Showing Off Your Beautiful rChart

Just like making rCharts is easy, showing off your beautiful creation is also super simple with lots of built in options. We will look at each of these options and provide examples. First let's grab some data. For this example, we will get daily S&P 500 prices from Yahoo! Finance using getSymbols and then use runMean to calculate a 200 day simple moving average.

require(quantmod)
require(reshape2)
require(rCharts)

#get sp500 prices from Yahoo! Finance
sp500 <- getSymbols("^GSPC", auto.assign = FALSE)

## Warning: downloaded length 101474 != reported length 200

#add 200 day moving average
sp500$ma <- runMean( sp500[,4], n = 200)
#get date, close, and 200day mov avg as a data.frame
sp500.df <- data.frame(index(sp500),coredata(sp500)[,c(4,7)])
colnames(sp500.df) <- c("date","close","200dMovAvg")
#melt to long format
sp500.melt <- melt( sp500.df, id.vars = 1 )
#get dates to a javascript favorite format
sp500.melt$date <- as.numeric(as.POSIXct(sp500.melt$date)) * 1000

Since we have some data, we can make a simple NVD3 lineWithFocus chart to use as our example. Although we are using NVD3, the sharing methods we will explore are available to all the libraries, including custom creations.

#do a NVD3 lineWithFocus chart
n1 <- nPlot(
  value ~ date,
  group = "variable",
  data = sp500.melt,
  type = "lineWithFocusChart",
  height = 400,
  width = 600
)
#set xAxis up to format dates
n1$xAxis(tickFormat = "#!function(d) {return d3.time.format('%b %d, %Y')(new Date(d))}!#")
#xAxis auto sets x2Axis, but I like a different format for x2Axis
n1$x2Axis(tickFormat = "#!function(d) {return d3.time.format('%Y')(new Date(d))}!#")
#set yAxis up to format numbers
n1$yAxis(tickFormat = "#!function(d) {return d3.format('0,.0')(d)}!#")
n1

So you see our chart. How did I do that? The focus of this tutorial is primarily standalone charts, but as an aside, let's quickly look at the mechanics of this. Since slidify shares the same author as rCharts, it very intelligently handles our rChart. When we call our chart n1, slidify checks to see if knitr is in progress, and if so it automatically uses the show(mode_='iframe') method from rCharts to save a standalone HTML for our chart and then include an <iframe> container in our main document to point to our chart. slidify expects that we might be including multiple rCharts, so it prefers an <iframe>to help prevent conflict between multiple rCharts and speed loading time for our presentation.

If instead, you would like your chart to be shown inline, you could do n1$show("inline") as below.

n1$show("inline")

Showing Off - To Yourself

Showing off to yourself can be very rewarding. Running

n1

will by default create a temp html file (don't worry it will be removed by R when you shut down R) that will open in your default browser. rCharts accomplishes this by the show method. Since it is an R5 reference class, the show method is called automatically.

Using shiny might make us really proud of ourselves. To see our chart with shiny we can run

n1$show(static = F)

and avoid the creation of a temp file. Shiny will start a web server, and your browser will open with our beautiful chart served by R.

Showing Off - GIST & RPubs

Showing off to an audience of one is not nearly as fun as showing off to the world In addition to the show method, rCharts provides a publish method once you are ready for public awe or ridicule. Currently, publish can upload your chart anonymously or with your registered name to Gist from Github or RPubs from RStudio. Gist is the default, so

n1$publish(description="My Beautiful NVD3 Chart")

will result in something like this. There are lots of Gist viewers, but the rCharts author Ramnath Vaidyanathan provides a viewer specifically designed for rCharts. This allows virtually every social sharing site available (Twitter, Pinboard, Facebook, Google+, etc.) and also provides Disqus commenting.

RPubs is another good option for sharing your creation, but it does not offer as many built-in social media options. However, the audience is probably all R users, so you'll still get some love. If we do something like

n1$publish(description = "My Beautiful rChart on RPubs", host="rpubs")

you should get a claim page that you can turn into your own RPub.

For both Gist and RPubs, rCharts respectfully remembers the id, so updates will update (not create) your uploaded masterpieces.

Showing Off to Reproduce

I mentioned earlier that rCharts viewer offers some nice features specifically designed for rCharts. In the spirit of reproducibility, it is always nice to not only share our chart but also the code to create it. rCharts encourages this behavior by making it easy with create_charts. If we save our code as an R script with .R as the extension, we can do something like this.

chartExample <- create_chart('ourbeautifulchartcode.R')
chartExample$publish( 'My Chart with Code' )

You will see now that our Gist will also include our code. This gives us extra opportunity to show off our brilliance.

Lots to Share

At some point, we hope you build up a whole collection of examples. The function make_example_pages will very likely change, but currently it will loop through all the .R scripts in the specified directory and produce a gallery of all the charts.

Here are three examples using make_example_pages:

Please Share

I hope you now understand how easy sharing your rChart is. Please share widely. There is also a gallery of rCharts, and hopefully soon adding to this will be just as easy as above.

Thanks

Thanks to Ramnath Vaidyanathan for seemingly thinking of everything and working so hard to include it in both slidify and rCharts.