Skip to contents

given formula ~ row_ids * text, where text is a character string, construct a matrix where row's are indexed by row_ids and columns are (default) bag-of-words. Perform PCA on this matrix.

Usage

pca_text(fo, tib, k, ...)

Arguments

...

Details

Extended arguments ... are passed to tidytext::unnest_tokens. for example ... as token = "ngrams", n=2 would construct matrix of bigrams.

TODO: make diagnose_text