computeCSI.m

load('SimulationResults/net1/TYLT_astronomy_net1_L.mat')
df1i=find(Oddball==12); %list of indices of the deviant
sf1i=find(Oddball==10);  %list of indices of the standard
%DA1=E_mean(Rec_Column,:); %all neurons of one column
%DA1=E_mean(Rec_Column-1,:);
DA1=E_mean(Rec_Column+1,:);

load('SimulationResults/net1/TYLT_astronomy_net1_H.mat')
df2i=find(Oddball==10);
sf2i=find(Oddball==12);
%DA2=E_mean(Rec_Column,:);
%DA2=E_mean(Rec_Column-1,:);
DA2=E_mean(Rec_Column+1,:);

%compute the peak responses of all 100 stim
maxE=zeros(1,n_stim);
for i=1:n_stim
m=max(DA1(Stim_Onsets(i):Stim_Onsets(i)+400)); %400 is duration of response
maxE(i)=m;  
%min seems to be at about 2.9956, subtract this from max?
end
d1=sum(maxE(df1i))/length(df1i); %filter out the deviants and compute mean
s1=sum(maxE(sf1i))/length(sf1i); %filter out the standards and compute mean

maxE=zeros(1,n_stim);
for i=1:n_stim
m=max(DA2(Stim_Onsets(i):Stim_Onsets(i)+400)); %400 is duration of response
maxE(i)=m;  
%min seems to be at about 2.9956, subtract this from max?
end
d2=sum(maxE(df2i))/length(df2i);
s2=sum(maxE(sf2i))/length(sf2i);

CSI=(d1+d2-s1-s2)/(d1+d2+s1+s2)

%CSI given in the text=0.632
%col=11 gives 0.7747
%col-1 gives 0.5526
%col+1 gives 0.6152