M_ABCdist.WC: Multi-source M-ABC fused with WeightedClust

View source: R/M_ABCdist_WC.R

M_ABCdist.WCR Documentation

Multi-source M-ABC fused with WeightedClust

Description

Runs ABCdist.SingleInMultiple on each source, then feeds the per-source dissimilarities to WeightedClust (type = "dist") to explore convex combinations over a weight grid and return the focus-weight clustering.

Usage

M_ABCdist.WC(
  List,
  distmeasure = rep("euclidean", length(List)),
  weighting = rep("var", length(List)),
  normalize = rep(list(NULL), length(List)),
  gr = c(),
  bag = rep(TRUE, length(List)),
  numsim = 1000L,
  numvar = rep(100L, length(List)),
  linkage = rep("ward.D", length(List)),
  NC = NULL,
  NC2 = NULL,
  StopRange = FALSE,
  wc_weight = seq(1, 0, -0.1),
  wc_weightclust = 0.5,
  wc_linkage = "ward",
  wc_alpha = 0.625,
  mds = FALSE,
  nugget_type = "between",
  nugget_args = list()
)

Arguments

List

A list of K data matrices over the same objects, objects (samples) in rows and features in columns.

distmeasure, weighting, normalize, bag, numvar, linkage

Per-source vectors (recycled to K) forwarded to ABCpp.SingleInMultiple.

gr

Unused; kept for compatibility.

numsim

Iterations per source.

NC

Base-cluster count per source (NULL = data-driven).

NC2

Optional number of clusters to cut the final tree into.

StopRange

Passed to WeightedClust.

wc_weight

Weight grid for WeightedClust.

wc_weightclust

Focus weight whose result is returned in Clust.

wc_linkage, wc_alpha

Final clustering controls for WeightedClust.

mds

Logical; MDS of the consensus dissimilarities.

nugget_type, nugget_args

Forwarded to the nugget weighting.

Value

A list of class "M_ABCdist_WC" with source_D, the full WC object, the focus DistM and Clust, and the weight_grid.

References

\insertCite

Amaratunga2008MosaiClusteR

See Also

M_ABCdist, WeightedClust

Examples

data(mosaic_toy)
fit <- M_ABCdist.WC(mosaic_toy$List, numsim = 40,
                    wc_weight = seq(1, 0, -0.25), wc_weightclust = 0.5)
dim(fit$DistM)

MosaiClusteR documentation built on Oct. 25, 2026, 1:08 a.m.