Enforcing model properties and behavior. (A–C) The convergence of the N_C estimate (A) and calculated P_O (B) and f_R (C) quantities are shown for each of the six optimization runs described in Fig. 2 B. The constrained and the true values are indicated by the gray and black dashed lines, respectively. As expected, N_C and P_O are inversely correlated; when one is constrained to be larger or smaller than the true value, the other one becomes smaller or larger, respectively. Likewise, f_R is correlated with the rate of recovery from inactivation; when f_R is constrained to be larger than the true value, the $k 4,3 0$ and $k 4,3 1$ estimates become smaller and more negative, respectively. All the enforced quantities quickly reach their enforced range (N_C) or value (P_O and f_R). Note that without being constrained, f_R is not well defined by the test data and does not converge to the true value. In contrast, N_C and P_O are well defined.