GuidedSparseKmeans.KLam: GuidedSparseKmeans.KLam

In LingsongMeng/GuidedSparseKmeans: Guided Sparse K-means

Description

Selection of Tuning Parameter K and lam in Guided Sparse K-means

Usage

GuidedSparseKmeans.KLam(x, z, pre.K = NULL, s.one, model, nstart = 20,
  maxiter = 15, silence = F)

Arguments

x	Gene expression matrix, n*p (rows for subjects and columns for genes).
z	One phenotypic variable from clinical dataset, a vector.
pre.K	Pre-knowledge of the number of clusters.
s.one	A proper value of the boundary of l1n weights.
model	The model fitted to obtain R2, please select model from 'linear', 'logit', 'exp', 'polr','cox'.
nstart	Specify the number of starting point for K-means.
maxiter	Maximum number of iteration.
silence	Output progress or not.

Details

Select tuning parameter K using gap statistics and tuning parameter lam using sensitivity analysis in Guided Sparse K-means integrating clinical dataset with gene expression dataset.

Value

A list consisting of

K.select	value of selected K.
lam.select	value of selected lam.
R2.per	R-squared or pseudo R-squared between phenotypic variable and expression value of each gene, a vector.
ARI.Cs	Adjusted ARI values for cluster results.
Jaccard.gene	Jaccard index values for gene selection results.

Author(s)

Lingsong Meng

LingsongMeng/GuidedSparseKmeans documentation built on May 11, 2020, 12:29 a.m.