Fig. 5From: Practical identifiability in the frame of nonlinear mixed effects models: the example of the in vitro erythropoiesisThe estimated parameter values of Model (8) are identifiable. Displayed as dots are the estimated population parameter values in the 25 convergent runs of SAEM for Model (8), color-coded by log-likelihood. Error bars are the FIM-derived standard errors of the estimates when available (when there is no error bar for a parameter value in a given run, it means that it was impossible for Monolix to invert the FIM). The population means and standard deviations are expressed in \(d^{-1}\). The error parameters are dimensionlessBack to article page