
@adjspikespsth/plot.m: Fixed bug that was not changing directories to the 
	cluster directory before trying to instantiate the stiminfo object. 
@frdiff/plot.m: Changed default RatePercentile to 75. Changed 
	HistFRNoiseCorr option to HistWinFRNoiseCorr to prevent confusing 
	windowed noise correlations with the PSTH noise correlation. Added 
	option to create scatter plot of the windowed noise correlations versus 
	the average rate in each window. The 2D histogram now plots only points 
	that are significant and above the rate percentile threshold. Now plots 
	the firing rates of each cell, as well as the thresholds for each cell, 
	in addition to the average rate. This way, we don't have to use a 
	nptgroup object to plot the firing rates of each cell. Added code to 
	plot histograms of the windowed noise correlations for all bins above 
	threshold for all cell pairs. 
@frdiff/plot.m: Now labels all windowed noise correlation plots as 
	'Windowed Noise Correlation'. 

August-02-2005
@frdiff/plot.m: Added heterogeniety boxplot of windowed noise correlation 
	values, which also displays the PSTH noise correlations. Modified all 
	code to check for significance to find values that are strictly less 
	than CorrPThresh. 

August-03-2005
@frdiff/plot.m: Added 3D scatter plot of percentile rates versus windowed 
	noise correlations. Now uses average rate percentile instead of average 
	rate in the 2D histograms of rate versus windowed noise correlations. 
	Fixed bug that failed to properly initialize windowed noise correlation 
	data when specifying HistBinsY. 

August-04-2005
@frdiff/plot.m: Changed RatePercentile argument back to 75 so that prctile 
	function call will work properly. The conversion from firing rates to 
	percentiles now includes scaling by 100 so we can use the same 
	RatePercentile argument. Added CLims argument to specifiy the scale of 
	imagesc so we can plot the intra-group and inter-group pairs on the 
	same color scale. Added code to create 2D histograms and 3D 
	scatterplots of windowed noise correlations versus rate percentile for 
	all pairs in an object. Fixed bug in the conversion from rate to 
	percentiles that failed to correctly calculate the number of bins in 
	each data set. Changed axis labels from Percentile Rate to Rate 
	Percentile since the latter seems more correct. Now correctly labels 
	the 2D histogram plots as Mean Rate Percentile instead of Mean Rate. 
	Now uses dotted lines to draw the percentile threshold in the FRMean 
	option so as to avoid obscuring the actual firing rates. Removed 
	"intra-group" from the title of the VAngCorr plot since it can also be 
	used for inter-group pairs. 

August-05-2005
@adjspikes/plus.m: Added code to ignore the new TossedSpikes when adding 
	old objects without that field. Empty arrays are used instead. 
@frdiff/plot.m: Added option to ignore 0's when computing the rate 
	percentile. 
@frdiff/plot.m: Now uses the percent sign instead of the word "percentile" 
	in the axis labels. 
@frdiff/plot.m: Fixed missing data points in the firing rate plots due to 
	zero bins being replaced by nan's when the NoZerosPercentile argument 
	is used. The percentile calculation now uses a copy of the frmat 
	variable so the plot is not affected.  

August-08-2005
@frdiff/plus.m: Modified to merge the new wincv field. 
@frdiff/plot.m: Renamed FRWinNoiseScatter3 argument to WinFRNoiseScatter3 
	to be more consistent with other arguments. Commented out code to 
	create 3D plots since it is hard to interpret the values. Now uses 
	scatter command instead with rate (in percent) of first cell versus the 
	second cell and the color of each point being the windowed noise 
	correlation value. Fixed bug that was using nbins, which could be 
	different for different cells if we are not including zeros in our 
	percentile calculation, instead of frmrows to create data vectors. 
	Added code to plot cv in percent of cell 1 versus cell 2 along with the 
	windowed noise correlation values in color. 
@frdiff/frdiff.m: Added code to compute coefficient of variation across 
	repetitions for each bin and store in new wincv field. 
@frdiff/frdiff.m: Now saves the standard deviation instead of the cv since 
	we can always get the cv by dividing the standard deviation by frmat. 
	Renamed data field wincv to winstd. 
@frdiff/plot.m: Added option to plot CV or standard deviation. Updated to 
	use winstd field instead of wincv. Now correctly labels CV and standard 
	deviation plots. 
@frdiff/plus.m: Updated to merge winstd field instead of wincv. 
@frdiff/frdiff.m: Now correctly removes the scaling factor of 100 when 
	saving the standard deviation. 
@frdiff/plot.m: Now correctly checks for the WinStdNoiseScatter option. 
	Added plot of standard deviation as a function of time, and modified 
	the plot of CV as a function of time to make use of the new winstd 
	field. 

August-09-2005
@firingrate/private/CalcFiringRate.m: Now correctly converts spike_matrix 
	to rate when the Rate argument is specified. 
@frdiff/plot.m: Added code to specify the scale of the colobar for scatter 
	plots using the CLims argument. Renamed NoZerosPercentile argument to 
	PercentileWithZeros since we are mostly calculating percentiles without 
	0's now. Renamed ScatterNoInsigP argument to ScatterWithInsigP since we 
	are mostly leaving out the insignificant values now. Added XLim and 
	YLim arguments to set the x- and y-axes limits. Added FontName, 
	FontSize, NoTitle and TickDir arguments to format plots for 
	manuscripts. Fixed bug that was trying to use frmat before initializing 
	it. 

August-20-2005
@frdiff/plot.m: Added option to compute firing rate differences only for 
	bins that are above the threshold for firing rate for both cells. 
	Reorganized code so that the conversion of firing rates to percentile 
	is performed at the beginning of each section (i.e. 
	if(isempty(Args.NumericArguments))) instead of in each separate option 
	inside each section. 

August-23-2005
@shufflesync/shufflesync.m: Added help comments to document the various 
	optional input arguments. 

August-24-2005
@shufflesync/shufflesync.m: Now recalculates surrogate data file, and data 
	file if RedoDataFile is also specified, even when the result file is 
	present. This way we can regenerate the files without having to remove 
	the result files. 

September-01-2005
@shufflesync/private/shufflesyncsurr.m: Modified comments to correct 
	grammar. Commented out redundant else statement. 
@shufflesync/plot.m: Modified comments to clarify the different subsets of 
	data that are plotted. 
@frdiff/plot.m: Removed "%" from labels of CV plots. Now correctly 
	transposes the indices of bins in which the firing rate of both cells 
	exceed the threshold before applying to frdiff which is organized in 
	rows instead of columns. This applies only to the DiffThreshBins 
	option. 
@shufflesync/private/computeSurrData.m: Modified comments to clarify 
	algorithm steps. 

September-08-2005
@frdiff/plus.m: Now adds the vector angle p-values. 
@frdiff/frdiff.m: Now saves the the p-value returned by the vector 
	similarity calculation. 
@latency/plot.m: Added option to specify threshold for cells with refresh 
	or frame rate artifacts. Now looks for frame rate power as well. Added 
	option to specify number of refreshes per frame for the different 
	refresh rates. Fixed bugs that was not selecting just the first index 
	that exceeded the refresh rate. Commented out code that used the mean 
	of the frequencies above and below the refresh rate to check for 
	artifacts since the resulting ratio was much closely spaced than just 
	using the frequency above the refresh rate. Now marks the cells with 
	refresh artifacts with an asterisk in the title when doing population 
	plots. 
@frdiff/plot.m: Now plots the firing rate differences along with the rates 
	of the individual cells when using the FRMean option. Now uses the 
	p-values for the vector angle calculation to plot the significant and 
	non-significant vector angles. 
@frdiff/frdiff.m: Added code to compute Spearman correlation coefficient. 
@frdiff/plus.m: Added code to add the Spearman correlation coefficients. 
@frdiff/plot.m: Added code to plot Spearman correlation coefficients. 

September-22-2005
@shufflesync/plot.m: Added ColorBar and CLims optional input arguments to 
	add colorbar and to specify color scale used in the pcolor plots.
	
October-18-2005
@frdiff/plot.m: Added NoDiff option to plot PSTH for both cells without 
	the difference so we can look at the bins that are above threshold
	for both cells more clearly. Now plots only the bins that are above
	threshold when the FRMean option is specified withe DiffThreshBins
	option. Moved DiffThreshBins calculations for the population data to
	the start of the population data section so that all applicable
	options like Corr, SpCorr, VAng, will work.

October-21-2005
@shufflesync/private/computeSurrData.m: Reordered initialization of psths 
	cell array so we eliminate need to alter memory size. 
@shufflesync/shufflesync.m: Updated help comments for the WindowIncrBins 
	and WindowBins arguments to use number of bins instead of ms. Added 
	code to compute extended window limits so we can compute the xcorr 
	without boundary effects. Now loads the data file if it is present in 
	order to have access to required variables. 
@shufflesync/private/shufflesyncsurr.m: Now uses the extended window limits 
	to check for repetitions with spikes in both cells. Might need to be 
	refined to find repetitions in which there are spikes in the central 
	window of one cell and spikes in the extended window of the other cell. 
	Optimized code to remove repeated find calls. Now uses the information 
	returned from unique and intersect to identify spikes. Optimizations 
	were applied to the following sections: xcorr of the data, xcorr of the 
	shift predictor and shuffling of the remainder of the repetitions. 
	Moved seeding of the random number generator out of the loop over 
	repetitions with paired spikes since we are not likely to exceed the 
	number of random numbers over all repetitions. Now initializes sptimes 
	cell array by setting the second entry first. Now uses a separate 
	variable called sptime when shuffling the remaining repetitions 
	(instead of using sptimes) so there is no confusion when the second 
	entry of sptimes is initialized at the beginning. 

October-25-2005
@shufflesync/private/shufflesyncsurr.m: Added code to compute xcorr without 
	boundary effects using central and extended windows. Now also checks 
	for refractory period violations in the extended window when generating 
	surrogates. 
@shufflesync/shufflesync.m: Fixed bug in call to the max function. Now 
	saves startlimits and endlimits in the data file so shufflesyncsurr 
	does not have to recreate it each time. 

October-26-2005
@shufflesync/private/shufflesyncsurr.m: Added check for empty trialspiken 
	before trying to use sub2ind with empty matrices. Now correctly sets 
	the start and end limits for all reps with spikes instead of just the 
	reps in repsboth. This way, the rest of the code can use the same start 
	and end limits to get the spike times. Added several fixes to get 
	around problems working with empty matrices. 
@shufflesync/shufflesync.m: Now no longer uses sparse matrices for the 
	endlimits array since code using empty matrices was not working with 
	sparse matrices. 
@shufflesync/shufflesync.m: Removed code that have been commented out. 
ShihChengChangeLog.txt: 
@shufflesync/private/shufflesyncsurr.m: Removed code that have been 
	commented out. 
@shufflesync/private/shufflesyncsurr.m: Added fixes to the shift predictor 
	code to correctly handle empty matrices. 

October-27-2005
@shufflesync/private/computeSurrData.m: Added check for empty spiketrains 
	which are now possible due to the use of extended windows. 
@shufflesync/plot.m: Fixed bug in which the input argument SyncSpikes was 
	not declared a flag. Added Windowed option to plot data window by 
	window (not completed). 
@shufflesync/private/launchshsy: New bash script to use batch to run 
	shufflesyncsurr and computeSurrData on a node.
@shufflesync/private/shsynodelaunch: New bash script to divide all pairs 
	among the available nodes and then launch the computations by calling 
	launchshsy on each node. 
@shufflesync/private/computeSurrData.m: Moved computation of mean of 
	histogram counts to avoid error when there were no spikes. 
@shufflesync/private/launchshsy: Replaced pathname with "~" with absolute 
	path. 
@shufflesync/private/launchshsy: Now correctly returns to the original 
	directory before trying to move to the next directory. 
@shufflesync/private/launchshsy: Now includes the correct path prefix when 
	finding data files to copy. 
@shufflesync/private/shsynodelaunch: Now no longer removes n4.txt which 
	contains the location of each dataset so we can figure out where to go 
	to get access to the surrogate data. 
@shufflesync/private/launchshsy: Now changes directory to the data 
	directory before issuing find command so that cpio will copy paths 
	relative to the data directory instead of including the entire 
	/home/syen/cat/newcatdata path. 
@shufflesync/private/launchshsy: Now correctly specifies the full path to 
	the file s1.txt for removal. 

October-28-2005
@shufflesync/private/shufflesyncsurr.m: Now uses histcie with the 
	'DropLast' option instead of hist to ignore the values outside the 
	range of interest. 
@shufflesync/shufflesync.m: Now creates and saves lagbins2 which will be 
	used with histcie to ignore values outside the range of interest. This 
	was necessary since doing hist with the data from extended windows was 
	leading to very large values in the first and last bins since they 
	include all the values outside the range of interest. 
@shufflesync/private/shufflesyncsurr.m: Fixed bug where hist was not 
	changed to histcie. 
@shufflesync/private/shufflesyncsurr.m: Now explicitly converts xcreps to 
	full matrix to avoid errors in histcie when xcreps is a sparse matrix. 
@shufflesync/plot.m: Now only plots the black (significant) dots by default 
	when using the XCorr argument. The other dots can be plotted when using 
	the new AllZScore3 argument. 

October-30-2005
@shufflesync/plot.m: Added code to plot the xcorr for the data and the 
	shift-predictor versus the distributions from the surrogates. 
@shufflesync/get.m: Added code to return number of analysis windows so that 
	we can use InspectGUI to pan through the different windows in a cell 
	pair. 
CatCmds.m: Merged with Charlie's version. 

October-31-2005
@shufflesync/plot.m: Added HighlightRasters argument so the default raster 
	plot does not highlight spikes, which can sometimes be hard to see. 
	Added code to display windowed rasters which can be used in combination 
	with the display of windowed xcorr to check the data. 
@shufflesync/plot.m: Added XLim argument so we can plot the windowed xcorr 
	on the same scale. Otherwise, because of the different shuffle 
	magnitude, the time lags will all be different. Padded data for pcolor 
	plot in windowed xcorr so that it now correctly displays all data and 
	labels the bins correctly. Now computes the maximum spike pairs in each 
	xcorr window before computing the histogram so that the points 
	representing the data will be visible. Now plots xcorr data exceeding 
	z-score threshold in black and everything else in white. Added colorbar 
	so we can compare data from different shuffle magnitudes. 

November-01-2005
@shufflesync/private/shufflesyncsurr.m: Fixed bug that was not saving spike 
	times from the data when creating shuffled surrogates for repetitions 
	not shuffled in the xcorr analysis.  
@shufflesync/plot.m: Moved PSTH section to after Windowed option so we can 
	use Windowed and PSTH to plot windowed PSTH data. Added code to plot 
	windowed PSTH (incomplete). 

November-04-2005
@shufflesync/plot.m: Added input argument to specify PSTH smoothing window. 
	Added code to plot windowed psth of data as well as the distribution of 
	each bin in the psth for the surrogates. 
@shufflesync/shufflesync.m: Now saves subbins and binSize which is used by 
	the plot function to align the windowed psth with the rasters. 

November-06-2005
@shufflesync/shufflesync.m: Now takes a PSTHWindow input argument which is 
	then saved to the surrogate data file so that it can be used by the 
	function computeSurrData. Now correctly recomputes data files even when 
	the result file is present. 
@shufflesync/private/computeSurrData.m: Added code to compute z-scores for 
	each bin of the smoothed PSTH's of both cells by comparing with the 
	mean and std of each bin in the smoothed PSTH's of the surrogates. The 
	result is now saved to the result file. 
@shufflesync/plot.m: Removed redundant lines of code and code that had been 
	commented out. 
@shufflesync/private/computeSurrData.m: Now correctly loads previous frame 
	for window numbers larger than 2 rather than 1. 

November-07-2005
@shufflesync/plot.m: Renamed AllZScore3 argument to AllZ3 to make it easier 
	to use. Added SmoothedPSTH argument to use the smoothed PSTH to look 
	for differences between the data and the surrogates. 
@shufflesync/plot.m: Added Z3Norm and Z3Psth arguments to turn on and off 
	bins with z-scores greater than 3 but do not meet one or the other of 
	the normality or psth-difference constraint. Now correctly finds the 
	bins in which the psth-difference is not above the z-score threshold. 

November-08-2005
@shufflesync/name2index.m: New function to convert time to window number so 
	we can enter a time while inspecting the windowed xcorr or psth and 
	have it plot the relevant window. 
@shufflesync/plot.m: Now uses white for significant bins in the xcorr and 
	black for the rest of the values to improve visibiltity on the dark 
	backgrounds. Added hack to try to relax the PSTH difference constraint. 
@shufflesync/private/computeSurrData.m: Now correctly uses PSTHWindow size 
	minus 1 to pad the PSTH and to extract the final values from the 
	smoothed PSTH. 
@shufflesync/plot.m: YLim now defaults to empty matrix instead of [-10 10] 
	so we can just check for empty matrix to see if we should use the range 
	in the data. This was necessary in order to fix the y-axis of the 
	windowed psth and xcorr plots. Added checks for empty sptrains in the 
	surrogate files as well as other minor bug fixes taken from recent 
	changes to computeSurrData.m. Now uses the binSize field saved in the 
	object since it is now the actual bin size instead of the requested bin 
	size. Now correctly loads the previous frame only for windows greater 
	than 2. Now correctly uses PSTHWindow size minus 1 to pad the PSTH and 
	to extract the final values from the smoothed PSTH. Added code to 
	convert PSTH to rates so we can plot them with the right scale. 
	Removed shift predictor data points from the windowed xcorr plot to 
	improve visibility. Now plots the data using a line plot. Now 
	highlights significant bins with a white square, similiar to the 
	windowed PSTH plot. 
@adjspikesxcorr/plot.m: Updated getDataDirs call to include 'DirString' 
	argument. 
@shufflesync/shufflesync.m: Now saves the actual bin size used to partition 
	each frame window as well as the number of stimulus repetitions. This 
	makes it easier for the plot function to access these frequently used 
	variables instead of having to load them from the data file. The 
	requested bin size which used to be stored in obj.data.binSize is still 
	available in obj.data.Args.BinSize. 

November-09-2005
@shufflesync/plot.m: Now only includes psth bins that will be used by a 
	window of size Args.PSTHWindow centered on the first bin in the central 
	window. Now selects the bins that correspond to centering the 
	PSTHWindow on the first and last bin in the central window. 
@shufflesync/private/computeSurrData.m: Added use of halfwinsize again to 
	correctly pad with just enough bins and to extract the right indices 
	representing the central window. 
@shufflesync/plot.m: Now only converts counts to rate when plotting so as 
	to avoid problems with rounding when computing histograms and to avoid 
	having to specify the bin size for the histograms.  
@shufflesync/plot.m: Now checks for empty YLim before use in XCorr plot. 
	Added missing semicolon after computing ratescale variable. 
@shufflesync/name2index.m: Added else statement so output argument will be 
	set by default. 

November-10-2005
@shufflesync/plot.m: Added ExpandPSTH option to skip resizing the windowed 
	PSTH plot to match the central window in the raster plot. This was 
	necessary so we can increase the visibility of the PSTH when not 
	plotted with the rasters. Added fix for pcolor bug in Matlab by 
	changing Renderer to Painters. Added MinXCCount argument to change the 
	minimum number of spike pairs to use as a constraint instead of the 
	noramality test. Now if the NormTest is anything but jbtest or 
	lillietest, it will use the minimum spike pairs constraint. This way 
	we can actually make use of the else statement in the code. Now plots 
	the windowed xcorr and psth data using larger symbols for increased 
	visibility. Added xlim call to xcorr plot in case we ever need to 
	constrain the time window. 

November-12-2005
@shufflesync/plot.m: Added option to plot xcorr data only without 
	highlighting bins that were significant. This was necessary to create 
	figures for the SFN05 talk. Added option to not show the title (again 
	for the SFN05 talk). Now considers psth to be different from the psth 
	of the surrogates if either of the cells contain two or more bins that 
	exceed the z-score threshold. This seems to work a lot better in not 
	rejecting too many significant correlations. Z3Psth option now plots 
	bins above the z-score threshold and found to be normally distributed, 
	or have counts larger than the minimum. Added code to remove the x-axis 
	labels when doing group plots. 
@shufflesync/plot.m: Added option to create summary plot over all shuffle 
	values. 

November-21-2005
@shufflesync/plot.m: Changed CLims input argument to CLim to make it more 
	consistent with XLim and YLim input arguments. Modified histogram bins 
	passed to histn so they correctly specifiy the bins edges instead of 
	the center of the bins. Added code to use CLim argument to set the 
	color scale when plotting summary plots. Now no longer removes the 
	x-axis unless the LabelsOff argument is specified. The old code that 
	checked for GroupPlotIndex did not work as it would always remove the 
	x-axis since the different group plots (corresponding to the different 
	shuffle values) are plotted on the same axis. 
@frdiff/plot.m: Added code to compute mean and max differences between 
	normalized firing rates. Added code to return arbitrary variables so we 
	can grab the normalized data for distribution testing. 

December-07-2005
@shufflesync/name2index.m: Renamed prefix to specify time to "t: " instead 
	of "w: " to make it more intuitive.  
@firingrate/private/CalcFiringRate.m: Fixed bug that was not saving the 
	proper histogram bin edges to the variable timebins. 

December-13-2005
getSurrogateSpikes.m: Added fprintf statement to track down 0 isi's. 

March-20-2006
@psparse/subsref.m: New class to compute population sparseness. 
@psparse/subsasgn.m: New class to compute population sparseness. 
@psparse/psparse.m: New class to compute population sparseness. 
@psparse/get.m: New class to compute population sparseness. 
@frdiff/plot.m: Added option to plot median differences instead of mean 
	differences. Transposed frdiffs and frdiffctrl so it is more consistent 
	with other variables. Updated code that used frdiffs and frdiffctrl to 
	take into account the transpose operation. Modified axis labels to be 
	more consistent with labels in figures for the heterogeneity paper. 
@psparse/test.m: New class to compute population sparseness. 
@psparse/plus.m: New class to compute population sparseness. 
@psparse/plot.m: New class to compute population sparseness. 

April-20-2006
@shufflesync/name2index.m: Now correctly looks for "t:" instead of "w:" in 
	the strncmpi command. 

April-24-2006
@variability/cmds.m: New file to save commands relevant to this class. 
@psparse/test.m: File renamed to cmds.m. 
@adjspikesxcorr/cmds.m: New file to save commands relevant to this class. 
@psparse/cmds.m: New file to save commands relevant to this class. 
@shufflesync/cmds.m: New file to save commands relevant to this class. 
@frdiff/cmds.m: New file to save commands relevant to this class.

May-02-2006
@poissonctrl/plus.m: New class to compute PSTH correlation and vector angle 
	for Poisson model.
@poissonctrl/subsasgn.m: New class to compute PSTH correlation and vector 
	angle for Poisson model.
@poissonctrl/poissonctrl.m: New class to compute PSTH correlation and 
	vector angle for Poisson model.
@poissonctrl/plot.m: New class to compute PSTH correlation and vector angle 
	for Poisson model.
@shufflesync/cmds.m: 
@poissonctrl/cmds.m: New class to compute PSTH correlation and vector angle 
	for Poisson model.
@poissonctrl/get.m: New class to compute PSTH correlation and vector angle 
	for Poisson model.
@poissonctrl/subsref.m: New class to compute PSTH correlation and vector 
	angle for Poisson model.

May-04-2006
@poissonctrl/poissonctrl.m: Added code to compute the joint event 
	probability. Added option to turn specific calculations on or off. Now 
	only runs the correlation analysis for default. Now also saves the mean 
	firing rate for the site so that we can create comparisons in the 
	instantaneous firing rate difference analysis. 
@poissonctrl/cmds.m: Added command to compute joint event probablity of the 
	Poisson null hypothesis. 
@jointevprob/cmds.m: New class to compute joint events using PSTH computed 
	at the frame rate. This allows us to compare the data to the data 
	obtained from the Poisson null hypothesis (calculated in the class 
	poissonctrl). 
@jointevprob/subsref.m: New class to compute joint events using PSTH 
	computed at the frame rate. This allows us to compare the data to the 
	data obtained from the Poisson null hypothesis (calculated in the class 
	poissonctrl). 
@jointevprob/plus.m: New class to compute joint events using PSTH computed 
	at the frame rate. This allows us to compare the data to the data 
	obtained from the Poisson null hypothesis (calculated in the class 
	poissonctrl). 
@jointevprob/get.m: New class to compute joint events using PSTH computed 
	at the frame rate. This allows us to compare the data to the data 
	obtained from the Poisson null hypothesis (calculated in the class 
	poissonctrl). 
@frdiff/cmds.m: Removed commands to compute Poisson model for inter-group 
	data since it is not really relevant. The Poisson model only tests for 
	Poisson variability around a mean rate for neighboring cells. Added 
	code to compare instantaneous firing rate differences in the data to 
	the Poisson model for each site instead of the Poisson model for each 
	pair. 
@poissonctrl/plus.m: Updated to include fields for the joint event 
	probability. 
@jointevprob/subsasgn.m: New class to compute joint events using PSTH 
	computed at the frame rate. This allows us to compare the data to the 
	data obtained from the Poisson null hypothesis (calculated in the class 
	poissonctrl). 
@jointevprob/plot.m: New class to compute joint events using PSTH computed 
	at the frame rate. This allows us to compare the data to the data 
	obtained from the Poisson null hypothesis (calculated in the class 
	poissonctrl). 
@jointevprob/jointevprob.m: New class to compute joint events using PSTH 
	computed at the frame rate. This allows us to compare the data to the 
	data obtained from the Poisson null hypothesis (calculated in the class 
	poissonctrl). 
@jointevprob/cmds.m: Added call to plot function.
@poissonctrl/cmds.m: Added commands to create heterogeneity plots for psth 
	correlation and joint event probability along with the data from the 
	Poisson null hypothesis. 

May-09-2006
@poissonctrl/cmds.m: Moved code to compute event probabilities to 
	poissonctrl constructor. Added code to compute percentage of windows 
	with population sparseness z-scores greater than 2. 
@poissonctrl/plus.m: Moved order of fields so ncells is added first. 
	Updated to add pszscore and kwp fields which are used by the 
	population sparseness calculation. 
@poissonctrl/poissonctrl.m: Added option to compute population sparseness 
	for Poisson control. The surrogate mean and std PS values are used to 
	convert data PS values to z-scores. Kruskal-Wallis test is also 
	performed to test if the data is different from the surrogates. Added 
	code to either set negative surrogate firing rates to 0 (default) or 
	to take the absolute value if the optional input argument 'Abs' is 
	specified. Added 'UseNCells' option so that population sparseness is 
	computed for all cells at a site. Previously, the code performed all 
	calculations for pairs. Now sets the default values to nan so if a 
	calculation is skipped the data in the object won't be misinterpreted. 
@poissonctrl/cmds.m: Added commands to compute percentage of windows with 
	rates higher than the 75th percentile and have z-scores greater than 2. 

May-22-2006
@variability/cmds.m: Added comments to code that computes which cells have 
	0 ISIs. 
@adjspikes/plus.m: Commented out code that concatenates the dlist field 
	since it is an irrevelant field that does not exist in newer objects. 

June-05-2006
CatCmds.m: Moved shufflesync commands to separate file shufflesync/cmds.m. 
@shufflesync/cmds.m: Added code for shufflesync objects from CatCmds.m. 
@shufflesync/private/computeSurrData.m: Added code to compute mean and std 
	of truncated normal distributions using mle. 
@shufflesync/plot.m: Modified to get handle to magenta lines before xlimits 
	are changed as it sometimes does not work otherwise. 
@shufflesync/private/shufflesyncsurr.m: Modified code to add option to 
	obtain the number of frames from the datafile instead of from the input 
	arguments. 

June-06-2006
@variability/cmds.m: Added code to look for duplicate spike times in all 
	surrogate files instead of just the first surrogate file (incomplete). 

June-16-2006
@shufflesync/plot.m: Commented out code to display the magenta lines for 
	the spike rasters as it is causing problems with Matlab 7. 
@adjspikes/plot.m: Added 'ReturnVars' input argument so it will work with 
	the new developmental version of InspectGUI. Added code to return 
	arguments including Args structure in order to work with the new 
	develeopmental version of InspectGUI. 
@frdiff/plot.m: Added input arguments common to nptdata objects that will 
	be taken out by new developmental version of InspectGUI. Added code to 
	return variables, including the Args structure so it will work with the 
	new developmental version of InspectGUI. 

June-20-2006
@ascorr/cmds.m: New class to compute correlation coefficients between a 
	pair of cells for individual repetitions. 
@ascorr/ascorr.m: New class to compute correlation coefficients between a 
	pair of cells for individual repetitions. 
@ascorr/plot.m: New class to compute correlation coefficients between a 
	pair of cells for individual repetitions. 
@ascorr/plus.m: New class to compute correlation coefficients between a 
	pair of cells for individual repetitions. 
@ascorr/subsref.m: New class to compute correlation coefficients between a 
	pair of cells for individual repetitions. 
@shufflesync/plot.m: Added comments to match if/else/end statements to make 
	code more readable. 
@ascorr/get.m: New class to compute correlation coefficients between a pair 
	of cells for individual repetitions. 
@ascorr/subsasgn.m: New class to compute correlation coefficients between a 
	pair of cells for individual repetitions. 
@frdiff/plot.m: Moved code to paint the background of the pairs that belong 
	to the same group the same color into a new function GroupedIntraPlot. 
@ascorr/plus.m: Updated to add correlation coefficients of the entire 
	response.
@latency/latency.m: Added option and code to plot the latency of multiple 
	frames. Need to modify to make it more general. 
@ascorr/plot.m: Added code to plot correlation between entire response. Now 
	uses the new function GroupedIntraPlot to delimit pairs that belong to 
	the same group. 
@ascorr/cmds.m: Now includes code to add PSTH correlation to ascorr/plot.
@ascorr/ascorr.m: Added code to compute corrcoef between the entire 
	responses of a pair of cells. 

July-08-2006
@ascorr/ascorr.m: Now calls histcie with 'DropLast' to avoid introducing a 
	false correlation with the zero at the end of the counts returned by 
	histcie. Added code to compute vector angles for individual repetitions 
	as well. 
@shufflesync/private/shufflesyncsurr.m: Fixed bug that was not setting 
	varargin to the variables startwindow and endwindow before using it. 
@frdiff/frdiff.m: Modified to take the transpose before calling 
	vecsimilarity since the function has been modified to take data in 
	columns rather than rows. 
@frdiff/cmds.m: Added commands to compute paired intra-inter PSTH 
	correlation differences.
@ascorr/cmds.m: Added commands to plot vector angles. 
@sparsity/private/Sparseness.m: Added options to compute sparsity for 
	different sized windows and to ignore the boundaries of the movie frame 
	(i.e. windows are equally spaced). 
@sparsity/plus.m: Removed code to add unnecessary object field dlist. 
	Hopefully won't cause problems with old objects. 
@ascorr/plot.m: Added option to plot vector angles. Modified to plot the 
	correlation and vector angles obtained by treating the responses of all 
	repetitions as one continuous time series only if the ShowCont argument 
	is used. Removed redundant LabelsOff code. 
@poissonctrl/cmds.m: Added code to plot poisson control data on top of data 
	from the correlation and vector angles of individual repetitions. 
@ascorr/plus.m: Added code to add vector angle fields. 
@sparsity/sparsity.m: Added options to compute sparsity for different sized 
	windows and to ignore the boundaries of the movie frame (i.e. windows 
	are equally spaced). Removed unnecessary object fields dlist and 
	setIndex. Now saves Args in the object so we can check if a saved 
	object matches the Args that a user has specified (not implemented 
	yet). 

July-10-2006
@ascorr/cmds.m: Added commands to compute correlation coefficients of movie 
	and m-sequence responses and to compute the paired t-test between them. 
@ascorr/ascorr.m: Added code to compute correlation coefficient using the 
	ispikes object. This way we can compare correlation between movie 
	responses and m-sequence responses. Added WindowSize argument to alter 
	window sizes used in computing histogram of spike counts before taking 
	the correlation. Only the simplest histogram is implemented right now, 
	which involves taking the entire response as one time series. Added 
	option to compute correlation coefficient without first subtracting the 
	mean out of the individual responses (does not appear to make any 
	difference). 

July-14-2006
@ascorr/plot.m: Added option to plot spike times in ispikes object to get 
	around problems with the plot function for the ispikes object. 
@ascorr/cmds.m: Added unix commands for creating list of common cells 
	between movie and msequence sessions. 
@ascorr/ascorr.m: Added code to break up m-sequence responses into pseudo 
	repetitions to compare to movie responses. Currently requires that the 
	user specifies the number of frames presented in the movie session for 
	the same cell, but ideally should be able to get the information 
	automatically (not implemented). Added code to compute a few measures 
	for determining responsiveness of m-sequence responses. Now uses frame 
	information from stiminfo object to set up histogram bins instead of 
	going from 0 to the max spike time, which might leave out a number of 
	frames with zero spikes. 
@sparsity/cmds.m: New file to keep track of commands making use of the 
	sparsity object. 

July-19-2006
@poissonctrl/cmds.m: 
@frdiff/plot.m: Fixed bug in returning output arguments.
@poissonctrl/plot.m: Added code to plot the results of the vector angle, 
	event probabilities, and population sparseness calculations. 
@poissonctrl/plus.m: Updated to add new fields to store median and 
	quartiles instead of mean and std. 
@ascorr/cmds.m: 
@poissonctrl/poissonctrl.m: Added code to compute correlations, vector 
	angles and population sparseness for surrogates with different 
	properties. New properties include surrogates with a specified multiple 
	of the standard deviation (StdMultiple), surrogates with variability 
	equivalent to the variability of all repetitions from all cells added 
	together (RepsVar), surrogates that have the same amount of average 
	variability in spike count for the entire repetition (CellsVar), 
	surrogates where the variability in each frame is computed from the 
	average variability of each of the cells (FrameVar), and surrogates 
	where the max variability was taken instead of the average (FanoMax). 
	Also replaced mean and std fields with median and quartiles since 
	distributions are not Gaussian.
@frdiff/cmds.m: 

October-07-2006
@responsiveness/plus.m: Now combines the new fields added to save the 
	results of testing distributions of correlation coefficients for data 
	and control.
@responsiveness/private/CalcResponsiveness.m: Now no longer takes stiminfo 
	object as input since we were only using the stiminfo object to get the 
	number of frames and repetitions. These numbers can be obtained from 
	the adjspikes object directly. Now uses a much more efficient method 
	to compute the responsiveness measure. Added code to create control 
	data and test for differences in the distributions of correlation 
	coefficients computed from the data and from the control data. By 
	default, the control data is created by shuffling spike counts between 
	repetitions but the shuffling can be expanded to include shuffling 
	between time bins if the 'RandTimeTrial' option is specified when 
	calling the responsiveness constructor. Added code to plot adjspikes 
	object together with the distributions of correlation coefficients 
	computed from the data and the control. 
@responsiveness/responsiveness.m: Removed 'stimInfoDir' argument since the 
	stiminfo object is no longer needed. The stiminfo object was used to 
	get the number of frames and the number of repetitions and these can be 
	obtained directly from the adjspikes object. Removed the 'Group' 
	argument since it is not being used. Added 'PlotDist' to plot 
	distributions of correlation coefficients for data and control while 
	creating the object. This is useful since saving the coefficients will 
	result in very large objects. Added 'RandTimeTrial' option to create 
	control data by shuffling spike counts over all time bins and 
	repetitions. This is an alternative to the control data generated by 
	default which only shuffles spike counts over repetitions without 
	shuffling time bins. Added fields to save results of testing 
	distributions of correlation coefficients for data and control. 
	Removed unused fields 'dlist' and 'setindex' from the object. 
@shufflesync/private/computeSurrData.m: Temporarily commented out code to 
	compute mean and standard deviations of truncated normal distribution 
	so we can recompile this code using the compiler. 
@sparsity/private/Sparseness.m: Added code to convert spike counts to 
	firing rates. 
@responsiveness/private/CalcResponsiveness.m: Now turns off divideByZero 
	warning before calling corrcoef function. Now generates a large number 
	of random surrogates to ensure a stable distribution of correlation 
	coefficients will be used to compare to the distribution from the data. 
@responsiveness/responsiveness.m: Added 'RandSets' argument to compute 
	distribution of correlation coefficients of a large number of surrogate 
	sets. 
@frdiff/frdiff.m: Added option to compute noise counts directly from the 
	adjspikes object instead of using the firingrate object. 
@poissonctrl/poissonctrl.m: Added 'SumVar' argument to generate surrogates 
	where the variance is equal to the sum of the variances of the 
	individual cells. Added 'StdMinFR' argument exclude bins below a 
	certain firing rate. Added 'CellPopVar' argument to compare variance 
	between repetitions from an individual cell versus repetitions between 
	cells. 
@poissonctrl/plot.m: Added option to plot mean and standard deviation of 
	population sparseness metric for poisson control data. 
@sparsity/sparsity.m: Added 'Rate' argument to convert spike counts to 
	firing rates. 

October-10-2006
@psparse/plot.m: Added 'BoxPlotColors' and 'BoxPlotSymbol' arguments to 
	over-ride the default colors. 
@frdiff/plot.m: Changed heterogeneity plot so we can plot more points. This 
	was necessary since we are now using 88 single units in all our 
	analyses. 

October-11-2006
@psparse/plot.m: Added argument to modify marker size in boxplot. 

November-06-2006
@poissonctrl/poissonctrl.m: Added code to turn off warnings about 
	divideByZero. Moved initialization code so that all fields are 
	initialized for all options. Improved code to extract unique 
	correlation values. 
@firingrate/cmds.m: New function used to store commands using objects of 
	this class.
@responsiveness/cmds.m: New function used to store commands using objects 
	of this class.

November-28-2006
@refractory/plot.m: Commented out code that plots the fitted exponential 
	since we are no longer doing that in the spikeRecovery function. 
spikeRecovery.m: Modified to skip fitting isi histogram to an exponential. 
@refractory/refractory.m: Updated to skip saving the fitted exponential. 

January-15-2007
spikeRecovery.m: Commented out computation of tspikes since it is no 
	longer used. 

January-24-2007
@variability/variability.m: Extensively modified by Xiao Xiao to compute 
	variability in windows that are multiples of the frame duration. 
@latency/plot.m: Added option and code to return variables so we can get 
	the indices of cells that have a refresh artifact. Fixed bar command so 
	it is compatible with Matlab 7. Changed default option to plot just the 
	latency data without replication. Added 'Replicate' option to repeat 
	the plot so frame following is more apparent. Added option to use 
	indices in subplot titles. 
@variability/plot.m: Modified by Xiao Xiao to work with new InspectGUI 
	function. 

January-25-2007
@adjspikes/private/AdjustFramesVector.m: Modified to ignore the 
	stimInfo.data.iniInfo.obj_type field so that the spikes are adjusted 
	properly for sessions where the obj_type field is irrelevant (the type 
	field contains the actual stimulus type presented). 

February-13-2007
@latency/plot.m: Now changes the EdgeColor as well as the FaceColor when 
	changing the color of bar plots. 
@variability/cmds.m: 
@variability/plot.m: Updated help comments to include new plot options. 
	Now changes the color of bar plots using a separate set command instead 
	of including it in the bar command which does not work in Matlab 7. 
getFreeFiringRate2.m: Added option and code to average free firing rate 
	over smaller window sizes. 
@variability/variability.m: Now loads trentropy data file into a variable 
	called stre so that the saved variable "entropy" does not clash with 
	the Matlab built-in function entropy. 


April-23-2007
@adjspikes/plot.m: Added title to plots with the 'Fast' argument so we can 
	see which cell is being plotted.
@poissonctrl/poissonctrl.m: Commented out redundant code that saved 
	timebins but was not using it. 
@poissonctrl/cmds.m: Fixed minor typo in comments. 
@repcorr/plot.m: New class to compute correlation coefficients between 
	individual repetitions instead of using the PSTH. Could be useful in 
	comparing movie responses to grating responses. 
@repcorr/subsasgn.m: New class to compute correlation coefficients between 
	individual repetitions instead of using the PSTH. Could be useful in 
	comparing movie responses to grating responses. 
@repcorr/get.m: New class to compute correlation coefficients between 
	individual repetitions instead of using the PSTH. Could be useful in 
	comparing movie responses to grating responses. 
ShihChengChangeLog.txt: 
@repcorr/plus.m: New class to compute correlation coefficients between 
	individual repetitions instead of using the PSTH. Could be useful in 
	comparing movie responses to grating responses. 
@repcorr/repcorr.m: New class to compute correlation coefficients between 
	individual repetitions instead of using the PSTH. Could be useful in 
	comparing movie responses to grating responses. 
@repcorr/subsref.m: New class to compute correlation coefficients between 
	individual repetitions instead of using the PSTH. Could be useful in 
	comparing movie responses to grating responses.