In sk6aus6/hashtag:

knitr::opts_chunk$set(
  collapse = TRUE,
  warning = FALSE,
  comment = "#>",
  fig.path = "README-"
)

library(readr)

hashtag

hashtag is a compilation of commands from different twitter and sentiment analysis packages that help you do simple twitter scrapes and analyses - fast.

So far, only "afinn" sentiment analysis is available, which classifies words as 'positive' or 'negative' with an appropriate score. Give it a try!

Note, the user must set up their own twitter API connection. I found this site helpful: rStatistics.net

Installation

devtools::install_github("sk6aus6/hashtag")

library(hashtag)
library(readr)
library(twitteR)

Quick demo

First connect to Twitter. I store my twitter API keys in a file buried deep in my computer:

keys <- read_csv('/Users/yam/OneDrive/shared files/Statslearningcourse/twitteR/keys.csv')

# setup_twitter_oauth(consumerKey, consumerSecret, accessToken, accessTokenSecret)

setup_twitter_oauth(keys$key[1],
                    keys$key[2],
                    keys$key[3], 
                    keys$key[4])

Then write a regular expression you want to search twitter for, and use this as an input for hash_sentiment()

hash_sentiment(regex = "coriander|cilantro", num.tweets = 800,
               method = "afinn", sentiment = "positive", num.summary = 3,
               scrape = TRUE)

If you want to look at different parts of the same twitter scrape, set scrape = FALSE

hash_sentiment(regex = "coriander|cilantro", num.tweets = 800,
               method = "afinn", sentiment = "negative", num.summary = 6,
               scrape = FALSE)