Fig. 1

Using sequential Monte Carlo to examine system stability. a, b The algorithm takes as input a model and a desired stability structure in phase space. c The system is evolved to achieve the stability of choice via intermediate populations. In this example model, there are two species and two parameters. For the model to be considered bistable, the phase plot of the two species of interest must have two distinct clusters, as shown in (c). d The parameter posterior distribution for the model to achieve the desired behaviour are given as an output. e Flow chart of the Python package, StabilityFinder, which implements the algorithm