Open Source

Software is an integral part of my research. Releasing my methods as code makes them transparent and reproducible, and lets applied researchers and policy analysts use them without reimplementing everything from scratch. Below are the tools I have built or contributed to, most of them Julia packages for econometrics and time series. A few older ones are archived: they generally still work, but I no longer actively maintain them. I also keep my teaching material open: notes and code from courses I’ve taught, listed after the software.

Software

TransmissionChannelAnalysis.jlJulia package for transmission channel analysis (TCA).

TCA MATLAB ToolboxMATLAB implementation of transmission channel analysis.

HamiltonFilters.jlLightweight Hamilton filter implementation for time series.

FredMDQD.jlSimplifies working with FRED-MD / QD data in Julia.

bigtimecontributorStructured lasso-penalized estimation for large VAR / VARX / VARMA models.

Archived

BayesFlux.jlBayesian neural network extension to Flux.jl (from master’s thesis).

BayesFluxRR wrapper for BayesFlux.jl.

TinyGibbs.jlSmall Gibbs sampler using the AbstractMCMC interface.

MacroEconometrics.jlEarly collection of macroeconometric utility functions.

Teaching Material

Probability Theory (EBC1024)Introductory R programming notes for the undergraduate probability course at Maastricht University.

Introduction to VAR ModelsAn interactive app for building intuition about vector autoregressive (VAR) models through experimentation.

High-Dimensional VARs with bigtimeA hands-on notebook on estimating sparse, high-dimensional VARs with the bigtime R package.

Shell, Git & GitHubExercise-driven materials for the command line, version control with Git, and collaboration on GitHub.