romicsPCA: romicsPCA()

In PNNL-Comp-Mass-Spec/RomicsProcessor: omics-oriented package for reproducible and sharable data processing pipelines

romicsPCA()

Description

Calculate the PCA of the data layer of the romics_object using the package FactoMineR. If the data layer contains some missing values those will be imputed using the missMDA::imputePCA() method (see the documentation of this function for more details). This function will return the PCA results and not a romics_object

Usage

romicsPCA(
  romics_object,
  ncp = 5,
  scale = TRUE,
  method = c("Regularized", "EM"),
  ncp.min = 0,
  ncp.max = 5,
  method.cv = c("gcv", "loo", "Kfold"),
  ...
)

Arguments

romics_object	has to be a log transformed romics_object created using romicsCreateObject() and transformed using the function log2transform() or log10transform()
ncp	inherited from missmda::imputePCA().
scale	inherited from missmda::imputePCA(). boolean. TRUE implies a same weight for each variable
method	inherited from missmda::imputePCA(). "Regularized" by default or "EM". TRUE implies a same weight for each variable
ncp.min	used only if ncp is not set. inherited from missmda::estim_ncpPCA().integer corresponding to the minimum number of components to test
ncp.max	used only if ncp is not set. inherited from missmda::estim_ncpPCA().integer corresponding to the minimum number of components to test
...	further arguments passed to or from other methods
row.w	inherited from missmda::imputePCA(). row weights (by default, a vector of 1 for uniform row weights)

Details

This function uses the dist() and hclust() functions to calculate the hierachical clustering and then plots the hclust with colors based on the current main_factor of the romics_object.

Value

Return the results of the PCA performed on the current version of the romics_object

Author(s)

Geremy Clair

PNNL-Comp-Mass-Spec/RomicsProcessor documentation built on March 18, 2023, 5:14 a.m.