multivariateCurveResolution: Multivariate curve resolution

Description Usage Arguments Value Author(s)

View source: R/multivariateCurveResolution.R

Description

Function to perform multivariate curve resolution by alternating total least squares.

Usage

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
multivariateCurveResolution(
  expressionMatrix,
  residualMatrix = NULL,
  numberOfComponents = 3,
  initAlgorithm = "simplisma",
  maxIterations = 2000,
  tolerance = 9e-04,
  randomSeed = 91120,
  computeGeneProfiles = TRUE,
  verbose = TRUE
)

Arguments

expressionMatrix

The expression matrix organized as (*n* X *m*), *n* is the number of genes and *m* is the number of samples. You should use the expressionMatrix output of prepareData().

residualMatrix

The residual matrix, whose dimensions are (*n* X *m*). You should use the residualMatrix output of prepareData(). If the residual matrix is specified, the function uses the error weighted multivariate curve resolution algorithm. If the residual matrix is NULL, the function uses the non-weighted multivariate curve resolution algorithm.

numberOfComponents

The number of number of components (profiles) to be extracted.

initAlgorithm

Which algorithm to use to initialize the P/C values. The choice is between "simplisma" (recommended) and "random".

maxIterations

The maximum number of iteration for the maximum likelihood algorithm.

tolerance

The convergence tolerance. Once this is reached, the algorithm stops.

randomSeed

The random seed number, only used if initAlgorithm is "random".

computeGeneProfiles

Whether to compute the gene profiler or not.

verbose

Whether to display the information about the computation or not.

Value

The output of is a list containing: Popt, the final estimate of the profile (spectral) matrix, where vectors (rows) are constrained to be non-negative and unit length. Copt, the final estimate of contribution matrix, constrained to be non-negative. estimatedMatrix, the final estimated matrix. geneProfiles, the gene profiles (only if computeGeneProfiles = TRUE).

Author(s)

Tobias K. Karakach, Federico Taverna


itikadi/EMOGEA documentation built on Dec. 20, 2021, 8:03 p.m.