server.R

# Shiny 2.2
library(dplyr)
library(ggplot2)
library("ggthemes")
library(scales)
library(shiny)
library(data.table)
library(plotly)
library(igraph)
library(networkD3)
library(markdown)
library(bizdays)
# To do:
# 1. Remove "Canceled:"
# 2. Provide a text entry box to remove certain events
# 3. Show numbers in % (checkbox option)
# 4. Calculate percentage of work time/given count of work days
# 5. barplot of events by months
# 6. Remove Appointment
# template <- read.csv("./data/df_template.csv", header = TRUE, sep = ",", comment.char="",stringsAsFactors = FALSE)

# To-do:
# .	Exclude meeting invites from specific person (from Quant Event?)
# .	Exclude certain events?
# .	Add "Download Plot" buttons [ggplot]
# . Change plot 1 to refelct input$n


shinyServer(function(input, output, session) {
  #output$instruction <- renderText({return( paste("Instruction to download csv file of your calendar","http://tinnes.co.uk/desktopcalendar/support/7/faq_importexport_exportoutlook2010.php"))})
  
  meet.df <- reactive({
    infile <- input$meet.df
    # if (is.null(infile)) {return(NULL)} # User has not uploaded a file yet
    if (is.null(infile)) {
      df <- read.csv("./data/df_template.csv", header = TRUE, sep = ",", comment.char="",stringsAsFactors = FALSE)
    } else{
      df <- read.csv(infile$datapath, header = TRUE, sep = ",", comment.char="",stringsAsFactors = FALSE) 
      }
    return (df)
  })

  display.df <- reactive({
    #validate(need(input$meet.df != "", "Please upload a data set"))
    display.df <- meet.df()
    # Changing All day event
    display.df$All.day.event <- ifelse(display.df$All.day.event == "True", TRUE, FALSE)
    
    if(input$showallday == TRUE){
      display.df <- subset(display.df, display.df$All.day.event == FALSE & display.df$Required.Attendees != "" & display.df$Meeting.Organizer != "")
      }
    else{display.df <- display.df}
    
    # Cleaning date
    display.df$start <- as.POSIXct(paste(display.df$Start.Date, display.df$Start.Time), format = "%m/%d/%Y %I:%M:%S %p")
    display.df$end   <- as.POSIXct(paste(display.df$End.Date, display.df$End.Time), format = "%m/%d/%Y %I:%M:%S %p")
    display.df$Start.Date <- as.Date.character(display.df$Start.Date, format = "%m/%d/%Y")
    display.df$End.Date <- as.Date.character(display.df$End.Date, format = "%m/%d/%Y")
    min.date <- input$dates[1]
    max.date <- input$dates[2]
    display.df <- subset(display.df, display.df$Start.Date >= min.date & display.df$End.Date <= max.date)
    
    # Cleaning time
    display.df$duration <- as.numeric(display.df$end-display.df$start)
    display.df$hour <- as.numeric(format(display.df$start, "%H")) # unclass((as.POSIXlt(display.df$start)))$hour
    display.df$wday <- lubridate::wday(display.df$start) - 1 # unclass((as.POSIXlt(display.df$start)))$wday #wday(display.df$start)
    
    display.df$Required.Attendees[display.df$Required.Attendees == ""] <- NA
    display.df$Optional.Attendees[display.df$Optional.Attendees == ""] <- NA
    
    display.df$startyr   <- format(display.df$start, "%Y")
    display.df$startmon  <- format(display.df$start, "%m")
    display.df$startyrmon <- format(display.df$start, "%Y-%m")
    display.df$startdate <- format(display.df$start, "%d")
    return(display.df)
  })
  # Columns to show (reactive)
  observe({ 
    x <- colnames(meet.df())
    # Can use character(0) to remove all choices
    if (is.null(x)) x <- character(0)
    col_show <- c("Subject", "Start.Date","Start.Time", "End.Date", "End.Time", "All.day.event", "Meeting.Organizer", "Required.Attendees", "Optional.Attendees")
    updateSelectizeInput(session, inputId = "show_vars",
                         label = "Columns to show:",
                         choices = x,
                         selected = x[x %in% col_show]
                         )
  })
  
  # My name
  people.freq.df <- reactive({
    meet.df <- display.df()
    n <- input$n
    split.func <- function(x){ return (unlist(strsplit(x, split = ";")))}
    split2.func <- function(x){ return (strsplit(x, split = ";"))}
    # To avoid NA as character
    meeting.org <- as.character(na.omit(unlist(sapply(meet.df$Meeting.Organizer, split.func))))
    required.attendee <- as.character(na.omit(unlist(sapply(meet.df$Required.Attendees, split.func))))
    optional.attendee <- as.character(na.omit(unlist(sapply(meet.df$Optional.Attendees, split.func))))
    
    all.attendee <- c(meeting.org, required.attendee, optional.attendee)
    
    people.freq.df <- as.data.frame(table(all.attendee)) %>%
      top_n(n, Freq) %>% #show n rows, will show >n rows if there're ties
      arrange(desc(Freq)) #%>% #show by descending order of Freq
    #rename(Name=all.attendee2) -> people.freq.df
    rot <- 45
    # sort <- "Freq" #or by "Alphabet"
    
    # if (sort == "Freq"){
    people.freq.df$all.attendee <- factor(people.freq.df$all.attendee)
    people.freq.df$all.attendee <- reorder(people.freq.df$all.attendee, -(people.freq.df$Freq))
    return(people.freq.df)
  })
  # My name
  observe({
    df <- people.freq.df()
    x <- df$all.attendee[1]
    # Can use character(0) to remove all choices
    # if (is.null(x)) x <- character(0)
    updateSelectizeInput(session, inputId = "myname",
                         label = "My name is",
                         choices = df$all.attendee,
                         selected = x
    )
  })
  # Date of data to show (reactive datarangeinput) ####
  output$dates <- renderUI({
    df <- meet.df()
    # print(df)
    min <- min(as.Date.character(df$Start.Date, format = "%m/%d/%Y"))
    max <- max(as.Date.character(df$Start.Date, format = "%m/%d/%Y"))
    dateRangeInput('dates', label = "Date input:",
                   start = min, end = max,
                   min = min, max = max)
  })

  output$table <- renderDataTable({
    #validate(need(input$meet.df != "", "Please upload a data set"))
    display.df()[, input$show_vars, drop = FALSE]
  })
  
  #output$date <- renderText({format(input$dateRange[1])})
  # Number of people to show (reactive sliderbar)
  observe({
    df <- display.df()
    n <- length(unique(df$Meeting.Organizer))
    updateSliderInput(session, "n", value = 20, min = 1, max = n)
  })
  
  plot1 <- reactive({
    meet_clean.df <- display.df()
    n <- input$n
    people.freq.df <- as.data.frame(table(meet_clean.df$Meeting.Organizer)) %>%
      top_n(n, Freq) %>% #show n rows, will show >n rows if there're ties
      arrange(desc(Freq)) #%>% #show by descending order of Freq
    # rename(Name=Var1) #-> people.freq.df
    rot <- 45
    # sort <- "Freq"
    
    # if (sort == "Freq"){
    people.freq.df$Var1 <- factor(people.freq.df$Var1)
    people.freq.df$Var1 <- reorder(people.freq.df$Var1, -(people.freq.df$Freq))
    people.freq.df    <- people.freq.df[people.freq.df$Var1 != input$myname,]
    # } else {
    #   people.freq.df$Var1 <- people.freq.df$Var1
    # }
    # 
    ggplot(data=people.freq.df, aes(x=Var1, y=Freq)) +
      geom_bar(stat = "identity") +
      geom_text(aes(label = Freq), vjust = -0.5) +
      labs(title = paste("Top", input$n, "people who initiated a meeting with me from", input$dates[1], "to", input$dates[2]),
            x = "Name") +
      
      theme_fivethirtyeight() +
      theme(axis.text.x = element_text(angle = rot, hjust = 1)) +
      scale_fill_fivethirtyeight() #doesn't 
  })
  output$plot1 <- renderPlot({
    #validate(need(input$meet.df != "", ""))
    return(plot1())
  })
  
  output$downloadPlot1 <- downloadHandler(
    filename = function() {
      paste('peopleplot', '.png', sep='')
    },
    content=function(file){
      png(file)
      print(plot1())
      dev.off()
    },
    contentType='image/png')

  plot2 <- reactive({
    #validate(need(input$meet.df != "", ""))
    meet.df <- display.df()
    n <- input$n
    split.func <- function(x){ return (unlist(strsplit(x, split = ";")))}
    split2.func <- function(x){ return (strsplit(x, split = ";"))}
    # To avoid NA as character
    meeting.org <- as.character(na.omit(unlist(sapply(meet.df$Meeting.Organizer, split.func))))
    required.attendee <- as.character(na.omit(unlist(sapply(meet.df$Required.Attendees, split.func))))
    optional.attendee <- as.character(na.omit(unlist(sapply(meet.df$Optional.Attendees, split.func))))

    all.attendee <- c(meeting.org, required.attendee, optional.attendee)
      
    people.freq.df <- as.data.frame(table(all.attendee)) %>%
      top_n(n, Freq) %>% #show n rows, will show >n rows if there're ties
      arrange(desc(Freq)) #%>% #show by descending order of Freq
      #rename(Name=all.attendee2) -> people.freq.df
    rot <- 45
    # sort <- "Freq" #or by "Alphabet"

    # if (sort == "Freq"){
      people.freq.df$all.attendee <- factor(people.freq.df$all.attendee)
      people.freq.df$all.attendee <- reorder(people.freq.df$all.attendee, -(people.freq.df$Freq))
      people.freq.df <- people.freq.df[people.freq.df$all.attendee != input$myname, ]
    # } else {
    #   people.freq.df$Name <- people.freq.df$Name
    # 
    # return(people.freq.df) #returning for "my name"
      
    ggplot(data=people.freq.df, aes(x=all.attendee, y=Freq)) +
      geom_bar(stat = "identity") +
      geom_text(aes(label = Freq), vjust = -0.5) +
      labs(title = paste("Top", input$n, "people who attended a meeting with me from", input$dates[1], "to", input$dates[2]),
           x = "Name") + 
      theme_fivethirtyeight() +
      theme(axis.text.x = element_text(angle = rot, hjust = 1)) +
      scale_fill_fivethirtyeight()
  })
  output$plot2 <- renderPlot({
    return(plot2())
  })
  
  output$downloadPlot2 <- downloadHandler(
    filename = function() {
      paste('peopleplot2', '.png', sep='')
    },
    content=function(file){
      png(file)
      print(plot2())
      dev.off()
    },
    contentType='image/png')
  
  monthplot <- reactive({
    meet.df <- display.df()
    ## Number of meetings per month (plot) ####
    if(input$sortbyFreq == TRUE){ #sorting by Frequency
      month.df <- as.data.frame(table(meet.df$startyrmon)) %>%
        arrange(desc(Freq)) #sort by Frequency
      month.df[,1] <- factor(month.df[,1])
      month.df[,1] <- reorder(month.df[,1], -(month.df[,2]))
    }
    else{month.df <- as.data.frame(table(meet.df$startyrmon))}
    
    colnames(month.df) <- c("Month", "Freq")
    rot = 45
    
    ggplot(month.df, aes(x=Month, y=Freq))+
      geom_bar(stat = "identity") +
      geom_text(aes(label = Freq), vjust = -0.5) +
      labs(x = "Year-Month", title = "Number of Meetings per Month") +
      geom_hline(yintercept=mean(month.df$Freq), col = "dark grey", linetype = "dashed")+
      geom_text(x=Inf, y=mean(month.df$Freq), label = round(mean(month.df$Freq)), hjust=1, vjust=-0.5)+
      theme_fivethirtyeight() +
      theme(axis.text.x = element_text(angle = rot, hjust = 1)) +
      scale_fill_fivethirtyeight()
  })
  output$monthplot <- renderPlot({
    return(monthplot())
  })
  output$downloadmonthplot <- downloadHandler(
    filename = function() {
      paste('monthplot', '.png', sep='')
    },
    content=function(file){
      png(file)
      print(monthplot())
      dev.off()
    },
    contentType='image/png')
  
  monthtimeplot <- reactive({
    meet.df <- display.df()
    # Duration of meetings per month (plot) ####
    if(input$sortbyFreq == TRUE){
    time.month.df <- as.data.frame(aggregate(meet.df$duration, by = list(meet.df$startyrmon), mean)) %>%
        arrange(desc(x)) #sort by Frequency
    colnames(time.month.df) <- c("Month", "Duration")
    time.month.df$Month <- factor(time.month.df$Month)
    time.month.df$Month <- reorder(time.month.df$Month, -(time.month.df$Duration))   
    } else{
    time.month.df <- as.data.frame(aggregate(meet.df$duration, by = list(meet.df$startyrmon), mean))
    colnames(time.month.df) <- c("Month", "Duration") 
    }
    rot <- 45
    
    ggplot(time.month.df, aes(x=Month, y=Duration)) +
      geom_bar(stat = "identity") +
      geom_text(aes(label = round(Duration)), vjust = -0.5) +
      labs(x = "Year-Month", y = "Duration (Minutes)", title = "Average Duration (Minutes) Per Month") +
      geom_hline(yintercept=mean(time.month.df$Duration), col = "dark grey", linetype = "dashed")+
      geom_text(x=Inf, y=mean(time.month.df$Duration), label = round(mean(time.month.df$Duration)), hjust=1, vjust=-0.5)+
      theme_fivethirtyeight() +
      theme(axis.text.x = element_text(angle = rot, hjust = 1)) +
      scale_fill_fivethirtyeight()
  })
  output$monthtimeplot <- renderPlot({
    return(monthtimeplot())
  })
  
  output$downloadmonthtimeplot <- downloadHandler(
    filename = function() {
      paste('monthtimeplot', '.png', sep='')
    },
    content=function(file){
      png(file)
      print(monthtimeplot())
      dev.off()
    },
    contentType='image/png')
  
  output$timeplot1 <- renderPlotly({
    temp <- display.df()
    temp <- subset(temp, temp$wday >= 1 & temp$wday <= 5)
    time.df <- as.data.frame(table(temp$wday, temp$hour))
    colnames(time.df) <- c("wday", "hour", "Freq")
    time.df$hour <- factor(time.df$hour, levels = levels(time.df$hour)[nlevels(time.df$hour):1])
    time.df$desc <- character(nrow(time.df))
    
    for (i in 1:nlevels(time.df$wday)){
      for (j in 1:nlevels(time.df$hour)){
        day <- unique(time.df$wday)[i]
        hour <- unique(time.df$hour)[j]
        
        index <- which(temp$wday == day & temp$hour == hour)
        desc <- as.character(names(sort(table(temp$Subject[index]), decreasing  = TRUE)))
        
        index2 <- as.numeric(which(time.df$wday == day & time.df$hour == hour))
        time.df$desc[index2] <- desc[1] #show top 3 meeting subject
      }
    }
    if (input$showname == TRUE){
      g <- ggplot(time.df, aes(x = wday, y = hour, fill = Freq, text = paste("Event:",desc)))
    }else{ g <- ggplot(time.df, aes(x = wday, y = hour, fill = Freq))}
    # g <- ggplot(time.df, aes(x = wday, y = hour, fill = Freq, text = paste("Event:",desc)))
      g <- g + 
         geom_tile(color="white", size=0.1) +
        scale_fill_gradient(low = "white", high = "red") + 
        labs(x="Day", y="Hour", title="Meetings per weekday & time of day")
        # theme_tufte(base_family="Helvetica")

    return(ggplotly(g))
    
  })
  
  output$netplot1 <- renderPlot({
    netplot.df <- display.df()
    netplot.df$all <- paste(netplot.df$Meeting.Organizer, netplot.df$Required.Attendees, sep = ";")
    split2.func <- function(x){ return (strsplit(x, split = ";"))}
    netplot.df$all2 <- sapply(netplot.df$all, split2.func)
    temp <- unlist(netplot.df$all2) #everyone's name
    temp <- temp[temp != ""] #manual clean up
    n <- as.numeric(length(unique(temp)))
    
    people.df <- data.frame(matrix(0, ncol = n, nrow = 1))
    colnames(people.df) <- unique(temp)
    for (i in 1:ncol(people.df)){
      for (j in 1:nrow(netplot.df)){
        people.df[j,i] <- grepl(colnames(people.df)[i], netplot.df$all2[j]) 
      }
    }
    data <- people.df
    
    total_occurrences <- colSums(data)
    index <- total_occurrences >= sort(total_occurrences, decreasing = TRUE)[input$net_n] #median(total_occurrences)
    data <- subset(data, select = index == TRUE)
    total_occurrences <- colSums(data)
    
    data_matrix <- as.matrix(data)
    co_occurrence <- t(data_matrix) %*% data_matrix
    graph <- graph.adjacency(co_occurrence,
                             weighted=TRUE,
                             mode="undirected",
                             diag=FALSE)
    if(input$random_net == FALSE) {set.seed(33)}
    return(plot(graph,
         #vertex.label=names(data),
         vertex.size=total_occurrences*input$node_size/500,
         edge.width=E(graph)$weight*input$net_width/1000))
  })
  
  output$netplot2 <- renderForceNetwork({
    netplot.df <- display.df()
    netplot.df$all <- paste(netplot.df$Meeting.Organizer, netplot.df$Required.Attendees, sep = ";")
    split2.func <- function(x){ return (strsplit(x, split = ";"))}
    netplot.df$all2 <- sapply(netplot.df$all, split2.func)
    temp <- unlist(netplot.df$all2) #everyone's name
    temp <- temp[temp != ""] #manual clean up
    n <- as.numeric(length(unique(temp)))
    
    people.df <- data.frame(matrix(0, ncol = n, nrow = 1))
    colnames(people.df) <- unique(temp)
    for (i in 1:ncol(people.df)){
      for (j in 1:nrow(netplot.df)){
        people.df[j,i] <- grepl(colnames(people.df)[i], netplot.df$all2[j]) 
      }
    }
    data <- people.df
    
    total_occurrences <- colSums(data)
    index <- total_occurrences >= sort(total_occurrences, decreasing = TRUE)[input$net_n] #median(total_occurrences)
    data <- subset(data, select = index == TRUE)
    total_occurrences <- colSums(data)
    
    data_matrix <- as.matrix(data)
    co_occurrence <- t(data_matrix) %*% data_matrix
    graph <- graph.adjacency(co_occurrence,
                             weighted=TRUE,
                             mode="undirected",
                             diag=FALSE)
    wc <- cluster_walktrap(graph)
    members <- membership(wc)
    graph_in <- igraph_to_networkD3(graph, group = members)
    
    return(forceNetwork(Links = graph_in$links, Nodes = graph_in$nodes, 
                 Source = 'source', Target = 'target', 
                 NodeID = 'name', Group = 'group', zoom = TRUE))
  })
  
  dateplot <- reactive({
    meet.df <- display.df()
    ## Number of meetings per month (plot) ####
    if(input$sortbyFreq == TRUE){ #sorting by Frequency
      date.df <- as.data.frame(table(meet.df$startdate)) %>%
        arrange(desc(Freq)) #sort by Frequency
      date.df[,1] <- factor(date.df[,1])
      date.df[,1] <- reorder(date.df[,1], -(date.df[,2]))
    }
    else{date.df <- as.data.frame(table(meet.df$startdate))}
    
    colnames(date.df) <- c("Date", "Freq")
    rot = 0
    
    ggplot(date.df, aes(x=Date, y=Freq))+
      geom_bar(stat = "identity") +
      geom_text(aes(label = Freq), vjust = -0.5) +
      labs(x = "Date", title = "Number of Meetings per Date") +
      geom_hline(yintercept=mean(date.df$Freq), col = "dark grey", linetype = "dashed")+
      geom_text(x=Inf, y=mean(date.df$Freq), label = round(mean(date.df$Freq)), hjust=1, vjust=-0.5)+
      theme_fivethirtyeight() +
      theme(axis.text.x = element_text(angle = rot, hjust = 1)) +
      scale_fill_fivethirtyeight()
  })
  output$dateplot <- renderPlot({
    return(dateplot())
  })
  output$downloaddateplot <- downloadHandler(
    filename = function() {
      paste('dateplot', '.png', sep='')
    },
    content=function(file){
      png(file)
      print(monthplot())
      dev.off()
    },
    contentType='image/png')
  output$introdateplot <- renderText({
    "Do you think if you had more meetings on a particular date?"
  })
  output$summarydateplot <- renderText({
    "While some dates are by design less likely to have meetings, because they are less likely to exist or to be business days (e.g., 31st or 4th for the U.S. holidays), 
    it could still be interesting to see the differences."})
  output$summary1 <- renderText({
    paste("From", input$dates[1], "to", input$dates[2],  ", I had", nrow(display.df()), "meetings in total of", sum(display.df()$duration) ,"hours.")})
  # bizdaycount <- reactive({
  #   if(is.null(input$bizdaycount)){
  #     bizdaycount <- bizdays(input$dates[1], input$dates[2], cal = cal) + 1  
  #   } else{
  #     bizdaycount <- numericInput(inputId, label, value,
  #                                     min, max, step)
  #   }
  #   return(bizdaycount)
  # })

  # observe({
  #   cal <- create.calendar("Actual", weekdays=c("saturday", "sunday"))
  #   # x <- bizdays(as.Date(input$dates[1], format = "%Y-%m-%d"), as.Date(input$dates[2], format = "%Y-%m-%d"), cal = cal) + 1
  #   # Can use character(0) to remove all choices
  #   # if (is.null(x)) x <- numeric(0)
  #   updateNumericInput(session, inputId = "bizdaycount",
  #                        label = "Work Hours per Day:",
  #                         value = bizdays(as.Date(input$dates[1], format = "%Y-%m-%d"), as.Date(input$dates[2], format = "%Y-%m-%d"), cal = cal) + 1,
  #                         min = 1,
  #                         max = as.numeric(as.Date(input$dates[2], format = "%Y-%m-%d") - as.Date(input$dates[1], format = "%Y-%m-%d")),
  #                        step=1
  #   )
  # })
  output$summary2 <- renderText({
    cal <- create.calendar("Actual", weekdays=c("saturday", "sunday"))
    bizdaycount <- bizdays(as.Date(input$dates[1], format = "%Y-%m-%d"), as.Date(input$dates[2], format = "%Y-%m-%d"), cal = cal) + 1
    workhours <- input$workhours
    paste0("Out of ", bizdaycount  ," business days (roughly estimated), on average, ", round(sum(display.df()$duration)/(bizdaycount*workhours)),"% of my work time was spent on meetings." )
  })
  # output$summary2 <- renderText({
  #   meet.df <- display.df()
  #   n <- input$n
  #   split.func <- function(x){ return (unlist(strsplit(x, split = ";")))}
  #   split2.func <- function(x){ return (strsplit(x, split = ";"))}
  #   # To avoid NA as character
  #   meeting.org <- as.character(na.omit(unlist(sapply(meet.df$Meeting.Organizer, split.func))))
  #   required.attendee <- as.character(na.omit(unlist(sapply(meet.df$Required.Attendees, split.func))))
  #   optional.attendee <- as.character(na.omit(unlist(sapply(meet.df$Optional.Attendees, split.func))))
  #   
  #   all.attendee <- c(meeting.org, required.attendee)
  #   
  #   people.freq.df <- as.data.frame(table(all.attendee)) %>%
  #     top_n(n, Freq) %>% #show n rows, will show >n rows if there're ties
  #     arrange(desc(Freq)) #%>% #show by descending order of Freq
  # 
  #   rot <- 45
  #   # sort <- "Freq" #or by "Alphabet"
  #   
  #   # if (sort == "Freq"){
  #   people.freq.df$all.attendee <- factor(people.freq.df$all.attendee)
  #   people.freq.df$all.attendee <- reorder(people.freq.df$all.attendee, -(people.freq.df$Freq))
  #   return(paste0("Most of them are with ", people.freq.df$all.attendee[2], "."))
  #   })
  
  output$summary3 <- renderText({
    meet.df <- display.df()
    n <- input$n
    split.func <- function(x){ return (unlist(strsplit(x, split = ";")))}
    split2.func <- function(x){ return (strsplit(x, split = ";"))}
    # To avoid NA as character
    meeting.org <- as.character(na.omit(unlist(sapply(meet.df$Meeting.Organizer, split.func))))
    required.attendee <- as.character(na.omit(unlist(sapply(meet.df$Required.Attendees, split.func))))
    optional.attendee <- as.character(na.omit(unlist(sapply(meet.df$Optional.Attendees, split.func))))
    
    all.attendee <- c(meeting.org, required.attendee)
    
    people.freq.df <- as.data.frame(table(all.attendee)) %>%
      top_n(n, Freq) %>% #show n rows, will show >n rows if there're ties
      arrange(desc(Freq)) #%>% #show by descending order of Freq
    #rename(Name=all.attendee2) -> people.freq.df
    rot <- 45
    # sort <- "Freq" #or by "Alphabet"
    
    # if (sort == "Freq"){
    people.freq.df$all.attendee <- factor(people.freq.df$all.attendee)
    return(paste0(round(people.freq.df$Freq[2]/nrow(display.df())* 100), "% of my meetings are with ", people.freq.df$all.attendee[2]))
  })
  
  
})