plot_mir_terms: Plot count of top terms associated with a miRNA name
In JulFriedrich/miRetrieve: miRNA Text Mining in Abstracts

Description Usage Arguments Details Value See Also

Plot count of top terms associated with a miRNA name.

plot_mir_terms(
  df,
  mir,
  top = 20,
  tf.idf = FALSE,
  token = "words",
  ...,
  stopwords = stopwords_miretrieve,
  stopwords_ngram = TRUE,
  normalize = TRUE,
  colour = "steelblue3",
  col.mir = miRNA,
  col.abstract = Abstract,
  col.pmid = PMID,
  title = NULL
)

`df`	Data frame containing miRNA names, abstracts, and PubMed-IDs.
`mir`	String. miRNA name of interest.
`top`	Integer. Number of top terms to plot.
`tf.idf`	Boolean. If `tf.idf = TRUE`, terms are weighed in a tf-idf fashion. miRNA names are considered as separate documents and terms often associated with one miRNA, but not with other miRNAs get more weight.
`token`	String. Specifies how abstracts shall be split up. Taken from `unnest_tokens()` in the tidytext package: "Unit for tokenizing, or a custom tokenizing function. Built-in options are "words" (default), "characters", "character_shingles", "ngrams", "skip_ngrams", "sentences", "lines", "paragraphs", "regex", (...), and "ptb" (Penn Treebank). If a function, should take a character vector and return a list of character vectors of the same length."
`...`	Additional arguments for tokenization, if necessary.
`stopwords`	Data frame containing stop words.
`stopwords_ngram`	Boolean. Specifies if stop words shall be removed from abstracts when using ngrams. Only applied when `token = 'ngrams'`.
`normalize`	Boolean. If `normalize = TRUE`, normalizes the number of abstracts to the total number of abstracts with a miRNA name in a topic. Cannot be applied with `tf.idf = TRUE`.
`colour`	String. Colour of bar plot.
`col.mir`	Symbol. Column containing miRNA names
`col.abstract`	Symbol. Column containing abstracts.
`col.pmid`	Symbol. Column containing PubMed-IDs.
`title`	String. Title plot.

Plot count of top terms associated with a miRNA name. Top terms associated with mir have to be in df as abstracts. Number of top terms to plot is regulated via the top argument. Terms can either be evaluated as their count or in a tf-idf fashion. If terms are evaluated as their count, they can either be evaluated as their raw count, e.g. in how many abstracts they are mentioned in conjunction with the miRNA name, or as their relative count, e.g. in how many abstracts containing the miRNA they are mentioned compared to all abstracts containing the miRNA. If terms are evaluated in a tf-idf fashion, miRNA names are considered as separate documents and terms often associated with one miRNA, but not with other miRNAs get more weight. plot_mir_terms() is based on the tools available in the tidytext package.

Bar plot displaying the count of the top terms associated with a miRNA name.

plot_wordcloud(), tidytext::unnest_tokens()

Other miR term functions: plot_wordcloud()

JulFriedrich/miRetrieve documentation built on Sept. 20, 2021, 11:37 p.m.