Extract top features — get_top_features • moiraine

Extracts the features with highest contribution to the latent dimensions constructed by an integration method. Can retain a specific number of top contributing features for each dataset and latent dimension, or all features above a minimum importance score.

Usage

get_top_features(
  method_output,
  n_features = 10,
  min_importance = NULL,
  latent_dimensions = NULL,
  datasets = NULL,
  mo_data = NULL
)

Arguments

method_output: Integration method output generated via the get_output() function.
n_features: Integer, the number of features to extract for each latent dimension and dataset. Ignored if min_importance is set. Default value is 10. Will include all ties.
min_importance: Numeric value between 0 and 1, minimum importance score used to select features. Default value is NULL, i.e. the top n_features features are selected instead.
latent_dimensions: Character vector of latent dimensions name. Default value is NULL (top contributing features will be returned for all latent dimensions).
datasets: Character vector of datasets name. Default value is NULL (top contributing features will be returned for all datasets).
mo_data: A MultiDataSet::MultiDataSet object.

Value

A tibble containing one row per feature and latent dimension, giving the weight and importance score of the feature for the corresponding latent dimension. If mo_data is supplied, information about the features from the features metadata will be added to the resulting table.