[R-sig-dyn-mod] question about the limit of conversion rate

[Thomas Petzoldt]

Hi,

your idea for a smoothed impulse function seems quite close to a good 
solution, what about:

x <- seq(-20, 20, 1e-3)
steep <- 20
plot(x, (tanh(steep * (x %% 10 - 8))
        - tanh(steep * (x %% 10 - 2)))/2 + 1,
      type = "l", col = "red")



thpe