# This is an XPP realization of the noisy ring model that appears in # Goldberg JA, Rokni U and Sompolinsky H. Patterns of Ongoing Activity and the # Functional Architecture of the Primary Visual Cortex. Neuron, 42:489-500 (2004). # must have the file CosIntCol20.tab in the same directory as NoisyRing.ode # In order to see the population activity do # Viewaxes, Array and then fill out # *Column 1:m11 # NCols:20 # Row 1:0 # NRows:4000 # RowSkip:5 # Zmin:0 # and adjust Zmax according to your simulation # The phase diagram of this model is shown in fig. 3A of the above paper. In order to # get a noisy "hill" of activity lambda must be larger than 1 and mu must be positive. # parameters (gain and mean of Gaussian noise) p lambda=1.8, mu=1 # mu here is equivalent to T/Sigma_n in the figure in the paper. # note lambda=1.1, mu=5 are other possible defaults # threshold linear gain function sl(x)=max(x,0) # stochastic integration, w is local uncorrelated Gaussian white noise weiner w[11..30] #the model m[11..30]'=-m[j]+sl(2*lambda*h([j-11])+w[j]+mu) #Coupling Matrix table coscol CosIntCol20.tab # this table is cosine coupling divided by 20 # recurrent feedback is given by h special h=mmult(20,20,coscol,m11); # order parameters of model rnull=sum(0,19) of (shift(m11,i'))/20 rfundc=sum(0,19) of (shift(m11,i')*cos(2*pi*i'/20))/20 rfunds=sum(0,19) of (shift(m11,i')*sin(2*pi*i'/20))/20 aux r0=rnull aux r1=sqrt(rfundc^2+rfunds^2) aux phiang=atan(rfunds/rfundc) # r1 is order 1 for a "hill of activity" and is order 1/sqrt(20) otherwise @ total=1000,xlo=0,xhi=1000,ylo=0,yhi=10,dt=0.05,bounds=10000,meth=euler d