Description Usage Arguments Value Note Author(s) References See Also Examples
Calculates Kuncheva's index between two vectors of features. In brief, the closer to 1 the more similar the vectors. The two vectors must have the same cardinality (i.e. same length).
1 | kuncheva(x, y, num.features)
|
x |
Character vector of feature names |
y |
Character vector of feature names |
num.features |
total number of features in the original dataset |
Returns the Kuncheva Index for the two vectors. It takes values in [0,1], with 0 meaning no overlap between two sets and 1 meaning two sets are identical.
The returned Kuncheva Index has been scaled from its original [-1,1] range to [0,1] in order to make it compatible with RPT.
Charles E. Determan Jr.
Kuncheva L. (2007) A stability index for feature selection. Proceedings of the 25th IASTED International Multi-Conference: Artificial Intelligence and Applications. pp. 390-395.
He. Z. & Weichuan Y. (2010) Stable feature selection for biomarker discovery. Computational Biology and Chemistry 34 215-225.
kuncheva
, sorensen
,
ochiai
, pof
, pairwise.stability
,
pairwise.model.stability
1 2 3 4 5 6 7 8 9 10 11 12 | # Kuncheva demo
# Assuming 50 metabolites were measured
# But only 10 were found significant
# For demonstration purposes only!!!
some.numbers <- seq(20)
# Metabolites identified from one run
v1 <- paste("Metabolite", sample(some.numbers, 10), sep="_")
# Metabolites identifed from second run
v2 <- paste("Metabolite", sample(some.numbers, 10), sep="_")
kuncheva(v1, v2, 50)
|
[1] 0.6875
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.