Associations.jl is a package for quantifying associations, independence testing and causal inference.
All further information is provided in the
documentation, which you can either
find online or build locally by running the docs/make.jl
file.
- Association API: includes measures and their estimators for pairwise, conditional and other forms of association from conventional statistics, from dynamical systems theory, and from information theory: partial correlation, distance correlation, (conditional) mutual information, transfer entropy, convergent cross mapping and a lot more!
- Independence testing API, which is automatically compatible with every association measure estimator implemented in the package.
- Causal (network) inference API integrating the association measures and independence testing framework.
Extending on features from ComplexityMeasures.jl, we also offer
- Discretization API for multiple (multivariate) input datasets.
- Multivariate counting and probability estimation API.
- Multivariate information measure API
To install the package, run import Pkg; Pkg.add("Associations")
.
Previously, this package was called CausalityTools.jl.