Figure 3.

Population analysis view with N = 10, R = 2. (A) Graphic view of the potential's coefficients values; the abscissa corresponds to an index associated to a prescribed order on monomials: the N first coefficients corresponds to single events (hence are related to firing rate); by clicking on the histogram bar one makes appear the coefficient value and the corresponding monomial. Ordinate give the value of the coefficient attached to the corresponding monomial. (B) Literal form of the potential. (C) Theoretical average value of monomials vs. their empirical average value. To which extents the model predicts the empirical data is drawn with this representation. This allows checking that constraints on empirical averages are respected by the model. (D) Decay of estimation errors vs. the number of iterations allowing us to graphically check the convergence of the estimation algorithm (red curve). In the figure several successive runs have been performed. The blue curve shows the evolution of error after each of these trials. (E) Confidence plot. C_p, C_m corresponds, respectively, to the upper and lower 3σ bound associated with Central Limit Theorem (see text).