CODEWITHSUNDEEP
rshinyggvis
Dambo
Dambo

Reputation: 3496

How to modularize a dynamic ggvis chart in shiny?

My app has a few charts, and I would like to draw them in a sperate script in order to have a cleaner server.R which then contains only the bind_shiny calls.

Also, because I would like to recycle some of the building blocks of my charts, I would like to split the script to draw the charts in smaller parts of code.

Right know, I have a charts.R file like this:

chart1 <- reactive({
  cars %>% 
    ggvis(~speed) 

})

chart2 <- reactive({
  chart1 %>% layer_bars() 
})

The reactivity is not needed of course, but it will in my original application.

Then, I would like to call the chart on server.R this way:

library(shiny)
library(ggvis)
source("charts.R")
shinyServer(function(input, output) {

chart2 %>% 
  bind_shiny("test1", "test1_ui")
})

ui.R

library(shiny)
library(ggvis)
shinyUI(fluidPage(

  fluidRow(titlePanel("My app")),

    fluidRow(
          column(12,
                 tags$h1("hello"),
                 ggvisOutput('test1')
                 )
    )
   )
  )

I have noticed everything works in charts.R when I put everything in one object, but it does not if I split the function that draws the chart in multiple pieces. However, something like this works

> a <-  cars %>% 
+   ggvis(~speed)
> a
Guessing layer_histograms()
Guessing width = 0.5 # range / 42
> a %>% layer_bars()

Anybody could explain how to fix the issue and why my two-steps approach works in R but not in Shiny?

Upvotes: 4

Views: 140

Answers (1)

Sumedh
Sumedh

Reputation: 4965

If your problem is that you can't see the plot, that is because in your second recative statement, you need to refer to chart1 as chart1(). In other words, change your charts.R to

Charts.R

chart1 <- reactive({
            cars %>% 
              ggvis(~speed)
 })

chart2 <- reactive({
           chart1() %>% layer_bars()
 })

If you want to learn more about reactivity, please check out this link.

Upvotes: 0

Related Questions