weekly_twitter_report.Rmd

---
title: "Twitter Topics of the Week"
author: 
- "Office of Decision Support"
- "University of Nevada, Las Vegas"
header-includes:
- \usepackage[utf8]{inputenc}
- \usepackage[T1]{fontenc}
- \usepackage{textcomp}
- \usepackage{float}
- \usepackage{longtable}
date: "`r format(Sys.Date(),'%B %d, %Y')`"
output: 
  pdf_document:
    latex_engine: xelatex
---

```{r setup, include = FALSE}
setwd("~/Documents/GitHub/Twitter-Sentiment-Analysis/")

library(tidyr)
library(dplyr)
library(xtable)
library(tm)
library(tidytext)
library(ggplot2)
library(openxlsx)

source("02 - Dictionary.R")
```

```{r Clean Up Data, echo = FALSE, warning = FALSE, message = FALSE}
df <- read.xlsx(paste0("data_",Sys.Date(),".xlsx"), stringsAsFactors = FALSE)

df <- df %>%
      # separate(X, c("word", "hash"), "_") %>%
      # separate(hash, c("hash", "num"), "[.]") %>%
      separate(X, c("word", "num"), "[.]") %>%
      select(-num) %>%
      filter(word != "University of Nevada") 

data("stop_words")
custom_stop_words <- data.frame(word = c("i’m", "bc", "https", "t.co", "im", "don’t", 1:999,
                                         "barstool__unlv", "rt", "unlv",
                                         "charliekirk11", "donaldtrumpjr", "kimguilfoyle", "drew", "realcan",
                                         "las", "vegas", "amp", "2019", "tonyposnanski",
                                         "aijyz4gbe8", "nvyoome",
                                         dict$LW, dict$LHash))
```

```{r, echo = FALSE, message = FALSE, results = 'asis'}
avg_score <- round(mean(df$sentScore), 2)
```
There were `r nrow(df) ` total tweets containing 'UNLV' this week.
Average sentiment score of the week was `r avg_score`.

```{r Sentiment Frequencies, echo = FALSE, warning = FALSE, message = FALSE, results = 'asis'}
df %>%
  group_by(sentType) %>%
  summarise(Total = n()) %>%
ggplot(aes(x = sentType, y = Total, fill = sentType)) +
  geom_col(stat = "identity") +
  geom_text(aes(label = Total,
                y = Total + 5))
```            

\section{Most Frequently Used Words}
```{r Most Frequently Used Words, echo = FALSE, warning = FALSE, message = FALSE, results = 'asis'}
tweet_corpus <- Corpus(VectorSource(as.vector(df$text))) 
tidy_tweets <- df %>%
  select(created, text) %>%
  unnest_tokens("word", text)

freq_words <- tidy_tweets %>%
              anti_join(stop_words) %>%
              anti_join(custom_stop_words) %>% 
              count(word) %>%
              arrange(desc(n)) %>%
              head(n = 10)

print(xtable(freq_words,
             align = c("l", "p{5cm}", rep("p{2.25cm}", ncol(freq_words)-1))),
      table.placement = "H",
      include.rownames = TRUE,
      comment = FALSE)
```

\newpage
\section{Most Retweeted of the Week}
```{r Most Retweets, echo = FALSE, warning = FALSE, message = FALSE, results = 'asis'}
retweets <- df %>%
     filter(retweetCount > 0) %>%
     filter(isRetweet == FALSE) %>%
     filter(is.na(replyToSN)) %>%
     arrange(desc(retweetCount)) %>%
     select(text, retweetCount, favoriteCount, sentType, sentScore) %>%
     head(n = 10)

print(xtable(retweets,
             align = c("l", "p{7cm}", rep("p{1.75cm}", ncol(retweets)-1))),        
      table.placement = "H",
      include.rownames = TRUE,
      hline.after = c(-1:rep(nrow(retweets))),
      comment = FALSE)
```

\newpage
\section{Most Favorited of the Week}
```{r Most Favorites, echo = FALSE, warning = FALSE, message = FALSE, results = 'asis'}
favorites <- df %>%
     filter(favoriteCount > 0) %>%
     filter(isRetweet == FALSE) %>%
     filter(is.na(replyToSN)) %>%
     arrange(desc(favoriteCount)) %>%
     select(text, retweetCount, favoriteCount, sentType, sentScore) %>%
     head(n = 10)

print(xtable(favorites,
             align = c("l", "p{7cm}", rep("p{1.75cm}", ncol(favorites)-1))),
      table.placement = "H",
      include.rownames = TRUE,
      hline.after = c(-1:rep(nrow(favorites))),
      comment = FALSE)
```

\newpage
\section{Most Negative of the Week}
```{r Most Negative, echo = FALSE, warning = FALSE, message = FALSE, results = 'asis'}
negatives <- df %>%
     filter(isRetweet == FALSE) %>%
     filter(is.na(replyToSN)) %>%
     arrange(sentScore) %>%
     select(text, retweetCount, favoriteCount, sentType, sentScore) %>%
     unique() %>%
     head(n = 10)

print(xtable(negatives,
             align = c("l", "p{7cm}", rep("p{1.75cm}", ncol(negatives)-1))),
      table.placement = "H",
      include.rownames = TRUE,
      hline.after = c(-1:rep(nrow(negatives))),
      comment = FALSE)
```

\newpage
\section{Most Positive of the Week}
```{r Most Positive, echo = FALSE, warning = FALSE, message = FALSE, results = 'asis'}
positives <- df %>%
     filter(isRetweet == FALSE) %>%
     filter(is.na(replyToSN)) %>%
     arrange(desc(sentScore)) %>%
     select(text, retweetCount, favoriteCount, sentType, sentScore) %>%
     unique() %>%
     head(n = 10)

print(xtable(positives,
             align = c("l", "p{7cm}", rep("p{1.75cm}", ncol(positives)-1))),
      table.placement = "H",
      include.rownames = TRUE,
      hline.after = c(-1:rep(nrow(positives))),
      comment = FALSE)
```

\newpage
\section{Tweets About Hot Topics}
```{r, echo = FALSE, warning = FALSE, message = FALSE, results = 'asis'}
threat <- df %>%
     filter(isRetweet == FALSE) %>%
     filter(is.na(replyToSN)) %>%
     arrange(sentScore) %>%
     select(text, retweetCount, favoriteCount, sentType, sentScore) %>%
     unique()

threat <- as.data.frame(df$text[grep("threat[s]", tolower(threat$text))])
names(threat) <- "Threat"
print(xtable(threat,
             align = c("l", "p{15cm}")),
      table.placement = "H",
      include.rownames = TRUE,
      hline.after = c(-1:rep(nrow(threat))),
      comment = FALSE,
      tabular.environment = "longtable")
```

\newpage
```{r Hot Topics, echo = FALSE, warning = FALSE, message = FALSE, results = 'asis'}
df <- filter(df, isRetweet == FALSE)

financial_aid <- as.data.frame(df$text[grep("financial aid", tolower(df$text))])
names(financial_aid) <- "Financial Aid"
print(xtable(financial_aid,
             align = c("l", "p{15cm}")),
      table.placement = "H",
      include.rownames = TRUE,
      hline.after = c(-1:rep(nrow(financial_aid))),
      comment = FALSE,
      tabular.environment = "longtable")
```

\newpage
```{r, echo = FALSE, warning = FALSE, message = FALSE, results = 'asis'}
money <- as.data.frame(df$text[grep("money", tolower(df$text))])
names(money) <- "Money"
print(xtable(money,
             align = c("l", "p{15cm}")),
      table.placement = "H",
      include.rownames = TRUE,
      hline.after = c(-1:rep(nrow(money))),
      comment = FALSE,
      tabular.environment = "longtable")
```

\newpage
```{r, echo = FALSE, warning = FALSE, message = FALSE, results = 'asis'}
parking <- as.data.frame(df$text[grep("parking", tolower(df$text))])
names(parking) <- "Parking"
print(xtable(parking,
             align = c("l", "p{15cm}")) ,
      table.placement = "H",
      include.rownames = TRUE,
      hline.after = c(-1:rep(nrow(parking))),
      comment = FALSE,
      tabular.environment = "longtable")
```

\newpage
```{r, echo = FALSE, warning = FALSE, message = FALSE, results = 'asis'}
traffic <- as.data.frame(df$text[grep("traffic", tolower(df$text))])
names(traffic) <- "Traffic"
print(xtable(traffic,
             align = c("l", "p{15cm}")) ,
      table.placement = "H",
      include.rownames = TRUE,
      hline.after = c(-1:rep(nrow(traffic))),
      comment = FALSE,
      tabular.environment = "longtable")
```

\newpage
```{r, echo = FALSE, warning = FALSE, message = FALSE, results = 'asis'}
wifi <- as.data.frame(df$text[grep("wifi", tolower(df$text))])
names(wifi) <- "Wifi"
print(xtable(wifi,
             align = c("l", "p{15cm}")) ,
      table.placement = "H",
      include.rownames = TRUE,
      hline.after = c(-1:rep(nrow(wifi))),
      comment = FALSE,
      tabular.environment = "longtable")
```