SEset (R package)

Tools to compute and analyze the set of statistically-equivalent (Gaussian, linear) path models which generate the same precision or (partial) correlation matrix.

This is an R package I developed as part of a paper on how statistical network models can be understood as causal discovery tools, but can be used more generally to understand the variety of different path models which imply the same set of statistical dependencies, expressed as either a partial or full correlation matrix.

The statistical-equivalence set of a given GGM expresses the uncertainty we have about the sign, size and direction of directed relationships based on the weights matrix of the GGM alone.

More details can be found in: Ryan, O., Bringmann, L.F., & Schuurman, N.K. (2022) The Challenge of Generating Causal Hypotheses using Network Models. PsyArxiv Pre-print.

This github page for this package can be found here

The CRAN page can be found here