Skip to contents

Extract output of integration method in standard format

Source: R/output_get.R
get_output.Rd

Extract samples score and features weight from the result of an integration method. The get_output() function provides a wrapper around the methods' specific get_output_*() functions.

Usage

get_output(method_output, use_average_dimensions = TRUE)

get_output_pca(method_output)

get_output_splsda(method_output)

get_output_spls(method_output, use_average_dimensions = TRUE)

get_output_diablo(method_output, use_average_dimensions = TRUE)

get_output_mofa2(method_output)

get_output_so2pls(method_output, use_average_dimensions = TRUE)

Arguments

method_output

The output of an integration method.

use_average_dimensions

Logical, should the (weighted) average of the samples scores for each latent dimension across the datasets be used? If FALSE, a separate set of sample scores will be returned for each dataset for each of the latent dimensions. Only applies to sPLS, DIABLO and sO2PLS results. Default value is TRUE.

Value

An S3 object of class output_dimension_reduction, i.e. a named list, with the following elements:

  • features_weight: tibble of features weight (loadings) for each latent dimension, with columns feature_id, dataset, latent_dimension, weight (unscaled feature weight for the corresponding latent dimension), importance (which corresponds to the scaled absolute weight, i.e. 1 = feature has the maximum absolute weight for the corresponding latent dimension and dataset, 0 = the feature was not selected for the corresponding latent dimension)

  • samples_score: tibble of samples score for each latent component, with columns sample_id, latent_dimension, score (unscaled samples score for the corresponding latent dimension)

  • variance_explained: tibble of the fraction of variance explained by each latent component for the relevant datasets.