ToolTip when you mouseover a ggplot on shiny

Question

I am building a shiny application.

I am plotting charts using ggplot.

When I mouseover the points on the graph, I want a tooltip showing one of the columns i

Answer 1

I've released, together with my colleague, a package called GGTips (it's not on CRAN) that do just that, adding tooltips on plots. We created our own solution because we have trouble in recreating our complex plots using plotly that's not 100% compatible with ggplot2. Git repo have a link to online demo.

Answer 2

If I understand the question correctly, this can be achieved with the recent update of the shiny package for both the ggplot2 and the base package. Using this example from Winston Chang and Joe Cheng http://shiny.rstudio.com/gallery/plot-interaction-basic.html , I was able to solve this problem. Hover is now an input argument into plotOutput() so that is added to the ui along with a verbatimTextOutput to display mtcars$wt for the point hovered over.

In the server I basically make a distance vector which calculates the distance from the mouse to any point in the plot and if this distance is less than 3 (works in this application) then it shows mtcars$wt for the closest point to your mouse. To be clear input$plot_hover returns a list of info about the location of the mouse and only the x and y elements are extracted from input$plot_hover in this example.

library(ggplot2)
library(Cairo)   # For nicer ggplot2 output when deployed on Linux

ui <- fluidPage(
    fluidRow(
        column(width = 12,
               plotOutput("plot1", height = 350,hover = hoverOpts(id ="plot_hover"))
        )
    ),
    fluidRow(
        column(width = 5,
               verbatimTextOutput("hover_info")
        )
    )
)

server <- function(input, output) {


    output$plot1 <- renderPlot({

        ggplot(mtcars, aes(x=mpg,y=disp,color=factor(cyl))) + geom_point()

    })

    output$hover_info <- renderPrint({
        if(!is.null(input$plot_hover)){
            hover=input$plot_hover
            dist=sqrt((hover$x-mtcars$mpg)^2+(hover$y-mtcars$disp)^2)
            cat("Weight (lb/1000)\n")
            if(min(dist) < 3)
                mtcars$wt[which.min(dist)]
        }


    })
}
shinyApp(ui, server)

I hope this helps!

Answer 3

You can also use a little bit JQuery and conditional renderUI to show a custom tooltip near the pointer.

library(shiny)
library(ggplot2)

ui <- fluidPage(

  tags$head(tags$style('
     #my_tooltip {
      position: absolute;
      width: 300px;
      z-index: 100;
     }
  ')),
  tags$script('
    $(document).ready(function(){
      // id of the plot
      $("#plot1").mousemove(function(e){ 

        // ID of uiOutput
        $("#my_tooltip").show();         
        $("#my_tooltip").css({             
          top: (e.pageY + 5) + "px",             
          left: (e.pageX + 5) + "px"         
        });     
      });     
    });
  '),

  selectInput("var_y", "Y-Axis", choices = names(mtcars), selected = "disp"),
  plotOutput("plot1", hover = hoverOpts(id = "plot_hover", delay = 0)),
  uiOutput("my_tooltip")
)

server <- function(input, output) {

  data <- reactive({
    mtcars
  })

  output$plot1 <- renderPlot({
    req(input$var_y)
    ggplot(data(), aes_string("mpg", input$var_y)) + 
      geom_point(aes(color = factor(cyl)))
  })

  output$my_tooltip <- renderUI({
    hover <- input$plot_hover 
    y <- nearPoints(data(), input$plot_hover)[ ,c("mpg", input$var_y)]
    req(nrow(y) != 0)
    verbatimTextOutput("vals")
  })

  output$vals <- renderPrint({
    hover <- input$plot_hover 
    y <- nearPoints(data(), input$plot_hover)[ , c("mpg", input$var_y)]
    # y <- nearPoints(data(), input$plot_hover)["wt"]
    req(nrow(y) != 0)
    # y is a data frame and you can freely edit content of the tooltip 
    # with "paste" function 
    y
  })
}
shinyApp(ui = ui, server = server)

---

EDITED:

After this post I searched internet to see whether it could be done more nicely and found this wonderful custom tooltip for ggplot. I believe it can hardly be done better than that.

Answer 4

Using plotly, you can just translate your ggplot into an interactive version of itself. Just call the function ggplotly on your ggplot object:

library(plotly)

data(mtcars)

shinyApp(
ui <- shinyUI(fluidPage(
  sidebarLayout(sidebarPanel( h4("Test Plot")),
    mainPanel(plotlyOutput("plot1"))
  )
)),

server <- shinyServer(
  function(input, output) {
    output$plot1 <- renderPlotly({
      p <- ggplot(data=mtcars,aes(x=mpg,y=disp,color=factor(cyl)))
      p <- p + geom_point()

      ggplotly(p)
    })
  }
))

shinyApp(ui, server)

For customizations of what is shown in the tooltip, look e.g. here.