Python framework for inference in Hawkes processes. PyHawkes implements a variety of Bayesian inference algorithms for discovering latent network structure given point process observations. Suppose you observe timestamps of Twitter messages, but you don’t get to see how those users are connected to one another. You might infer that there is an unobserved connection from one user to another if the first user’s activity tends to precede the second user’s. This intuition is formalized by combining excitatory point processes (aka Hawkes processes) with random network models and performing Bayesian inference to discover the latent network.