Create weighted concordance lines

This code assumes that the lists of context words are stored in columns starting with _cws, followed by the name of the lemma and a model name from which first-order parameter settings can be extracted (all separated by periods). Then it creates the weighted concordance lines and stores them in columns following the same name pattern, but starting with _ctxt. In addition, it creates a non weighted column _ctxt.raw.

Usage

weightConcordance(variables, cws, lemma, ...)

Arguments

variables

Dataframe with one row per token ID and at least columns prefixed by _cws. with semicolon-separated context words.

cws

Dataframe with one row per token ID per context word as outputted by setupConcordancer.

lemma

Name of the lemma, to process model names

...

Arguments to be passed to weightLemma and getContext, in order to adapt to different ways of coding parameter settings. See vignette('weightConcordance').

Value

Enriched variables dataframe with columns containing weighted concordance lines.

Examples

if (FALSE) {
cws <- setupConcordancer(lemma, input_dir)
variables <- readr::read_tsv('path/to/file', lazy = F)
ctxts <- weightConcordance(variables, cws, lemma)
}