Current R packages and team initiatives associated to the project include:

  • missing values: missMDA for principal component methods with missing values and imputation of mixed data; Rmistatic a platform to provide references, implementations, workflows of analyses.
  • multi-blocks methods: FactoMineR to visualize groups of variables/observations with dimensionality reduction. 112 000 download/month; sgPLS for Sparse Group PLS.
  • causal inference: a task view with implementations and their functionalities
  • functional data: curveDepth for reading curves (brain fibers, the ex involves 1000 3D coord.),  calculating distance between curves and for computing Tukey curve depth.
  • measures of (non-linear) dependence: IndependenceTests, HellCor.


Comments are closed.