During the model building process, simulations can be performed on the any model at any time. Simulations are then used to compare model behavior to the data under constant and noisy conditions. Starting with an initial population density N(0), standard deviation s, and number of iterations Y (input from the keyboard), the value of R is calculated from the current R-function and is then used to obtain the value of N(1) from the step-ahead forecasting equation. This value is then used as the input to the next iteration. Simulation can continue for up to 100 iterations.
The simulated trajectory is first displayed as a time series plot, and then as a phase portrait. The original data are superimposed on the phase portrait so that the user can compare the simulated and natural dynamics. Trajectories that fall in the same area of phase space, and that display the same patterns as the data may be considered as valid trajectories. Changing the standard deviation can be used to obtain better congruence between simulated and real trajectories. In other words, the model can be "tuned" by modifying the degree of random variability, but cannot be tuned by changing its biological parameters.