NullModel: Calculate the Normalized Counts and Related Training...

In NBLDA: Negative Binomial Linear Discriminant Analysis

Calculate the Normalized Counts and Related Training Parameters.

Description

Fit a training set to the NBLDA model and estimate normalized counts. The related model parameters, which are used while normalizing training sets, are also returned to normalize test sets using training set parameters.

Usage

NullModel(x, type = c("mle", "deseq", "quantile", "none", "tmm"))

NullModelTest(null.out, xte = NULL)

Arguments

x	an n-by-p data frame or matrix of count data. Samples should be in the rows.
type	the normalization method. See control for details.
null.out	an object returned from NullModel.
xte	an n-by-p count matrix or data frame of test set. These counts are normalized using the training set parameters.

Value

a list with the normalized counts and the training set parameters that are used for normalizing the raw counts.

Note

These functions are copied from the PoiClaClu package and modified here to make "tmm" and "none" methods available.

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$x
xte <- counts$xte

x.out <- NullModel(x, "mle")
x.out$n ## Normalized counts using "mle" method

xte.out <- NullModelTest(x.out, xte)
xte.out$n  # Normalized counts for test set using train set parameters.

NBLDA documentation built on March 18, 2022, 7:51 p.m.