Dynamically change plots based on user input in Shiny

Question

I'm trying to create a shiny app that generates plots based on the user selection of a subset of a loaded dataframe. For example, I have the following dataset:

library(shiny)
library(data.table)

df <- rbind(
  data.table( cat = rep('X', 40), grp = rep(LETTERS[1:4], each=10), x = rep(1:10, times=4), y = rnorm(40) ),
  data.table( cat = rep('Y', 30), grp = rep(LETTERS[1:3], each=10), x = rep(1:10, times=3), y = rnorm(30) ),  
  data.table( cat = rep('Z', 20), grp = rep(LETTERS[4:6], each=10), x = rep(1:10, times=2), y = rnorm(20) )
)

Based on the value for cat that the user selects in the UI, I want shiny to produce charts for each value of grp. So, if the user selects 'X', then there will be 4 plots produced; if they select 'Y' there will be three, and if they select 'Z' there will be 3.

I also want to specify how each chart is generated based on the value of grp. So if grp is A,D or E I want it produce a line plot, otherwise it should produce a scatterplot (only if that grp has that value of course).

Below is the code for my (broken) shiny app:

server <- function(input, output) {

  rv <- reactiveValues(
    i  = NULL,
    df = NULL
  )

  observe({ rv$i <- input$i })

  observe({ rv$df <- df[cat == rv$i] })

  output$test <- renderUI({
    plotList <- lapply( LETTERS[1:6], function(x) plotOutput(x) )

    do.call( tagList, unlist(plotList, recursive=FALSE))
  })

  for(i in LETTERS[1:6]){
    local({
      my_i <- i

      output[[my_i]] <- renderPlot({
        if( my_i %in% c('A','D','E')) {
          with(rv$df[grp == my_i], plot(x,y, type='l'))
        } else {
          with(rv$df[grp == my_i], plot(x,y))
        }
      })
    })
  }

}

ui <- fluidPage(
  titlePanel('Title'),

  sidebarLayout(
    sidebarPanel(
      helpText('Select the Category you would like to view.'),

      selectInput('i', 'Category', c('X','Y','Z'), selectize=TRUE)
    ),

    mainPanel(
      uiOutput('test')
    )
  )
)

shinyApp(ui, server)

does the answer work for you?

Tonio Liebrand
– Tonio Liebrand

2018-12-11 10:37:50 +00:00
Commented Dec 11, 2018 at 10:37 — Tonio Liebrand
– Tonio Liebrand, Commented Dec 11, 2018 at 10:37

Tonio Liebrand · Accepted Answer · 2018-12-07 13:38:22Z

A reproducible example can be found at the bottom.

A few hints:

1) Using reactive contexts:

In your for Loop at the bottom of the Server Code you are using the reactive variable rv, so you will have to run the Code in a reactive Content. So wrap it in observe().

2) Create a list of Outputs:

If I am not mistaken you used some of the Code in this answer: dynamically add plots to web page using shiny.

It is a good starting Point. For the part of the taglist it might be easier to simplify to:

output$test <- renderUI({
    lapply(unique(rv$df$grp), plotOutput)
})

You can also add tagList(), but it is not necessary here,...

3) Correcting the sample data:

You might want to update the df variable:

  data.table(cat = rep('Z', 20), grp = rep(LETTERS[4:6], each=10), 
             x = rep(1:10, times=2), y = rnorm(20) )

Here your have three letters, so you might change it to LETTERS[5:6] or update the other numbers.

Full reproducible example:

library(shiny)
library(data.table)

df <- rbind(
  data.table( cat = rep('X', 40), grp = rep(LETTERS[1:4], each=10), x = rep(1:10, times=4), y = rnorm(40) ),
  data.table( cat = rep('Y', 30), grp = rep(LETTERS[1:3], each=10), x = rep(1:10, times=3), y = rnorm(30) ),  
  data.table( cat = rep('Z', 30), grp = rep(LETTERS[4:6], each=10), x = rep(1:10, times=3), y = rnorm(30) )
)
server <- function(input, output) {

  rv <- reactiveValues(
    i  = NULL,
    df = NULL
  )

  observe({ rv$i <- input$i })

  observe({ rv$df <- df[cat == rv$i] })

  observe({
    for(letter in unique(rv$df$grp)){
      local({
        let <- letter
        output[[let]] <- renderPlot({
          if( let %in% c('A','D','E')) {
            with(rv$df[grp == let], plot(x, y, type='l'))
          } else {
            with(rv$df[grp == let], plot(x,y))
          }
        })
      })
    }
  })

  output$test <- renderUI({
    lapply(unique(rv$df$grp), plotOutput)
  })

}

ui <- fluidPage(
  titlePanel('Title'),
  sidebarLayout(
    sidebarPanel(
      helpText('Select the Category you would like to view.'),
      selectInput('i', 'Category', c('X','Y','Z'), selectize=TRUE)
    ),

    mainPanel(
      uiOutput('test')
    )
  )
)

shinyApp(ui, server)

Collectives™ on Stack Overflow

Dynamically change plots based on user input in Shiny

1 Answer 1

Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related