Causality research based on phase space reconstruction.

Based on phase space reconstruction theory, the root mean square error is used as a quantitative criterion for identifying the appropriate embedding dimension and time step and selecting the optimal configuration for these factors. The phase space is then reconstructed, and the convergent cross-mapping algorithm is applied to analyse the causality between time series. The causality among the variables in the Lorenz equation is first discussed, and the response of this causality to the integration step of numerical solutions to the Lorenz equation is analyzed.