Optional return value to work with `cluster::clusGap`

[MahShaaban]

I tried to use the timeclust function to calculate the appropriate number of clusters using cluster::clusGap but got an error because cluster::clusGap expects a return value as list with an item cluster. This is in fact the case for popular clustering functions in R such as stats::kmeans and e1071::cmeans.

I understand that TCseq uses S4 classes, so I wonder whether an optional argumnet, say cluster_only = TRUE can be added to give the option of returning a list with an item cluster.

I made a simple example to show the error and a work around.

# loading required libraries
library(TCseq)
library(cluster)

# run example of timeclust
x <- matrix(rnorm(1600,sd=0.3),
            nrow = 200,
            dimnames = list(paste0('peak', 1:200), 1:8))
clust_res <- timeclust(x,
                       algo = 'km',
                       k = 4)

# using timeclust with clusGap (error)
clusGap(x,
        FUNcluster = timeclust,
        algo = 'km',
        K.max = 10)
#> Error: $ operator not defined for this S4 class

# work around
timeclust2 <- function(...) {
  # run timeclust normally
  res <- timeclust(...)
  
  # return a list of one item (cluster slot)
  return(list(cluster = res@cluster))
}

# using timeclust2 with clusGap
clusGap(x,
        FUNcluster = timeclust2,
        algo = 'km',
        K.max = 10,
        B = 10)
#> Clustering Gap statistic ["clusGap"] from call:
#> clusGap(x = x, FUNcluster = timeclust2, K.max = 10, B = 10, algo = "km")
#> B=10 simulated reference sets, k = 1..10; spaceH0="scaledPCA"
#>  --> Number of clusters (method 'firstSEmax', SE.factor=1): 1
#>           logW   E.logW       gap      SE.sim
#>  [1,] 4.065481 4.517970 0.4524893 0.011765972
#>  [2,] 4.004416 4.438673 0.4342573 0.012349933
#>  [3,] 3.961315 4.391034 0.4297197 0.009777780
#>  [4,] 3.921194 4.347815 0.4266205 0.008408944
#>  [5,] 3.894419 4.312041 0.4176216 0.009219877
#>  [6,] 3.869805 4.283160 0.4133550 0.013342736
#>  [7,] 3.843492 4.257513 0.4140212 0.007701472
#>  [8,] 3.814787 4.231085 0.4162979 0.008534093
#>  [9,] 3.794906 4.209470 0.4145634 0.010870945
#> [10,] 3.767030 4.186994 0.4199637 0.010462993

^{Created on 2019-04-12 by the reprex package (v0.2.1)}