R/bk_method.R
FM_index_profdpm.Rd
Calculating Fowlkes-Mallows index using the pci
function from
the profdpm R package. This function uses C code (thanks to
Matthew Shotwell's work) and is a bit faster than the R code
(from my simple tests - it is about 1.2-1.3 times faster - which is not much
but it might be useful).
As opposed to the FM_index_R
function, the FM_index_profdpm
function does NOT calculate the expectancy or the variance of the FM Index
under the null hypothesis of no relation.
This function also allows us to compare our calculations with an independent writing of a function calculating the same statistic.
FM_index_profdpm( A1_clusters, A2_clusters, assume_sorted_vectors = FALSE, warn = dendextend_options("warn"), ... )
A1_clusters | a numeric vector of cluster grouping (numeric) of items, with a name attribute of item name for each element from group A1. These are often obtained by using some k cut on a dendrogram. |
---|---|
A2_clusters | a numeric vector of cluster grouping (numeric) of items, with a name attribute of item name for each element from group A2. These are often obtained by using some k cut on a dendrogram. |
assume_sorted_vectors | logical (FALSE). Can we assume to two group vectors are sorter so that they have the same order of items? IF FALSE (default), then the vectors will be sorted based on their name attribute. |
warn | logical (default from dendextend_options("warn") is FALSE). Set if warning are to be issued, it is safer to keep this at TRUE, but for keeping the noise down, the default is FALSE. |
... | Ignored. |
The Fowlkes-Mallows index between two vectors of clustering groups.
From Wikipedia:
Fowlkes-Mallows index (see references) is an external evaluation method that is used to determine the similarity between two clusterings (clusters obtained after a clustering algorithm). This measure of similarity could be either between two hierarchical clusterings or a clustering and a benchmark classification. A higher the value for the Fowlkes-Mallows index indicates a greater similarity between the clusters and the benchmark classifications.
Fowlkes, E. B.; Mallows, C. L. (1 September 1983). "A Method for Comparing Two Hierarchical Clusterings". Journal of the American Statistical Association 78 (383): 553.
Shotwell, Matthew S. "profdpm: An R Package for MAP Estimation in a Class of Conjugate Product Partition Models." Journal of Statistical Software 53: 1-18.
http://en.wikipedia.org/wiki/Fowlkes-Mallows_index
if (FALSE) { set.seed(23235) ss <- TRUE # sample(1:150, 10 ) hc1 <- hclust(dist(iris[ss, -5]), "com") hc2 <- hclust(dist(iris[ss, -5]), "single") # dend1 <- as.dendrogram(hc1) # dend2 <- as.dendrogram(hc2) # cutree(dend1) FM_index_profdpm(cutree(hc1, k = 3), cutree(hc1, k = 3)) # 1 set.seed(1341) FM_index_profdpm(cutree(hc1, k = 3), sample(cutree(hc1, k = 3)), assume_sorted_vectors = TRUE ) # 0.38037 FM_index_profdpm(cutree(hc1, k = 3), sample(cutree(hc1, k = 3)), assume_sorted_vectors = FALSE ) # 1 again :) FM_index_profdpm(cutree(hc1, k = 3), cutree(hc2, k = 3)) # 0.8059 FM_index_profdpm(cutree(hc1, k = 30), cutree(hc2, k = 30)) # 0.4529 fo <- function(k) FM_index_profdpm(cutree(hc1, k), cutree(hc2, k)) lapply(1:4, fo) ks <- 1:150 plot(sapply(ks, fo) ~ ks, type = "b", main = "Bk plot for the iris dataset") }