stabm

stabm provides an implementation of many measures which assess the stability of feature selection. The following stability measures are currently included:

stabm::listStabilityMeasures()
#>                           Name Corrected Adjusted Minimum Maximum
#> 1               stabilityDavis     FALSE    FALSE       0       1
#> 2                stabilityDice     FALSE    FALSE       0       1
#> 3             stabilityHamming     FALSE    FALSE       0       1
#> 4   stabilityIntersectionCount      TRUE     TRUE    <NA>       1
#> 5  stabilityIntersectionGreedy      TRUE     TRUE    <NA>       1
#> 6     stabilityIntersectionMBM      TRUE     TRUE    <NA>       1
#> 7    stabilityIntersectionMean      TRUE     TRUE    <NA>       1
#> 8             stabilityJaccard     FALSE    FALSE       0       1
#> 9               stabilityKappa      TRUE    FALSE      -1       1
#> 10         stabilityLustgarten      TRUE    FALSE      -1       1
#> 11           stabilityNogueira      TRUE    FALSE      -1       1
#> 12         stabilityNovovicova     FALSE    FALSE       0       1
#> 13             stabilityOchiai     FALSE    FALSE       0       1
#> 14                stabilityPhi      TRUE    FALSE      -1       1
#> 15              stabilitySomol      TRUE    FALSE       0       1
#> 16         stabilityUnadjusted      TRUE    FALSE      -1       1
#> 17               stabilityWald      TRUE    FALSE     1-p       1
#> 18                 stabilityYu      TRUE     TRUE    <NA>       1
#> 19           stabilityZucknick     FALSE     TRUE       0       1

Installation

You can install the released version of stabm from CRAN with:

install.packages("stabm")

For the development version, use devtools:

devtools::install_github("bommert/stabm")