R/sentiment.R

In gvegayon/twitterreport: Out-of-the-Box Analysis and Reporting Tools for Twitter

#' Computes sentiment score
#' 
#' Computes sentiment score by counting the number of positive and negative terms.
#'
#' @param text Character vector with the text to be classified
#' @param pos Positive words lexicon
#' @param neg Negative words lexicon
#' @param pos_s Numeric vector of scores for each word in the pos vec
#' @param neg_s Numeric vector of scores for each word in the neg vec
#' @param lan Languaje of the lexicon (can be either 'en' or 'es')
#' @param normalize Whether or not to normalize the values of each words' score.
#' @details By default uses an english lexicon downloaded from 
#' http://www.cs.uic.edu/~liub/FBS/sentiment-analysis.html. In particular
#' if no list of words is provided, the function loads the dataset 
#' [warriner_et_al_en()] or [warriner_et_al_es()], depending
#' on the `lang` argument, both of which contains a vector of words with
#' their respective *valence* (pleasantness of the stimulus).
#' 
#' The arguments `pos_s` and `neg_s` allow to provide with scores
#' (valence) to each vector of words, so instead of adding 1 or -1 for every
#' positive or negative word the function can add some other value specified by
#' these arguments.
#' 
#' By default, when using the default lexicon, the function loads the english
#' lexicon (`lan='en'`). Otherwise, if `lan` is set to 'es', it
#' will load the spanish lexicon.
#' 
#' When `normalize` is set to `TRUE` (default), the function will normalize
#' the scores values for each word such as all values are within -1 and 1. Note that
#' this does not implies that the final sentiment score will be in that range as
#' well since it is a result of the addition of the scores.
#' 
#' @seealso [sentiment_lexicon_pos_en()] [sentiment_lexicon_neg_en()]
#' @return Numeric Vector with scores
#' @examples 
#'  # Example data
#'  text <- c(
#'    "I'm feeling sad today",
#'    "This is terrible, bad",
#'    "I'm doing ok",
#'    "I'm feeling great",
#'    "I'm feeling nothing",
#'    "This is sick, but good at the same time")
#'  
#'  # Getting the scores
#'  tw_sentiment(text)
#'  tw_sentiment(text, normalize=FALSE)
#' @family statistics
#' @export
tw_sentiment <- function(
  text, pos=NULL, neg=NULL, pos_s=NULL, neg_s=NULL, lan='en', normalize=TRUE) {
  
  # Processing text
  x <- sapply(tolower(text), strsplit, split="[^a-zA-Z']")
  x <- lapply(x, gsub, pattern="[^a-zA-Z]", replacement="")
  
  path <- system.file('data/',package='twitterreport')
  
  # Checking language
  if (!(lan %in% c('en','es'))) stop('Invalid language, only \'es\' or \'en\' are supported')
  
  # Checking list of words
  if (!length(pos)) {
    pos<-read.table(sep=';',header = TRUE,file=
    paste0(path,'/warriner_et_al_',lan,'.csv.xz'), as.is=TRUE)
    pos_s <- pos$valence
    pos   <- pos$word
  }
  if (!length(neg)) {
    neg   <- character()
    neg_s <- numeric()
  }
  
  # Checking scores
  if (!length(pos_s)) pos_s <- rep(1, length(pos))
  if (!length(neg_s)) neg_s <- rep(-1, length(neg))
  
  cpp_tw_sentiment(x, pos, neg, pos_s, neg_s, normalize = normalize)
}