best_iter: best_iter

In songjiajia2018/LINEAGE: Label-Free Identification of Endogenous Informative Single Cell Mitochondrial RNA Mutation for Lineage Analysis and Clonal Evolution

best_iter

Description

best_iter performs a non-parallel iteration process of main to guarantee a more stable and reliable cell clustering/clone tracing result.

Usage

best_iter(d, centers = c(2, 3), nmarker = c(10, 15, 20), repeats = 30)

Arguments

d	A list containing 12 submatrices with different mutation types. Output of data_prepare(). Required.
centers	Integer. The number of clusters used in Kmeans procedure. Default: c(2,3).
nmarker	Integer. The number of markers showed in final result. Default: c(10,15,20).
repeats	Integer. The number of iterations. Default: 30.

Details

Non-parallel iterative optimization

Value

A list containing results of all repeats, including the best result.

Examples

data("TF1_clones")
data=TF1_clones$data
d=data_prepare(data)
a=dim(d[[1]])[2]
if(a>100){
    nmarker=c(15,20)
    centers=3
}else{
    nmarker=c(10,15)
    centers=2
}
all_results=best_iter(d, centers, nmarker, repeats=30)

songjiajia2018/LINEAGE documentation built on Oct. 17, 2022, 6:17 a.m.