convertRowID: A function to change the row names of a matrix.

In rikenbit/scTGIF: Cell type annotation for unannotated single-cell RNA-Seq data

A function to change the row names of a matrix.

Description

To avoid to specify the duplicated row names against matrix, multiple aggregation rules are implemented.

Usage

convertRowID(input, rowID, LtoR,
    aggr.rule=c("sum", "mean", "large.mean", "large.var", "large.cv2"))

Arguments

input	A matrix filled with number (n * m).
rowID	A vector to specify the row names of input (length: n).
LtoR	A corresponding table to covert the row names of input as different type of IDs. (Left: current row names -> Right: new row names)
aggr.rule	The aggregation rule to change the row names of input and collapse/select the values, if the row names changed by LtoR are duplicated. sum: Collapses multiple row vectors by summation. mean: Collapses multiple row vectors by mean. large.mean: Select a vector having the largest mean in the duplicated vectors. large.var: Select a vector having the largest variance in the duplicated vectors. large.cv2: Select a vector having the largest CV2 in the duplicated vectors.

Value

output	A matrix, in which the row names are changed, according to the aggregation rule user specified.
ctable	The corresponding table explaining the relationship between previous row names and changed row names of input.

Author(s)

NA

Examples

    input <- matrix(1:20, nrow=4, ncol=5)
    rowID <- c("A", "B", "C", "D")
    LtoR <- rbind(
      c("A", "3"),
      c("B", "2"),
      c("C", "4"),
      c("D", "7"))
    (input2 <- convertRowID(input, rowID, LtoR, "sum"))
    (input3 <- convertRowID(input, rowID, LtoR, "mean"))
    (input4 <- convertRowID(input, rowID, LtoR, "large.mean"))
    (input5 <- convertRowID(input, rowID, LtoR, "large.var"))
    (input6 <- convertRowID(input, rowID, LtoR, "large.cv2"))

rikenbit/scTGIF documentation built on Jan. 14, 2023, 4:18 a.m.