canberra_stability: Canberra Stability

In OmicsMarkeR: Classification and Feature Selection for 'Omics' Datasets

Description

Calculates canberra stability between two ranked lists. In brief, the raw canberra distance is scaled to a [0,1] distribution by the maximum canberra metric. Lastly, this value is subtracted from 1 to provide the same interpretation as the other stability metrics whereby 1 is identical and 0 is no stability.

Usage

canberra_stability(x, y)

Arguments

x	numeric vector of ranks
y	numeric vector of ranks with compatible length to x

Value

Returns the canberra stability for the two vectors

Author(s)

Charles E. Determan Jr.

References

Jurman G., Merler S., Barla A., Paoli S., Galea A., & Furlanello C. (2008) Algebraic stability indicators for ranked lists in molecular profiling. Bioinformatics 24(2): 258-264.

He. Z. & Weichuan Y. (2010) Stable feature selection for biomarker discovery. Computational Biology and Chemistry 34 215-225.

Examples

# Canberra demo
v1 <- seq(10)
v2 <- sample(v1, 10)
canberra(v1, v2)

canberra_stability(v1, v2)

Example output

[1] 2.752969
[1] 0.3943468

OmicsMarkeR documentation built on April 28, 2020, 6:54 p.m.