InteractiveWordCloud.Rmd

---
title: 'Tutorial (week 2): Interactive Word Cloud'
runtime: shiny
output:
  html_document: default
  html_notebook: default
---

This is an *Interactive* [R Markdown](http://rmarkdown.rstudio.com) Notebook. It generates an HTML notebook that would allow users to interactively explore your analysis results. 

We will use the presidential inaugural speech word clouds as examples. 

# Step 0 - Install and load libraries
```{r, message=FALSE, warning=FALSE}
packages.used=c("tm", "wordcloud", "RColorBrewer", 
                "dplyr", "tidytext")

# check packages that need to be installed.
packages.needed=setdiff(packages.used, 
                        intersect(installed.packages()[,1], 
                                  packages.used))
# install additional packages
if(length(packages.needed)>0){
  install.packages(packages.needed, dependencies = TRUE,
                   repos='http://cran.us.r-project.org')
}

library(tm)
library(wordcloud)
library(RColorBrewer)
library(dplyr)
library(tidytext)
```

This notebook was prepared with the following environmental settings.

```{r}
print(R.version)
```

# Step 1 - Read in the speeches
```{r}
folder.path="../data/inaugurals/"
speeches=list.files(path = folder.path, pattern = "*.txt")
prex.out=substr(speeches, 6, nchar(speeches)-4)

ff.all<-Corpus(DirSource(folder.path))
```

# Step 2 - Text processing

See [Basic Text Mining in R](https://rstudio-pubs-static.s3.amazonaws.com/132792_864e3813b0ec47cb95c7e1e2e2ad83e7.html) for a more comprehensive discussion. 

For the speeches, we remove extra white space, convert all letters to the lower case, remove [stop words](https://github.com/arc12/Text-Mining-Weak-Signals/wiki/Standard-set-of-english-stopwords), remove empty words due to formatting errors, and remove punctuation. Then we compute the [Document-Term Matrix (DTM)](https://en.wikipedia.org/wiki/Document-term_matrix). 

```{r}
ff.all<-tm_map(ff.all, stripWhitespace)
ff.all<-tm_map(ff.all, content_transformer(tolower))
ff.all<-tm_map(ff.all, removeWords, stopwords("english"))
ff.all<-tm_map(ff.all, removeWords, character(0))
ff.all<-tm_map(ff.all, removePunctuation)

tdm.all<-TermDocumentMatrix(ff.all)

tdm.tidy=tidy(tdm.all)

tdm.overall=summarise(group_by(tdm.tidy, term), sum(count))
```

# Step 3 - Inspect an overall wordcloud
```{r, fig.height=6, fig.width=6}
wordcloud(tdm.overall$term, tdm.overall$`sum(count)`,
          scale=c(5,0.5),
          max.words=100,
          min.freq=1,
          random.order=FALSE,
          rot.per=0.3,
          use.r.layout=T,
          random.color=FALSE,
          colors=brewer.pal(9,"Blues"))
```

# Step 4 - compute TF-IDF weighted document-term matrices for individual speeches. 
As we would like to identify interesting words for each inaugural speech, we use [TF-IDF](https://en.wikipedia.org/wiki/Tf%E2%80%93idf) to weigh each term within each speech. It highlights terms that are more specific for a particular speech. 

```{r}
dtm <- DocumentTermMatrix(ff.all,
                          control = list(weighting = function(x)
                                             weightTfIdf(x, 
                                                         normalize =FALSE),
                                         stopwords = TRUE))
ff.dtm=tidy(dtm)
```

# Step 5- Interactive visualize important words in individual speeches
```{r, warning=FALSE}
library(shiny)

shinyApp(
    ui = fluidPage(
      fluidRow(style = "padding-bottom: 20px;",
        column(4, selectInput('speech1', 'Speech 1',
                              speeches,
                              selected=speeches[5])),
        column(4, selectInput('speech2', 'Speech 2', speeches,
                              selected=speeches[9])),
        column(4, sliderInput('nwords', 'Number of words', 3,
                               min = 20, max = 200, value=100, step = 20))
      ),
      fluidRow(
        plotOutput('wordclouds', height = "400px")
      )
    ),

    server = function(input, output, session) {

      # Combine the selected variables into a new data frame
      selectedData <- reactive({
        list(dtm.term1=ff.dtm$term[ff.dtm$document==as.character(input$speech1)],
             dtm.count1=ff.dtm$count[ff.dtm$document==as.character(input$speech1)],
             dtm.term2=ff.dtm$term[ff.dtm$document==as.character(input$speech2)],
             dtm.count2=ff.dtm$count[ff.dtm$document==as.character(input$speech2)])
      })

      output$wordclouds <- renderPlot(height = 400, {
        par(mfrow=c(1,2), mar = c(0, 0, 3, 0))
        wordcloud(selectedData()$dtm.term1, 
                  selectedData()$dtm.count1,
              scale=c(4,0.5),
              max.words=input$nwords,
              min.freq=1,
              random.order=FALSE,
              rot.per=0,
              use.r.layout=FALSE,
              random.color=FALSE,
              colors=brewer.pal(10,"Blues"), 
            main=input$speech1)
        wordcloud(selectedData()$dtm.term2, 
                  selectedData()$dtm.count2,
              scale=c(4,0.5),
              max.words=input$nwords,
              min.freq=1,
              random.order=FALSE,
              rot.per=0,
              use.r.layout=FALSE,
              random.color=FALSE,
              colors=brewer.pal(10,"Blues"), 
            main=input$speech2)
      })
    },

    options = list(height = 600)
)
```




# Further readings

+ [Text mining with `tidytext`](http://tidytextmining.com/).
+ [Basic Text Mining in R](https://rstudio-pubs-static.s3.amazonaws.com/132792_864e3813b0ec47cb95c7e1e2e2ad83e7.html)