Load twitterAnalsis package
library(twitterAnalysis)
#> Loading required package: dplyr
#>
#> Attaching package: 'dplyr'
#> The following objects are masked from 'package:stats':
#>
#> filter, lag
#> The following objects are masked from 'package:base':
#>
#> intersect, setdiff, setequal, union
#> Loading required package: stopwords
#> Loading required package: tm
#> Loading required package: NLP
#>
#> Attaching package: 'tm'
#> The following object is masked from 'package:stopwords':
#>
#> stopwords
#> Loading required package: wordcloud
#> Loading required package: RColorBrewer
#> Loading required package: twitteR
#>
#> Attaching package: 'twitteR'
#> The following objects are masked from 'package:dplyr':
#>
#> id, location
#> Loading required package: syuzhetObtain Twitter API credentials
From https://developer.twitter.com/ obtain API access credentials.
consumer_key <- 'xxx'
consumer_secret <- 'xxx'
access_token <- 'xxx'
token_secret <- 'xxx'Quick use example
The functions of the package are all meant to be used in tandem with each other.
Step 1: Setup Twitter API credentials
api_user <- user_info(consumer_key, consumer_secret, access_token, token_secret)Step 2: get 30 most recent tweets of a user (for example,
@elonmusk)
elon_df <- load_twitter_by_user('elonmusk', 100, api_user)
#> [1] "Using direct authentication"
head(elon_df)
#> screenName id created favoriteCount retweetCount
#> 1 elonmusk 1621886041245237253 2023-02-04 14:59:00 2242 84
#> 2 elonmusk 1621881994375299073 2023-02-04 14:42:55 3146 142
#> 3 elonmusk 1621881617978429440 2023-02-04 14:41:25 1444 87
#> 4 elonmusk 1621793704175362050 2023-02-04 08:52:05 13767 426
#> 5 elonmusk 1621719831249567744 2023-02-04 03:58:32 6621 301
#> 6 elonmusk 1621719712429129728 2023-02-04 03:58:04 23223 653
#> text
#> 1 @cb_doge 🤣
#> 2 @d4t4wr4ngl3r 🤣🤣
#> 3 @BillyM2k @cb_doge 🤣 it’s a fair cop
#> 4 @TheJackForge 🤣
#> 5 @tobi !
#> 6 @TrungTPhan 🤣Step 3: Preform pre-processing
elon_clean_df <- generalPreprocessing(elon_df)
head(elon_clean_df)
#> screenName id created favoriteCount retweetCount
#> 1 elonmusk 1621886041245237253 2023-02-04 14:59:00 2242 84
#> 2 elonmusk 1621881994375299073 2023-02-04 14:42:55 3146 142
#> 3 elonmusk 1621881617978429440 2023-02-04 14:41:25 1444 87
#> 4 elonmusk 1621793704175362050 2023-02-04 08:52:05 13767 426
#> 5 elonmusk 1621719831249567744 2023-02-04 03:58:32 6621 301
#> 6 elonmusk 1621719712429129728 2023-02-04 03:58:04 23223 653
#> text text_clean
#> 1 @cb_doge 🤣 🤣
#> 2 @d4t4wr4ngl3r 🤣🤣 🤣🤣
#> 3 @BillyM2k @cb_doge 🤣 it’s a fair cop 🤣 it’s a fair cop
#> 4 @TheJackForge 🤣 🤣
#> 5 @tobi ! !
#> 6 @TrungTPhan 🤣 🤣Step 4: Preform Sentiment Analysis
elon_sentiment_df <- sentiment_labeler(elon_clean_df, 'text')
#> Warning: `spread_()` was deprecated in tidyr 1.2.0.
#> ℹ Please use `spread()` instead.
#> ℹ The deprecated feature was likely used in the syuzhet package.
#> Please report the issue to the authors.
head(elon_sentiment_df)
#> screenName id created favoriteCount retweetCount
#> 1 elonmusk 1621886041245237253 2023-02-04 14:59:00 2242 84
#> 2 elonmusk 1621881994375299073 2023-02-04 14:42:55 3146 142
#> 3 elonmusk 1621881617978429440 2023-02-04 14:41:25 1444 87
#> 4 elonmusk 1621793704175362050 2023-02-04 08:52:05 13767 426
#> 5 elonmusk 1621719831249567744 2023-02-04 03:58:32 6621 301
#> 6 elonmusk 1621719712429129728 2023-02-04 03:58:04 23223 653
#> text text_clean sentiment
#> 1 @cb_doge 🤣 🤣 neutral
#> 2 @d4t4wr4ngl3r 🤣🤣 🤣🤣 neutral
#> 3 @BillyM2k @cb_doge 🤣 it’s a fair cop 🤣 it’s a fair cop positive
#> 4 @TheJackForge 🤣 🤣 neutral
#> 5 @tobi ! ! neutral
#> 6 @TrungTPhan 🤣 🤣 neutralStep 5: Generate a WordCloud
elon_plt_clean_df <- clean_tweets(elon_sentiment_df)
elon_count_df <- count_words(elon_plt_clean_df)
plt <- create_wordcloud(elon_count_df)
Optional: We can also determine the proportion of
neutral, positive, and netaive
tweets with count_tweets().
count_tweets(elon_sentiment_df)
#> $negative
#> [1] 0.1041667
#>
#> $neutral
#> [1] 0.6458333
#>
#> $positive
#> [1] 0.25