// Automatically display figure 2B and provide a
// Grapher for generating curves for figure 2A.
// This simulation seems to produce different quantitative (but same
// qualitative) results
// from those shown in Fig 2B with regard to height at x=1000 and relative minima
// of unstimulated axon (x=~1100).
// The simulation output has been verified to be a linear superposition
// of two exponentials with lambda1 and lambda2 given in the paper.
// The simulation output for N=2 is best fit by
// stimulated a1*exp(-($1-1000)/l1)+ a2*exp(-($1-1000)/l2)
// unstimulated a1*exp(-($1-1000)/l1) - a2*exp(-($1-1000)/l2)
// "a1", 0.818485
// "a2", 0.181507
// "l1", 28.359
// "l2", 129.129
This is consistent with a private communication from
Hemant Bokil <firstname.lastname@example.org>
I checked my notes for the anaytical calculations and the complete
V_A(x)=0.5*(1-f)*sqrt(r_i * r_m)*exp(-x/l2)+0.5*f*sqrt(r_i*r_m)
where l1 and l2 are as listed in the paper. For plotting 2b, i divided
both these expressions by V_A(0) and this leads to amplitudes of 0.8207
and -0.1791, for V_B(x) (with f=0.5; 2 neurons), very close to your
answers (perhaps a rounding error ?).