Current speciation models in Kaphi assume constant rates of speciation (birth) and extinction (death). It would be great if we can implement rate variation over time. This is implemented in TreePar with epochs (time intervals where global rates switch to interval-specific values). I'm interested in implementing uncorrelated rates (sample rates from parametric distributions) or autocorrelated rates (sample rate change from distribution), where these events may occur at nodes or according to another Poisson process over the tree. We can modify existing code, e.g., diversitree trees function.
Actually this is already implemented in a way with the BiSSE and MuSSE models in diversitree, although those follow a general discrete distribution (specified rate classes).
Current speciation models in Kaphi assume constant rates of speciation (birth) and extinction (death). It would be great if we can implement rate variation over time. This is implemented in TreePar with epochs (time intervals where global rates switch to interval-specific values). I'm interested in implementing uncorrelated rates (sample rates from parametric distributions) or autocorrelated rates (sample rate change from distribution), where these events may occur at nodes or according to another Poisson process over the tree. We can modify existing code, e.g., diversitree
trees
function.