finderror.m

function [errorval, val] = finderror(x,type,mit,mxt,wrapped,burst)

% This function will find the error values for spike data
% generated by lsd (the 'RAW' data format)
%
% errorval=finderror(x,type,mit,mxt,wrapped,burst)
%
% x = The data in 'raw' format (from lsd).
% type = The error type.
% mit,mxt = The times within which to find the error in milliseconds.
% wrapped = Whether the data is wrapped or not.
% burst = whether to look through all spikes or just burst spikes

mit=mit*10;   %need to define the limits
mxt=mxt*10-1;

if wrapped == 1 %wrapped is ON
	l = 1;
	for j = 1:x.numtrials
		for k = 1:x.nummods
			if burst == 0
				s = x.trial(j).mod{k} - x.trial(j).modtimes(k);   %because it is wrapped
			else
				s = x.btrial(j).mod{k} - x.trial(j).modtimes(k);   %because it is wrapped
			end
			m = find( s > mit & s <= mxt );
			val(l) = length(m);
			l = l+1;
		end
	end
else %wrapped is OFF
	l = 1;
	a = [];
	for j = 1:x.numtrials
		for k = 1:x.nummods
			if burst == 0
				s=x.trial(j).mod{k}-x.trial(j).modtimes(1);   %not wrapped
			else
				s=x.btrial(j).mod{k}-x.trial(j).modtimes(1);   %not wrapped
			end
			m=find(s>mit & s<=mxt);
			a=[a;length(m)];
		end
		a=sum(a);
		val(l)=a;
		a=[];
		l=l+1;
	end
end

errorval=errorfun(val,type);
val = val';

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
function err = errorfun(edata,type)
% Computes the Error Data

switch(type)
	
	case {'Standard Error', 'SE'}
		err=std(edata)';
		err=sqrt((err.^2/length(edata)));
	case {'Standard Deviation', 'SD'}
		err=std(edata)';
	case {'2 StdDevs', '2SD'}
		err=std(edata)'*2;
	case '3 StdDevs'
		err=std(edata)'*3;
	case {'2 StdErrs', '2SE'}
		err=std(edata)';
		err=sqrt((err.^2/length(edata)))*2;
	case 'Variance'
		err=std(edata)'.^2;
	case 'Fano Factor'
		if max(edata)==0
			err=0;
		else
			err=var(edata)/mean(edata);
		end
	case 'Coefficient of Variation'
		if max(edata)==0
			err=0;
		else
			err=std(edata)/mean(edata);
		end
	case 'Allan Factor'
		if max(edata)==0
			err=0;
		else
			err=var(diff(edata))/(2*mean(edata));
		end
end