ggvis visualization does not appear in main pane

I'm new to shiny and this one has been giving me such a difficult. I tried several suggestions I found to the last reactive but none worked. I am not sure what I am doing wrong.

I tried vis <- reactive({}) and vis %>% bind_shiny() that did not work. Any suggestions will be greatly appreciated.

The ui.R appears but the visualization does not and I do not get an error message

server.R

library(shiny)
library(ggvis)
library(dplyr)


dataS <-read.csv("https://raw.githubusercontent.com/indianspice/IS608/master/Final%20Project/Data/shinydata.csv",
              stringsAsFactors = FALSE)


function(input, output, session) {
#Filter breaches
breaches <- reactive({
    records <- input$records
    minyear <- input$year[1]
    maxyear <- input$year[2]

    # Apply filters
    b <- dataS %>%
    filter(
    TotalRecords >= records,
    Year >= minyear,
    Year <= maxyear
) %>%
arrange(records)

#Filter by breach
if (input$breach != "All") {
    breach <- paste0("%", input$breach, "%")
    b <- b %>% filter(Breach %like% breach)
}

#Filter by company
if (!is.null(input$company) && input$company != "") {
    company<- paste0("%", input$director, "%")
    b <- b %>% filter(Company %like% company)
}     

reactive({
    xvar_name <- names(axis_vars)[axis_vars == input$year]
    yvar_name <- names(axis_vars)[axis_vars == input$records]

    xvar <- prop("x", as.symbol(input$xvar))
    yvar <- prop("y", as.symbol(input$yvar))

    breaches %>%
        ggvis(x=xvar, y=yvar, stroke = ~breach) %>%
        layer_points() %>%
        add_axis("x", title = xvar_name) %>%
        add_axis("y", title = yvar_name) %>%
        add_legend("stroke", title = "Breach Type", 
                   values = c("Hacking or Malware", 
                              "Unintended Disclosure",
                              "Insider",
                              "Portable Device",
                              "Stationary Device",
                              "Unknown",
                              "Payment Card Fraud",
                              "Physical Loss")) %>%
        scale_nominal("stroke", domain = c("Hacking", 
                                           "Unintended",
                                           "Insider",
                                           "Portable",
                                           "Stationary",
                                           "Unknown",
                                           "Payment",
                                           "Physical"),
                          range = c("red", "orange"))  %>%
        bind_shiny("ggvis", "ggvis_ui")
   }) 

})


}

ui.R

library(shiny)
library(ggvis)

dataS <- read.csv("https://raw.githubusercontent.com/indianspice/IS608/master/Final%20Project/Data/shinydata.csv",
             stringsAsFactors = FALSE)


fluidPage(
    titlePanel("Data Breaches in the United States"),
    #fluidRow(
       column(4,
               h4("Filter Data"),
               sliderInput("records", "Number of records breached",
                           min = 10,
                           max = 1000000,
                           value = 10000,
                           step = 500),
               sliderInput("year", "Year breach reported",
                           sep = "",
                           min = 2005, 
                           max = 2017, 
                           value = c(2007, 2010)),
               selectInput("breach", "Type of breach",
                           c("All", 
                             "Hacking or Malware", 
                             "Unintended Disclosure", 
                             "Insider", 
                             "Portable Device", 
                             "Stationary Device",
                             "Unknown", 
                             "Payment Card Fraud", 
                             "Physical Loss")),
               selectInput("organzation", "Select type of organization",
                           choices = unique(dataS$TypeofOrganization)),
               selectInput("company", "Select company",
                           choices = unique(dataS$Company)
               ),
              textInput("companyName", "Enter company name")
           ),

    #),
    mainPanel(
        uiOutput("ggvis_ui"),
        ggvisOutput("ggvis")
    )


    )

Data

Company TypeofBreach            TypeofOrganization     TotalRecords Year
Bullitt Unintended Disclosure   Educational Institutions    676     2009
Roane   Portable Device         Educational Institutions    14783   2009
Halifax Portable Device         Healthcare Medical Provider 33000   2009
Suffolk Unintended Disclosure   Educational Institutions     300    2009
Penrose Physical Loss           Healthcare Medical Providers  175   2009

You are defining a reactive inside a reactive, which is bad. You should define your reactive (changing) data breaches using reactive - that's fine. Then, you should observe changes of that data using observe:

observe({
   breaches() ... <do something>
   ...
   %>% bind_shiny("ggvis", "ggvis_ui")
})

and then, at the end, use bind_shiny. See the following minimal example for an introduction how to do it (inspired by ggvis help pages):

library(shiny)
runApp(list(
  ui = fluidPage(
    sliderInput("slider", "Select rows from mtcars to consider", min=1, max = nrow(mtcars), step = 1, value = c(1,10)),
    ggvisOutput("p"),
    uiOutput("p_ui")
  ),
  server = function(input, output) {

    # define the data according to some input
    data <- reactive({
      mtcars[ input$slider[1] : input$slider[2], ]
    })

    # observe changes in the data and update ggvis plot accordingly
    observe({
     data %>%
        ggvis(~wt, ~mpg) %>%
        layer_points() %>%
        bind_shiny("p", "p_ui")
    })
  }
))