rOpenSci Post on ggplotly

rOpenSci Post on ggplotly

Make your ggplots shareable, collaborative, and with D3
========================================================

[Ggplotly](https://github.com/ropensci/plotly) and [Plotly's R API](https://plot.ly/api/r) let you make ggplot2 plots, add `py$ggplotly()`, and make your plots interactive, online, and drawn with D3. Let's make some.


```{r echo=FALSE}
options(bitmapType="cairo")
#see: http://cran.r-project.org/web/packages/opencpu/NEWS
library(knitr)
```

# 1. Getting Started and Examples

Here is Fisher's iris data.

```{r}
library(ggplot2)
ggiris <- qplot(Petal.Width, Sepal.Length, data=iris, color=Species)
print(ggiris)
```

Let's make it in Plotly. Install:

```{r, eval=FALSE}
install.packages("devtools")
library(devtools)
install_github("plotly", "ropensci")
```

Load.

```{r}
library(plotly)
```

Sign up [online](https://plot.ly), use our public keys below, or sign up like this:

```{r, eval=FALSE}
signup("new_username", "your_email@domain.com")
```

That should have responded with your new key. Use that to create a plotly interface object, or use ours:

```{r, eval=TRUE}
py <- plotly("RgraphingAPI", "ektgzomjbx")
```

It just works.

```{r, plotly=TRUE}
py$ggplotly(ggiris)
```

The call opens a browser tab. Or in an Rmd document, the plot is embedded if you specify the plotly=TRUE chunk option (see [source](https://gist.github.com/msund/10820797)). If you're running this from the source, it makes all the graphs at once in your browser. Reaction my first time: here be dragons. 

If you click the "data and graph" link in the embed, it takes you to Plotly's GUI, where you can edit the graph, see the data, and share your plot with collaborators. 

1.2 Maps
---------------------

Next: Maps!

```{r, fig.width=15, tidy=FALSE, fig.align='center', frameBorder="0"}
data(canada.cities, package="maps")
viz <- ggplot(canada.cities, aes(long, lat))+
  borders(regions="canada", name="borders")+
  coord_equal()+
  geom_point(aes(text=name, size=pop), colour="red",
               alpha=1/2, name="cities")
```

Call Plotly.

```{r, plotly=TRUE}
py$ggplotly(viz)
```

1.3 Scatter
---------------------

Want to make a scatter and add a [smoothed conditional mean](http://docs.ggplot2.org/current/geom_smooth.html)? Here's how to do it in Plotly. For the rest of the plots, we'll just print the Plotly version to save space. You can hover on text to get data, or click and drag across a section to zoom in.

```{r, fig.width=15, tidy=FALSE, plotly=TRUE}
model <- lm(mpg ~ wt + factor(cyl), data=mtcars)
grid <- with(mtcars, expand.grid(
    wt = seq(min(wt), max(wt), length = 20),
    cyl = levels(factor(cyl))
))

grid$mpg <- stats::predict(model, newdata=grid)

viz2 <- qplot(wt, mpg, data=mtcars, colour=factor(cyl)) + geom_line(data=grid)
py$ggplotly(viz2)
```

1.4 Lines
---------------------

Or, take `ggplotly` for a spin with the organge dataset:

```{r, fig.width=15, fig.align='center', tidy=FALSE, plotly=TRUE}
orange <- qplot(age, circumference, data = Orange, colour = Tree, geom = "line")
py$ggplotly(orange)
```

1.5 Alpha blend
---------------------

Or, make plots [beautiful](http://mandymejia.wordpress.com/2013/11/13/10-reasons-to-switch-to-ggplot-7/). 

```{r, fig.width=15, tidy=FALSE, plotly=TRUE, cache.lazy=TRUE}
prettyPlot <- ggplot(data=diamonds, aes(x=carat, y=price, colour=clarity))
prettyPlot <- prettyPlot + geom_point(alpha = 1/10)
    py$ggplotly(prettyPlot)
```

1.6 Functions
---------------------

Want to [draw functions](http://stackoverflow.com/questions/1853703/plotting-functions-in-r) with `curve`?

```{r, fig.width=15, tidy=FALSE, plotly=TRUE}
eq <- function(x) {x*x}
tmp <- data.frame(x=1:50, y=eq(1:50))

# Make plot object
p <- qplot(x, y, data=tmp, xlab="X-axis", ylab="Y-axis")
c <- stat_function(fun=eq)

py$ggplotly(p + c)
```

# 2. A GitHub for data and graphs

Like we might work together on code on GitHub or a project in a Google Doc, we can edit graphs and data together on Plotly. Here's how it works: 

- Your URL is shareable. 
- Public use is free.
- You can set [the privacy](http://plot.ly/api/r/docs/privacy) of your graph. 
- You can edit and add to plots from our GUI or with R or [APIs](https://plot.ly/api) for Python, MATLA, Julia, Perl, Arduino, Raspberry Pi, and REST. 
- You get a profile of graphs, like [Rhett Allain](https://plot.ly/~RhettAllain/) from Wired Science.
- You can [embed interactive graphs in iframes](http://plot.ly/api/r/docs/iframes).

2.1 Inspiration and team
---------------------

Plotly's API is part of `[rOpenSci](ropensci.org)`, being developed by the brilliant [Toby Hocking](http://cbio.ensmp.fr/~thocking/), and on [GitHub](https://github.com/ropensci/plotly). Your thoughts, issues, and pull requests are welcome. Right now, you can make scatter and line plots; let us know what you'd like to see next.

The project was inspired by [Hadley Wickham](https://github.com/hadley/) and the elegance and precision of [`ggplot2`](http://ggplot2.org/). Thanks to [Scott Chamberlain](scottchamberlain.info), [Joe Cheng](https://github.com/jcheng5), and [Elizabeth Morrison-Wells](https://twitter.com/efvmw) for their help.

# 3. ggthemes and Plotly

Using `[ggthemes](https://github.com/jrnold/ggthemes)` opens up another set of custom graph filters for styling your graphs. To get started, you'll want to install `ggthemes`.

```{r eval=FALSE}
library("devtools")
install_github("ggthemes", "jrnold")
```

and load your data.

```{r}
library("ggplot2")
library("ggthemes")
dsamp <- diamonds[sample(nrow(diamonds), 1000),]
```

Inverse gray.

```{r, fig.width=15, tidy=FALSE, plotly=TRUE}
gray <- (qplot(carat, price, data = dsamp, colour = cut) + theme_igray())
py$ggplotly(gray)
```

The Tableau scale.

```{r, fig.width=15, tidy=FALSE, plotly=TRUE}
tableau <- (qplot(carat, price, data = dsamp, colour = cut) + theme_igray() + scale_colour_tableau())
py$ggplotly(tableau)
```

[Stephen Few's](http://www.perceptualedge.com/articles/visual_business_intelligence/rules_for_using_color.pdf) scale.

```{r, fig.width=15, tidy=FALSE, plotly=TRUE}
few <- (qplot(carat, price, data = dsamp, colour = cut) + theme_few() + scale_colour_few())
py$ggplotly(few)

```

Hey Matt,
I forked your Gist and made the following changes: https://gist.github.com/mkcor/d42b835d46c70a2611d2/revisions
I cannot open a PR but feel free to apply the patch if worthwhile or inspire from my edits.
Thanks!
Marianne