Rumana Lakdawala - Joris Mulder - Roger Leenders

Simulating relational event history data for model checking & theory building

My first PhD project involved developing a flexible and fast data generating package ‘remulate’, for the tie-oriented Relational Event model (REM). I further extended the package to include extensions of REM including the actor oriented DyNAM model of Stadtfeld (2017) and allowing memory effects to persist over time.

Software package


I am the developer of remulate, an open source data generation R package. Remulate provides a set of functions that generate relational event data under the tie-oriented (Relational Event Model) and actor-oriented (DyNAM) models. Several statistics commonly used in literature have been implemented within remulate, to allow users to generate data from complex models. Relational event data can also be generated with various memory effects such as exponential decay or a memory window.


Furthermore I am also a contributor of remstimate, an R package that provides a set of functions for modeling both tie-oriented (Relational Event Model) and actor-oriented (DyNAM) models. Several optimization methods (frequentist or Bayesian) have been implemented in order to optimize either the log-likelihood or the posterior density of the specified model. In the current version of the package the estimation methods available are: Maximum Likelihood, Gradient Descent, Adaptive Gradient Descent (ADAM), Bayesian Sampling-Importance- Resampling and Hamiltonian Monte Carlo. I contribute to the package along with Giuseppe, Fabio, and Diana.