topRank: Identify features with the highest influence on clustering.

In celda: CEllular Latent Dirichlet Allocation

Description

topRank() can quickly identify the top 'n' rows for each column of a matrix. For example, this can be useful for identifying the top 'n' features per cell.

Usage

topRank(matrix, n = 25, margin = 2, threshold = 0, decreasing = TRUE)

Arguments

matrix	Numeric matrix.
n	Integer. Maximum number of items above 'threshold' returned for each ranked row or column.
margin	Integer. Dimension of 'matrix' to rank, with 1 for rows, 2 for columns. Default 2.
threshold	Numeric. Only return ranked rows or columns in the matrix that are above this threshold. If NULL, then no threshold will be applied. Default 0.
decreasing	Logical. Specifies if the rank should be decreasing. Default TRUE.

Value

List. The 'index' variable provides the top 'n' row (feature) indices contributing the most to each column (cell). The 'names' variable provides the rownames corresponding to these indexes.

Examples

data(sampleCells)
topRanksPerCell <- topRank(sampleCells, n = 5)
topFeatureNamesForCell <- topRanksPerCell$names[1]

celda documentation built on Nov. 8, 2020, 8:24 p.m.