收敛交叉映射（convergent cross mapping，CCM）滥觞、2012年Science论文单向因果：Detecting Causality in Complex Ecosystems

Framework for identifying causation, case (ii) Unidirectional causality. Here, species X influences the dynamics of Y, but Y has no effect on X (Fig. 3C and fig. S1B). This describes an amensal or commensal relationship, or where X represents external environmental forcing. 确定因果关系的框架，案例（ii）单向因果关系。在这里，物种X影响Y的动力学，但Y对X没有影响（图3C和图S1B）。这描述了修正关系或共生关系，或其中X表示外部环境强迫。

Figure 3C examines the system when ${\beta }_{x,y}=0$. Notice that with moderately strong forcing from X (via ${\beta }_{y,x}$), even though Y exerts no effect, there may still be partial cross mapping of Y arising from the contemporaneous dependence of Y on X. However, this statistical effect is not convergent (shown by the asymptotic level curves with respect to L in Fig. 3E). With extremely strong forcing, the intrinsic dynamics of the forced variable become subordinate to the forcing variable, leading to the well-studied phenomenon of “synchrony” (27). The red plateau in Fig. 3E shows that bidirectional convergence can occur with strong forcing. Thus, strong forcing (synchrony) must be ruled out for CCM to unequivocally imply bidirectional coupling, although it still implies membership in a common dynamic system. 图3C测试了${\beta }_{x,y}=0$时的系统。请注意，对于来自X的中等强度强迫（通过${\beta }_{y,x}$），即使y没有施加任何影响，也可能存在y对X的同时依赖性引起的y的部分交叉映射。然而，这种统计效应并不收敛（如图3E中关于L的渐近水平曲线所示）。在极强的强迫作用下，受迫变量的内在动力学从属于受迫变量，导致“同步”现象得到充分研究（27）。图3E中的红色平台表明，在强强迫下，可以发生双向收敛。因此，必须排除强强迫（同步），CCM才能明确表示双向耦合，尽管它仍然意味着属于一个共同的动力系统。