Shiny code for SegmentViewer Simulation

LICENSE

# The MIT License (MIT)
#
# Copyright (c) 2012 Schaun Jacob Wheeler
#
# Permission is hereby granted, free of charge, to any person obtaining a copy
# of this software and associated documentation files (the "Software"), to deal
# in the Software without restriction, including without limitation the rights
# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
# copies of the Software, and to permit persons to whom the Software is
# furnished to do so, subject to the following conditions:
#
# The above copyright notice and this permission notice shall be included in all
# copies or substantial portions of the Software.
#
# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
# SOFTWARE.

server.R

library(shiny)
library(plyr)
library(scales)
library(ggplot2)
library(reshape2)

set.seed(12435)

state <- sample(c("CA", "WA", "OR", "AZ", "NV", "UT"), 40000, replace = T)
lowgrade <- sample(c(0,1), 40000, replace = T)
lowgrade[sample(40000, 10000)] <- sample(c(5,7), 10000, replace = T)
lowgrade[sample(40000, 7000)] <- 9

highgrade <- sample(5:6, 40000, replace = T)
highgrade[lowgrade == 5] <- sample(c(8,12), sum(lowgrade == 5), replace = T)
highgrade[lowgrade == 7] <- 9
highgrade[lowgrade == 9] <- 12

income <- sample(1:10, 40000, replace = T) * 10000
revenue <- c(rnorm(20000, mean = 2000, sd = 2000),
             rnorm(10000, mean = 2000, sd = 1000),
             rnorm(10000, mean = 2000, sd = 4000))
revenue[revenue < 0] <- 0
insert_zeros <- sample(1:sum(revenue != 0),sum(revenue != 0)/1.2)
revenue[insert_zeros] <- 0
revenue <- sample(revenue, 40000)

schools <- data.frame(state, 
                      lowgrade, 
                      highgrade, 
                      income, 
                      revenue,
                      stringsAsFactors = F)

sch_tab <- ddply(schools, 
                 .variables = c("state", "lowgrade", "highgrade", "income"),
                 function(df){
                   c("n" = nrow(df), 
                     "sum" = sum(df$revenue),
                     "mean" = mean(df$revenue))
                 })

reg_line <- predict(lm(sum~mean, data = sch_tab), 
                    newdata = data.frame("mean" = sch_tab$mean))

sch_tab$segment <- rep("More Profitable\nMore Difficult", nrow(sch_tab))
sch_tab$segment2 <- rep("Less Profitable\nMore Difficult", nrow(sch_tab))

sch_tab$segment[sch_tab$sum <= reg_line &
                  sch_tab$mean < median(sch_tab$mean)] <- "Less Profitable\nMore Difficult"
sch_tab$segment[sch_tab$sum > reg_line &
                  sch_tab$mean >= median(sch_tab$mean)] <- "More Profitable\nLess Difficult"
sch_tab$segment[sch_tab$sum <= reg_line &
                  sch_tab$mean >= median(sch_tab$mean)] <- "Less Profitable\nLess Difficult"

sch_tab$segment <- factor(sch_tab$segment, 
                          levels = c("More Profitable\nLess Difficult", 
                                     "More Profitable\nMore Difficult",
                                     "Less Profitable\nLess Difficult",
                                     "Less Profitable\nMore Difficult"))

refs <- data.frame("labels" = c(paste("State:",sort(unique(schools$state))),
                                paste("Low grade:", sort(unique(schools$lowgrade))),
                                paste("High grade:", sort(unique(schools$highgrade))),
                                paste("Income:", dollar(sort(unique(schools$income))))),
                   "variables" = c(rep("state",length(unique(schools$state))),
                                   rep("lowgrade", length(unique(schools$lowgrade))),
                                   rep("highgrade", length(unique(schools$highgrade))),
                                   rep("income", length(unique(schools$income)))),
                   "values" = c(sort(unique(schools$state)),
                                sort(unique(schools$lowgrade)),
                                sort(unique(schools$highgrade)),
                                sort(unique(schools$income))),
                   stringsAsFactors = FALSE) 

# Define server logic required to plot various microsegments
shinyServer(function(input, output) {
  
  sch_table <- reactive(function() {
    
    df <- sch_tab
    
    reward2 <- refs[refs$labels %in% input$reward,]
    risk2 <- refs[refs$labels %in% input$risk,]
    
    
    df$segment2[(df$state %in% reward2[reward2$variables == "state", "values"] |
                   df$lowgrade %in% reward2[reward2$variables == "lowgrade", "values"] |
                   df$highgrade %in% reward2[reward2$variables == "highgrade", "values"] &
                   df$income %in% reward2[reward2$variables == "income", "values"]) &
                  (df$state %in% risk2[risk2$variables == "state", "values"] |
                     df$lowgrade %in% risk2[risk2$variables == "lowgrade", "values"] |
                     df$highgrade %in% risk2[risk2$variables == "highgrade", "values"] |
                     df$income %in% risk2[risk2$variables == "income", "values"])] <- "More Profitable\nLess Difficult"
    
    df$segment2[(df$state %in% reward2[reward2$variables == "state", "values"] |
                   df$lowgrade %in% reward2[reward2$variables == "lowgrade", "values"] |
                   df$highgrade %in% reward2[reward2$variables == "highgrade", "values"] |
                   df$income %in% reward2[reward2$variables == "income", "values"]) &
                  !(df$state %in% risk2[risk2$variables == "state", "values"] |
                      df$lowgrade %in% risk2[risk2$variables == "lowgrade", "values"] |
                      df$highgrade %in% risk2[risk2$variables == "highgrade", "values"] |
                      df$income %in% risk2[risk2$variables == "income", "values"])] <- "More Profitable\nMore Difficult"
    
    df$segment2[!(df$state %in% reward2[reward2$variables == "state", "values"] |
                    df$lowgrade %in% reward2[reward2$variables == "lowgrade", "values"] |
                    df$highgrade %in% reward2[reward2$variables == "highgrade", "values"] |
                    df$income %in% reward2[reward2$variables == "income", "values"]) &
                  (df$state %in% risk2[risk2$variables == "state", "values"] |
                     df$lowgrade %in% risk2[risk2$variables == "lowgrade", "values"] |
                     df$highgrade %in% risk2[risk2$variables == "highgrade", "values"] |
                     df$income %in% risk2[risk2$variables == "income", "values"])] <- "Less Profitable\nLess Difficult"
    
    df
    
  })
  
  output$segplot <- reactivePlot(function() {
    df <- sch_table()
    
    df <- melt(df, id.vars = c("state", "lowgrade", "highgrade", "income", "n", 
                               "sum", "mean"), variable.name = "Version",
               value.name = "Segment")
    df$Version <- as.character(df$Version)
    df$Version[df$Version == "segment"] <- "Optimized"
    df$Version[df$Version == "segment2"] <- "Customized"
    df$Version <- factor(df$Version, levels = c("Optimized", "Customized"))
    df$Segment <- factor(df$Segment, 
                         levels = c("More Profitable\nLess Difficult", 
                                    "More Profitable\nMore Difficult",
                                    "Less Profitable\nLess Difficult",
                                    "Less Profitable\nMore Difficult"))
    
    out <- ggplot(df, aes(x = mean, y = sum, color = Segment)) + 
      geom_point(size=2) +   
      scale_color_manual(values = c("blue", "green", "red", "black")) + 
      scale_x_continuous("Avg. Revenue (Ease of Sale)", labels = dollar) + 
      scale_y_continuous("Tot. Revenue (Profitability)", labels = dollar) + 
      facet_grid(Version~.) +
      ggtitle("Segmentation Comparison:\nStatistically Optimized vs. Individually Customized Solutions") +
      theme_bw()
    
    print(out)
  })
  
  output$segtable1 <- reactiveTable(function() {
    df <- sch_table()
    df$Optimized <- df$segment
    ddply(df, 
          .variables = "Optimized",
          function(df){
            c("Tot. Revenue" = dollar_format(0)(sum(df$sum)),
              "Avg. Revenue" = dollar_format(0)(sum(df$mean)),
              "# Schools" = comma(sum(df$n)))
          })
  })
  
  output$segtable2 <- reactiveTable(function() {
    df <- sch_table()
    df$Customized <- df$segment2
    ddply(df, 
          .variables = "Customized",
          function(df){
            c("Tot. Revenue" = dollar_format(0)(sum(df$sum)),
              "Avg. Revenue" = dollar_format(0)(sum(df$mean)),
              "# Schools" = comma(sum(df$n)))
          })	
  })
})

ui.R

library(shiny)
library(scales)

state <- c("CA", "WA", "OR", "AZ", "NV", "UT")
lowgrade <- c(0, 1, 5, 7, 9)
highgrade <- c(5,6,8,9,12)
income <- (1:10) * 10000

shinyUI(pageWithSidebar(
  
  # Application title
  headerPanel("Microsegment Explorer"),
  
  sidebarPanel(
    checkboxGroupInput("reward", "Select microsegments you expect to be profitable:",
                       c(paste("State:",sort(unique(state))),
                         paste("Low grade:", sort(unique(lowgrade))),
                         paste("High grade:", sort(unique(highgrade))),
                         paste("Income:", dollar(sort(unique(income)))))),
    checkboxGroupInput("risk", "Select microsegments you expect to be easy to sell:",
                       c(paste("State:",sort(unique(state))),
                         paste("Low grade:", sort(unique(lowgrade))),
                         paste("High grade:", sort(unique(highgrade))),
                         paste("Income:", dollar(sort(unique(income))))))
  ),
  
  mainPanel(
    plotOutput("segplot", width = "80%", height = "500px"),
    tableOutput("segtable1"),
    tableOutput("segtable2")
  )
))