test.asv

clear

dt = 0.1e-3;
T = 2.2;
Ntime = T/dt;


[A,c,x] = getInput(dt.*(1:Ntime),0.0);
J = size(A,1);
Nneuron = 400;

lambda = 10;
F = randn(J,Nneuron);
vn = vecnorm(F,2);
F = 0.03*F./vn;
C = -F'*F;
Thresh = 5.5e-4;

C = -F'*(A+lambda*eye(J))*F;

[rO, O, V] = runnet(dt, lambda, F ,c, C,Nneuron,Ntime, Thresh);
plot(dt*(1:Ntime), F*rO)






rO=zeros(Nneuron,Ntime);%filtered spike trains
O=zeros(Nneuron,Ntime); %spike trains array
V=zeros(Nneuron,Ntime); %amebrane poterial array

for t=2:Ntime

    Input = c(:,t-1) + (A+eye(J)*lambda)*x(:,t-1);
    V(:,t)=(1-lambda*dt)*V(:,t-1)+dt*F'*Input+C*O(:,t-1)+0.00*randn(Nneuron,1);%the membrane potential is a leaky integration of the feedforward input and the spikes

 
    [m,k]= max(V(:,t) - Thresh-0.0*randn(Nneuron,1));%finding the neuron with largest membrane potential
        
    if (m>=0)  %if its membrane potential exceeds the threshold the neuron k spikes  
        O(k,t)=1; % the spike ariable is turned to one
    end

    rO(:,t)=(1-lambda*dt)*rO(:,t-1)+1*O(:,t); %filtering the spikes
    x(:,t) = F*rO(:,t);
end

options = odeset("MaxStep",1e-4);
[tt,y] = ode45(@(t,x) A*x + c(:,floor(t/dt)+1),[0 T-dt],x(:,1),options);

figure
clf
h = 1;
plot(dt.*(1:Ntime),x(h,1)+x(h,:));
hold on
plot(tt,y(:,h))