Create a text corpus from graph text attribute data

This function creates a text corpus from node or edge text attribute data in an igraph.

corpusFromGraph(
  g = NULL,
  txt_attr = NULL,
  type = "vertex",
  iconv = FALSE,
  html_decode = TRUE,
  rm_url = TRUE,
  rm_num = TRUE,
  rm_punct = TRUE,
  rm_twit_hashtags = FALSE,
  rm_twit_users = FALSE,
  sw_kind = "SMART",
  rm_words = NULL,
  stem = FALSE
)

Arguments

g: an igraph graph object.
txt_attr: Character string. Name of graph text attribute. Default is NULL.
type: Character string. Graph attribute type. Default is "vertex".
iconv: Logical. Use the iconv function to attempt UTF8 conversion. Default is FALSE.
html_decode: Logical. HTML decode text. Default is TRUE.
rm_url: Logical. Remove URL's. Default is TRUE.
rm_num: Logical. Remove numbers. Default is TRUE.
rm_punct: Logical. Remove punctuation. Default is TRUE.
rm_twit_hashtags: Logical. Remove twitter hashtags. Default is FALSE.
rm_twit_users: Logical. Remove twitter user names. Default is FALSE.
sw_kind: Character string. Stopword dictionary. Refer stopwords kind parameter. Default is "SMART".
rm_words: Character vector. User defined stopwords. Default is NULL.
stem: Logical. Apply word stemming. Default is FALSE.

Value

A tm text corpus object.