Filter to words of interest

Filter to words of interest

trim_betas(betas, filter_by, threshold = 0, n_features = NULL)

Arguments

betas

A data.frame whose rows correspond to topics and whose columns are m, the model ID, and the words for which estimates are made.

filter_by

a character specifying if the data (betas) should be filtered by the average "beta" across topics or by the "distinctiveness" of the features.

threshold

Words (features) with less than this average beta or distinctiveness across all topics are ignored

n_features

Maximum number of features to show