FindBestTransform: Find the Power Transformation Parameter.

In dncR/NBLDA: Negative Binomial Linear Discriminant Analysis

Description

Use this function to find a constant value of alpha to be used for transforming count data. The power transformation parameter alpha, which approximately fits transformed data to the Poisson log-linear model, is selected using a grid search within the interval [0, 1].

Usage

FindBestTransform(x, grid.length = 50)

Arguments

x	an n-by-p data frame or matrix of count data. Samples should be in the rows.
grid.length	how many distinct points of alpha should be searched within the interval [0, 1]? Default is 50.

Value

the value of alpha to be used within the power transformation.

Note

This function is copied from PoiClaClu package and modified to control the total number of grid search.

Author(s)

Dincer Goksuluk

Examples

set.seed(2128)
counts <- generateCountData(n = 20, p = 10, K = 2, param = 1, sdsignal = 0.5, DE = 0.8,
                            allZero.rm = FALSE, tag.samples = TRUE)

x <- counts$x
FindBestTransform(x)

dncR/NBLDA documentation built on Feb. 22, 2022, 12:59 a.m.