TreeStab {moreparty}

Stability assessment of conditional inference trees

Description

Assesses the stability of conditional inference trees through the partition of observations in the terminal nodes and the frequency of the variables used for splits.

Usage

TreeStab(ct, B = 20)

Arguments

ct

A tree of class constparty (as returned by ctree from partykit package).

B

Numerical value. The number of bootstrap replications. Default is 20.

Details

The study of splitting variables used in the original tree and in bootstrap trees in directly inspired from the approach implemented in stablelearner package. The other side of this functions also uses bootstrap trees, this time to compute the Jaccard index of concordance between partitions, to assess the stability of the partition of observations in the terminal nodes of the tree.

Value

A list of two elements :

partition

average Jaccard index of concordance between the partition (terminal nodes) of ct and the partitions of bootstrap trees

variables

a data frame with splitting variables in rows and two statistics in columns : their frequency of use in the tree vs in the bootstrap trees, and

Author(s)

Nicolas Robette

References

Hothorn T, Hornik K, Van De Wiel MA, Zeileis A. "A lego system for conditional inference". The American Statistician. 60:257–263, 2006.

Hothorn T, Hornik K, Zeileis A. "Unbiased Recursive Partitioning: A Conditional Inference Framework". Journal of Computational and Graphical Statistics, 15(3):651-674, 2006.

Philipp M, Zeileis A, Strobl C (2016). "A Toolkit for Stability Assessment of Tree-Based Learners". In A. Colubi, A. Blanco, and C. Gatu (Eds.), Proceedings of COMPSTAT 2016 - 22nd International Conference on Computational Statistics (pp. 315-325). The International Statistical Institute/International Association for Statistical Computing. Preprint available at https://EconPapers.RePEc.org/RePEc:inn:wpaper:2016-11

See Also

ctree

Examples

  data(iris)
  iris2 = iris
  iris2$Species = factor(iris$Species == "versicolor")
  iris.ct = partykit::ctree(Species ~ ., data = iris2)
  TreeStab(iris.ct, B = 10)