Skip to contents

Run PCA on MultiDataSet

Source: R/multidataset_pca.R
run_pca.Rd

Runs a Principal Component Analysis on an omics dataset from a MultiDataSet object. This is a wrapper function around the get_dataset_matrix() and run_pca_matrix() functions.

Usage

run_pca(
  mo_data,
  dataset_name,
  n_pcs = 10,
  scale = "none",
  center = TRUE,
  method = NULL
)

Arguments

mo_data

A MultiDataSet::MultiDataSet object.

dataset_name

Character, name of the omics dataset on which a PCA should be run.

n_pcs

numeric, number of Principal Components to compute. Default value is 10.

scale

character, type of scaling that should be applied to the dataset before running the PCA. Should be one of 'none', 'pareto', 'vector', 'uv' (see pcaMethods::pca()). Default value is 'none'.

center

boolean, should the dataset be centred prior to running the PCA? Default value is TRUE.

method

character, type of PCA that should be applied to the dataset. See pcaMethods::listPcaMethods(). for a list of available methods. Default value is 'svd' for datasets with no missing value, and 'nipals' for datasets with missing values.

Value

A pcaMethods::pcaRes object containing the result from the PCA analysis. The attribute dataset_name specifies the name of the dataset analysed.

Details

To facilitate the use of dynamic branching with the targets package, the dataset_name attribute of the resulting object is set as the value of the dataset_name parameter, and can be accessed via attr(res_pca, "dataset_name").