Probabilistic reconciliation of hierarchical time series forecasts via conditioning. Implements analytical Gaussian reconciliation, Bottom-Up Importance Sampling (BUIS), MCMC reconciliation of count series, and mixed-type hierarchies.

Computationally efficient estimation of orthant probabilities of high-dimensional Gaussian vectors. Also implements conservative excursion set estimation under Gaussian random field priors.

Computes pseudo-realizations from the posterior distribution of a Gaussian Process for excursion set estimation, using a simulation strategy that minimizes expected distance in measure.

Computes and visualizes profile extrema functions for arbitrary functions, with GP emulation and uncertainty quantification support for expensive-to-evaluate functions.

Criteria and algorithms for sequentially estimating level sets of a multivariate numerical function, for deterministic and noisy computer experiments (Kriging-based inversion).

Python port of bayesRecon. Probabilistic reconciliation of hierarchical forecasts, published in the Journal of Open Source Software (2025).

Gaussian process based library for learning from preference and choice data. Implemented in Jax and PyTorch

Sparse Information Filter for fast GP regression. Scales to millions of data points via mini-batch SGD; approximately 4× faster than SVGP at comparable accuracy.

Closed-form posterior inference for GP classification, preference learning, ordinal regression, and mixed problems using Skew Gaussian Processes.

Stochastic Recursive GP regression using a Kalman filter formulation, estimating model parameters by recursively propagating gradients over mini-batches.