I’ve long been interested in how we can understand complex systems without dissecting out the interdependence that causes them to come to life. I deeply admire and value the control and certainty that a well-constructed experiment can provide, but from the data analysis side, I think there’s much more we can do to gain a preliminary understanding of potentially interesting causal relationships.
As a step in that direction, I’ve created CauseMap, an implementation of convergent cross mapping, a method developed by the Sugihara group at UCSD. The method is fairly different from what most have been exposed to, so I’ve tried to consolidate some useful explanatory resources on the project website. Check it out at cyrusmaher.github.io/CauseMap.jl!