Constructing and Modifying Sequence Statistics for relevent Using informR in R

Abstract

The informR package greatly simplifies the analysis of complex event histories in R by providing user friendly tools to build sufficient statistics for the relevent package. Historically, building sufficient statistics to model event sequences (of the form a / b) using the egocentric generalization of Butts’ (2008) relational event framework for modeling social action has been cumbersome. The informR package simplifies the construction of the complex list of arrays needed by the rem() model fitting for a variety of cases involving egocentric event data, multiple event types, and/or support constraints. This paper introduces these tools using examples from real data extracted from the American Time Use Survey.