Open Source Software
Open-source software is central to how I work, as a maintainer of Python packages in Bayesian methodology and climate data pipelines, and as a contributor to established R packages in the Stan ecosystem. For the tooling behind this work, see the Stack page.
Python package for learning prior distributions of model parameters from expert knowledge.
Python package for harmonizing two time series using zero- and first-order gradient information.
Pipeline for constructing a composite greenhouse gas dataset from ground-based and satellite observations, suitable as forcing input for climate models in CMIP.
Efficient leave-one-out cross-validation and WAIC for Bayesian models.
Plotting functions for posterior distributions, prior/posterior checks, and MCMC diagnostics.
Tools for manipulating and summarizing posterior and prior distributions.
High-level R interface for fitting Bayesian regression models via Stan.
Reproducible manuscript workflow. JupyText notebooks compiled to publication-ready LaTeX PDFs.
Controlled vocabularies for the input4MIPs project, defining the allowed metadata terms across datasets.
Generation of greenhouse gas concentration inputs (forcings) for CMIP7's ScenarioMIP.