trainForest: Train randomForest on a dataset.

In BPijuanSala/anSeq: Tools to analyse RNA-seq data

Description

Cluster cells using the spearman rank correlation as a similarity metric and (1-correlation)/2 as a distance metric.

Usage

trainForest(data, groups, numTrees = 1000, rankData = T)

Arguments

data	Dataframe or matrix containing the counts (genes x cells). It is recommended to use only the HVG for this.
groups	Vector specifying which group each cell belongs to. Cells should be as names and group ID as vector content.
numTrees	Number of trees to grow in randomForest. Default: 1000. Cells should be as names and group ID as vector content.
rankData	Logical. Rank data before training based on counts? Default: TRUE

Value

list with randomForest object (RForest) and a vector of the final genes used for training.

Author(s)

Blanca Pijuan Sala.

References

randomForest.

See Also

randomForest, allocateCellsForest

BPijuanSala/anSeq documentation built on May 30, 2019, 11:47 p.m.