Characterizing nonlinear dynamics from noisy trajectories

This dataset was created to train and test automated systems to determine the qualitative behavior (mono- or multi-stable, oscillatory, excitable...) of planar dynamical systems from ensemble of trajectories. We provide pre-simulated trajectories for: an ensemble of dynamical systems, with fixed point structures provided for an ensemble of dynamical systems generated from random cubic polynomial flows; A simple model of non-reciprocal dynamics for various paramter values A 6-dimensional model of the cell cycle developed by Tyson, PNAS 1993. See readme.txt for details on each. Code to generate the trajectories is available at https://github.com/NicoRomeo/DynCarto/. These datasets were generated for the paper "Characterizing nonlinear dynamics by contrastive cartography", by N Romeo, C Chi, AR Dinner, ER Jerison, PRX Life 2025. Please cite the original paper if this dataset is used for research. We acknowledge support from the Burroughs Wellcome Foundation through a Career Award at the Scientific Interface, the Clare Boothe Luce Foundation; the University of Chicago Biological Physics and Center for Living Systems Fellowships, and the National Science Foundation through the Physics Frontier Center for Living Systems (PHY-2317138). Computing resources were provided by the University of Chicago Research Computing Center.