clean_news: Clean retrieved news articles

In MikeJohnPage/newsrivr: Newsriver API R Client

Description

clean_news wrangles the messy data fetched by get_news, returning a tidy tibble with sensible defaults.

Usage

clean_news(data, min_nchar = 300, as_date = TRUE, drop_vars = TRUE,
  to_lower = TRUE, distinct = TRUE, drop_na = FALSE,
  tif_corpus = FALSE)

Arguments

data	Tbl, returned from get_news.
min_nchar	Integer, specifying the minimum number of characters of articles to be kept in the corpus.
as_date	Logical, indicating whether dates should be transformed to class "Date".
drop_vars	Logical, indicating whether all variables (other than title, text, discoverDate, & website.domainName) should be dropped. The Newsriver API (typically) returns 26 variables, many of which contain sparse metadata.
to_lower	Logical, indicating whether the title and text variables should be transformed to lowercase.
distinct	Logical, indicating whether only articles with either distinct title or text values should be kept.
drop_na	Logical, indicating whether to drop rows containing missing values.
tif_corpus	Logical, indicating whether the tibble should be a TIF valid corpus.

Examples

## Not run: 
clean_news(data = my_tbl)

clean_news(my_tbl, min_nchar = 500, tif_corpus = TRUE)

## End(Not run)

MikeJohnPage/newsrivr documentation built on Jan. 4, 2021, 7:48 p.m.