5 projects
spacebench
Spatial confounding poses a significant challenge in scientific studies where unobserved spatial variables influence both treatment and outcome, leading to spurious associations. SpaCE provides realistic benchmark datasets and tools for systematically valuating causal inference methods for spatial confounding. Each dataset includes training data with spatial confounding, true counterfactuals, a spatial graph with coordinates, and realistic semi-synthetic outcomes.
pycre
Python implementation of Causal Rule Ensemble
pycausalgps
Matching on generalized propensity scores with continuous exposures
nsaphx
nsaphx is a Python package for causal inference studies using the potential outcome framework. It offers a flexible and extensible framework to apply computational instructions to input data, including exposure, outcome, and confounders. The package uses directed acyclic graphs and database storage for efficient computation and storage.
tsprocess
Ground motion time series processing tools
