ngrams: Calcualte ngrams

In wiertz/rusecqp: Corpus linguistic analysis for Corpus Workbench corpora

Description

ngrams calculates ngrams for a corpus or subcorpus. Ngram calculation uses a lot of memory and may fail for large corpora or subcorpora or causing the system to hang. For computational efficiency, results are not cleaned and may contain some artifacts (i.e. ngrams with less than the specified token count).

Usage

ngrams(cqp_corpus, ngram_length, pattr = "word", ignore_punct = T)

Arguments

ngram_length	numeric. Length of ngrams.
pattr	character. Positional attribute to use for ngram calculation.
ignore_punct	logical. If TRUE, punctuation is ignored in ngram calculation.
corpus	corpus created with get_corpus.

Value

data.table with ngrams and frequencies.

Examples

## Not run: trigrams <- ngrams(my_corpus, 3, min_count = 10)

wiertz/rusecqp documentation built on Feb. 9, 2022, 1:30 p.m.