Plotting demo using ggplot2. Check out at http://spark.rstudio.com/dgrapov/1Dplots/.

global.R

#initialize
library(datasets)
library(ggplot2) 

#helper function (convert vector to named list)
namel<-function (vec){
		tmp<-as.list(vec)
		names(tmp)<-as.character(unlist(vec))
		tmp
	}

server.R

# shiny server side code for each call
shinyServer(function(input, output, session){
	#update variable and group based on dataset
	output$variable <- renderUI({ 
		obj<-switch(input$dataset,
           "iris" = iris,
           "mtcars" = mtcars)	 
		var.opts<-namel(colnames(obj))
		selectInput("variable","Variable:", var.opts) # uddate UI 				 
		}) 
		
	output$group <- renderUI({ 
		obj<-switch(input$dataset,
           "iris" = iris,
           "mtcars" = mtcars)	 
		var.opts<-namel(colnames(obj))
		selectInput("group","Groups:", var.opts) # uddate UI 				 
		}) 
		
	output$caption<-renderText({
		switch(input$plot.type,
			"boxplot" 	= 	"Boxplot",
			"histogram" =	"Histogram",
			"density" 	=	"Density plot",
			"bar" 		=	"Bar graph")
		})
			
	
	output$plot <- renderUI({
		plotOutput("p")
	})
		
	#plotting function using ggplot2
	output$p <- renderPlot({

	plot.obj<<-list() # not sure why input$X can not be used directly?
	plot.obj$data<<-get(input$dataset) 
	plot.obj$variable<<-with(plot.obj$data,get(input$variable)) 
	plot.obj$group<<-with(plot.obj$data,get(input$group)) 
	
	#dynamic plotting options
	plot.type<-switch(input$plot.type,
			"boxplot" 	= 	geom_boxplot(),
			"histogram" =	geom_histogram(alpha=0.5,position="identity"),
			"density" 	=	geom_density(alpha=.75),
			"bar" 		=	geom_bar(position="dodge")
		)
		
	require(ggplot2)
	#plotting theme
	.theme<- theme(
				axis.line = element_line(colour = 'gray', size = .75), 
				panel.background = element_blank(),  
				plot.background = element_blank()
				 )	 
	if(input$plot.type=="boxplot")	{		#control for 1D or 2D graphs 
		p<-ggplot(plot.obj$data, 
				aes(
					x 		= plot.obj$group, 
					y 		= plot.obj$variable,
					fill 	= as.factor(plot.obj$group)
					)
				) + plot.type
				
				if(input$show.points==TRUE)
				{ 
					p<-p+ geom_point(color='black',alpha=0.5, position = 'jitter')
				}
				
		} else {
		
		p<-ggplot(plot.obj$data, 
				aes(
					x 		= plot.obj$variable,
					fill 	= as.factor(plot.obj$group),
					group 	= as.factor(plot.obj$group),
					#color 	= as.factor(plot.obj$group)
					)
				) + plot.type
		}
		
	 p<-p+labs(
			fill 	= input$group,
			x 		= "",
			y 		= input$variable
			)  +
	.theme
	print(p)
	})	
})

ui.R

# UI for app
shinyUI(pageWithSidebar(
	# title
	headerPanel("Select Options"),
	
	#input
	sidebarPanel
	(
		selectInput("dataset","Data:", 
						list(iris = "iris", mtcars = "mtcars")
						),
		uiOutput("variable"), 	# depends on dataset ( set by output$variable in  server.R)
		uiOutput("group"),  		# depends on dataset	( set by output$group in  server.R)
		selectInput("plot.type","Plot Type:", 
						list(boxplot = "boxplot", histogram = "histogram", density = "density", bar = "bar")
						),
		checkboxInput("show.points", "show points", TRUE)
	),	

	# output				
	mainPanel(
		h3(textOutput("caption")),
		#h3(htmlOutput("caption")),
		uiOutput("plot") # depends on input 
	)
))

Thanks for the code.

From local drive, somehow the code's not working. Showing the following message:

Error: evaluation nested too deeply: infinite recursion / options(expressions=)?

Now it's working. Thanks again.

I don't know if it matters but it seems to be working for me. Thanks for the code!

is it possible to update output$group dependent on output$variable ?

This is beautiful, thank you!

This is great and very neat. Thank you!!
Do you have a similar example of a line chart with multiple y axis to share, where the input variables are taken from the user, data is reformatted and then line chart created?