kwm: Produce a model object based on regular expressions.

In mdlincoln/kwm: Explicit Regex Matching Implemented As Model-like Objects

Description

Produce a model object based on regular expressions.

Usage

kwm(include = character(), exclude = character(), varname,
  search_fun = stringr::regex, search_opts = NULL)

## S3 method for class 'kwm'
predict(object, newdata, progress = interactive(),
  return_names = FALSE, parallel = TRUE, ...)

Arguments

include	Character. List of regular expression patterns, any one of which will return a positive match.
exclude	Character. List of regular expression patterns, any one of which will return a negative match, overriding any other matches with include
varname	Character. A character vector of length one with the column name to be tested.
search_fun	One of the stringr::modifiers functions from stringr::stringr. Defaults to stringr::regex.
search_opts	List of arguments to pass to search_fun.
object	A kwm model object
newdata	A data.frame containing a text column
progress	Logical. Display a progress bar?
return_names	Logical. Add original text as names to the resulting logical vector? Defaults to false.
parallel	Logical. Compute predictions in parallel? Requires doParallel.
...	Other arguments passed from functions. Not used currently.

Value

kwm returns a model object of class kwm

predict.kwm reutrns a logical vector.

mdlincoln/kwm documentation built on May 14, 2019, 2:15 p.m.