Bigsimr.jl Package

The Bigsimr package provides a helpful set of tools for simulating high-dimensional multivariate data with arbitrary marginal distributions. Particularly, Bigsimr implements:

  • Simulation of multivariate data via Gaussian copulas (NORTA algorithm)
  • Converting between different types of correlations (Pearson, Spearman, and Kendall)
  • Computing the nearest positive definite correlation matrix (quadratically convergent algorithm)
  • Pearson correlation matching to account for non-linear transformations
  • Generating random positive semi-definite correlation matrices