R/tweet_topics.R

In saotd: Sentiment Analysis of Twitter Data

#' @title Tweet Topics
#'
#' @description Determines the Latent topics within a data frame by using Latent
#'   Dirichlet Allocation (LDA) model parameters.  Uses the `ldatuning` package 
#'   and outputs an ldatuning plot.  Prepares Tweet text, creates DTM, conducts 
#'   LDA, display data terms associated with each topic.
#'
#' @param DataFrame Data Frame of Twitter Data.
#' @param clusters The number of latent clusters.
#' @param method method = "Gibbs"
#' @param set_seed Seed for reproducible results.
#' @param num_terms The desired number of terms to be returned for each topic.
#' 
#' @importFrom dplyr mutate group_by anti_join inner_join count select transmute quo rename
#' @importFrom stringr str_replace_all
#' @importFrom tidytext cast_dtm
#' @importFrom topicmodels LDA topics terms
#' 
#' @return Returns LDA topics.
#' 
#' @examples
#' \dontrun{
#' library(saotd)
#' data <- raw_tweets
#' LDA_data <- tweet_topics(DataFrame = data,
#'                          clusters = 8,
#'                          method = "Gibbs",
#'                          set_seed = 1234,
#'                          num_terms = 10)
#'
#' LDA_data
#' }
#' @export

tweet_topics <- function(DataFrame,
                         clusters,
                         method = "Gibbs",
                         num_terms = 10,
                         set_seed = 1234) {

  # input checks
  if (!is.data.frame(DataFrame)) {
    stop("The input for this function is a data frame.")
  }

  if (!is.numeric(clusters)) {
    stop("The input must be a numerical value.")
  }

  # configure defusing operators for packages checking
  text <- dplyr::quo(text)
  key <- dplyr::quo(key)
  word <- dplyr::quo(word)
  n <- dplyr::quo(n)

  # web url
  wu <- "https://t.co/[A-Za-z\\d]+|http://[A-Za-z\\d]+|&amp;|&lt;|&gt;|RT|https"

  # function main body
  lda_prep <- DataFrame %>%
    dplyr::mutate(
      text = base::iconv(
        x = DataFrame$text,
        from = "latin1",
        to = "ASCII",
        sub = ""),
      text = stringr::str_replace_all(
        string = text,
        pattern = "#",
        replacement = ""), # Remove hashtag
      text = stringr::str_replace_all(
        string = text,
        pattern = "[:punct:]",
        replacement = ""), # Remove punctuation
      text = stringr::str_replace_all(
        string = text,
        pattern = "RT",
        replacement = ""), # Remove retweet note
      text = stringr::str_replace_all(
        string = text,
        pattern = "&amp",
        replacement = ""), # Remove Accelerated Mobile Pages (AMP) note
      text = stringr::str_replace_all(
        string = text,
        pattern = wu,
        replacement = "")) %>%  # Remove links
    dplyr::group_by(key) %>%
    tidytext::unnest_tokens(word, text) %>%
    dplyr::anti_join(tidytext::stop_words) %>%
    dplyr::count(key, word, sort = TRUE) %>%
    tidytext::cast_dtm(key, word, n)

  # Run LDA using Gibbs sampling
  ldaout <- topicmodels::LDA(
    lda_prep,
    k = clusters,
    method = method,
    control = list(seed = set_seed))

  # probabilities associated with each topic assignment
  data.topics <- topicmodels::topics(ldaout, 1)
  data.terms <- as.data.frame(
    topicmodels::terms(ldaout,
                       num_terms),
    stringsAsFactors = FALSE)
  print(data.terms)

  # Creates a data frame to store the Lesson Number and the most likely topic
  tweettopics.df <- as.data.frame(data.topics)
  tweettopics.df <- dplyr::transmute(
    tweettopics.df,
    LessonId = rownames(tweettopics.df),
    Topic = data.topics)
  tweettopics.df$ArticleNo <- as.character(tweettopics.df$LessonId)

  # Clean up and rename columns to match previous data frames
  tweettopics <- tweettopics.df %>%
    dplyr::select(c("ArticleNo", "Topic")) %>%
    dplyr::rename(c("key" = "ArticleNo"))

  # Join original Twitter data frame with Tweet topics
  tweet_topics <- dplyr::inner_join(
    x = DataFrame,
    y = tweettopics,
    by = "key")

  invisible()

  return(tweet_topics)

}