unnest_source_term_list: Format Term List as Individual Regexes

In trinker/termco: Counts of Terms and Substrings

Description

For validation purposes it is useful to test an uncollapsed and unnested term list to validate how well individual regexes are performing without iterative hierarchical processing. This function allows the user to read in an external term list file and format as a term list object made up of the individual regular expressions for each category. The regular expression maintain category names and are numbered sequentially. This list can then be modeled and passed to the validation tools to understand regular expression performance that might not be clear when one builds a larger model containing many tiers. After validating, the user can use the information to rearrange the regular expressions within categories at various tiers depending on the level of confidence associated with a regex.

Usage

unnest_source_term_list(file, indices = NULL, ...)

Arguments

file	Path to external term list.
indices	Indices of the elements to retain (used to take part of a term list).
...	ignored.

Value

Returns an unnested & uncollapsed term list.

Examples

## Not run: 
library(dplyr)
library(textshape)

good_cats <- system.file("termco_docs/categories.R", package = "termco")

good_cats %>%
    unnest_source_term_list()

good_cats %>%
    unnest_source_term_list() %>%
    tidy_list('category', 'regex')

## End(Not run)

trinker/termco documentation built on Jan. 7, 2022, 3:32 a.m.