Sentiment-Analysis

# Load Requried Packages
library(SnowballC)
library(tm)
library(twitteR)
library(dplyr)
library(syuzhet)
library(tidytext)
library(wordcloud)
library(ggplot2)
library(leaflet)
library(gganimate)
library(lubridate)
library(maps)
library(ggthemes)
library(tidyr)
library(viridis)#color palettes
library(radarchart)#for wheels emotions
consumer_key <- 'put ur id'
consumer_secret <- 'put ur id'
access_token <- 'put ur id'
access_secret <- 'put ur id'
#connection
setup_twitter_oauth(consumer_key, consumer_secret, access_token, access_secret)
tweets <- searchTwitter("#india", n=100,lang = "en")
View(tweets)
n.tweet <- length(tweets)
#to dataframe
tweets.df <- twListToDF(tweets) 
View(tweets.df)
#corpus
mycorpus <-Corpus(VectorSource(tweets.df$text))
#clean the data
mycorpus <-tm_map(mycorpus,tolower)
mycorpus <-tm_map(mycorpus,removeNumbers)
mycorpus <-tm_map(mycorpus,removePunctuation)
mycorpus <-tm_map(mycorpus,stripWhitespace)
mycorpus <-tm_map(mycorpus,removeWords,stopwords())

#Document term matrix
doc_matrix <-DocumentTermMatrix(mycorpus,control = list(tolower=TRUE,
                                                      stopwords=TRUE,
                                                       removePunctuation=TRUE,
                                                       stripWhitespace=TRUE))
#View(doc_matrix[1:10,1:4])
#wordcloud
wordcloud(mycorpus,min.freq = 3,colors = brewer.pal(9,"Set2"),random.order = FALSE,rot.per = 0.90)
#sentiment
nrc_senti <- get_sentiments("nrc")
data <- tibble(1:nrow(tweets.df),text=tweets.df$text)
data <- data %>% unnest_tokens(word,text)
#stop words removal
View(stop_words)
data <- data %>% anti_join(stop_words)
#count each word
data %>% count(word,sort = TRUE)

#for 10 emotions in nrc
emotions <-data %>% inner_join(nrc_senti) %>% group_by(sentiment) %>% count() %>% arrange(n) 
emotions %>% ggplot(aes(x=sentiment,y=n,fill=n))+geom_col()+xlab("Sentiments")+ylab("Count")+theme(axis.text.x = element_text(angle = 90, hjust = 1))
chartJSRadar(emotions)#
bing <- get_sentiments("bing")
bing_analysis <-data %>% inner_join(bing) %>% group_by(sentiment) %>% count() %>% arrange(n)
bing_analysis  %>% ggplot(aes(x=sentiment,y=n,fill=sentiment))+geom_col()+xlab("Sentiment")+ylab("Count")
#typeof(data)
#google map
lat <- tweets.df$latitude
long <- tweets.df$longitude
#Location based
location <- data.frame(x_lat=lat,y_long=long)
#general map
world_basemap <- ggplot() +
  borders("world", colour = "gray85", fill = "gray80")+theme_map()
world_basemap+geom_point(location,aes(x=x_lat,y=y_long),color ="purple",alpha=0.5)