vegawidget

The goal of vegawidget is to render Vega-Lite and Vega specifications as htmlwidgets, and to provide you a means to communicate with a Vega chart using JavaScript or Shiny. Its ambition is to be a low-level interface to the Vega(-Lite) API, such that other packages can build upon it to offer higher-level functions to compose Vega(-Lite) specifications.

This is the key difference with the vegalite package: it provides a set of higher-level functions to compose specifications, whereas vegawidget concerns itself mainly with the rendering of the htmlwidget.

To be clear, although Vega-Lite offers a grammar-of-graphics, this package does not offer a user-friendly framework to compose graphics, like those provided by ggplot2 or ggvis. However, this package may be useful to:

• build re-usable Vega and Vega-Lite specifications for deployment elsewhere, if you can tolerate the frustration of building specifications using lists.
• develop higher-level, user-friendly packages to build specific types of plots, or even to build a general ggplot2-like framework, using this package as a foundation (or inspiration).

Installation

You can install vegawidget from CRAN with:

install.packages("vegawidget")

You can also install vegawidget from GitHub with:

# install.packages("devtools")
devtools::install_github("vegawidget/vegawidget")

This package supports these versions of Vega libraries:

library("vegawidget")
vega_version()
#> $vega_lite
#> [1] "2.6.0"
#> 
#> $vega
#> [1] "4.0.0"
#> 
#> $vega_embed
#> [1] "3.25.0"

The first released version of vegawidget supports Vega-Lite 2.6.0, the last release of Vega-Lite that will run using JavaScript version ES5. The next version of Vega-Lite, 3.0.0, will use JavaScript ES6.

In practical terms, this means that this version of vegawidget will be the only version that will display properly using version 1.1.x of the RStudio IDE.

As soon as Vega-Lite makes its 3.0.0 release, we will make another vegawidget release - for which you will need to use a (presently preview) version of the RStudio IDE at least 1.2.x.

Introduction

For a comprehensive introduction to Vega-Lite, please visit the its web site. An interactive tutorial to vegawidget is available at shinyapps.io.

Vega(-Lite) specifications are just text, formatted as JSON. However, in R, we can use lists to build specifications:

spec_mtcars <-
  list(
    `$schema` = vega_schema(), # specifies Vega-Lite
    description = "An mtcars example.",
    data = list(values = mtcars),
    mark = "point",
    encoding = list(
      x = list(field = "wt", type = "quantitative"),
      y = list(field = "mpg", type = "quantitative"),
      color = list(field = "cyl", type = "nominal")
    )
  ) %>% 
  as_vegaspec()

The as_vegaspec() function is used to turn the list into a vegaspec; many of this package’s functions are built to support, and render, vegaspecs:

spec_mtcars

The appearance of the chart above depends on where you are reading it:

• On this package’s pkgdown site, it is rendered as part of an HTML environment, showing its full capabilities.

• At its GitHub code site, the chart is further rendered as a static PNG.

This is package a low-level interface to Vega-Lite and the Vega ecosystem, which has a lot of powerful capabilities, highlighted in this series of articles:

• Specify using vegaspec: how to construct and render a vegaspec.
• Render using vegawidget: advanced rendering options.
• Extend using Shiny: how to interact with Vega charts using Shiny.
• Extend using JavaScript: how to interact with Vega charts using JavaScript.
• Create Images: how to create and save PNG or SVG images.
• Work with Dates and Times: dates and times in Vega(-Lite) work a little differently from R.
• Import into Other Packages: how to import vegawidget functions into your package, then re-export them.

To share your Vega(-Lite) creation on the Blocks website, you can use the vegablock package.

Integration with other packages

Although there is an article dedicated to this aspect of the package, it warrants further emphasis.

This package provides functions to render Vega(-Lite) specifications; although it provides some helpers, it does not provide higher-level functions to build specifications. Rather, this is left to other packages. Even though you can use its functions directly, you are invited to import and re-export them for use in your package.

Accordingly, this package offers a templating function, use_vegawidget(), to help you integrate vegawidget functions into your package. For example, it is used to import and re-export vegawidget functions for the altair package.

Known limitations

The two biggest known limitations are both associated with vegaspecs that contain datasets specified using remote URLs:

• The RStudio IDE does not let you access remote URLs for security reasons. A chart that accesses remote data will not render in the IDE; it will render in a browser window.

• The image functions use nodejs to create SVG strings. We are not able to configure the proxy in our node scripts, so if your computer uses a proxy to access remote data, our functions will not create the images properly.

Acknowledgements

• Alicia Schep has been instrumental in guiding the evolution of the API, and for introducing new features, particularly the JavaScript and Shiny functions.
• Haley Jeppson and Stuart Lee have provided valuable feedback and contributions throughout the package’s development.
• Bob Rudis and the vegalite package provided a lot of the inspiration for this work, providing a high-level interface to Vega-Lite.
• The Altair developers, for further popularizing the notion of using a programming language (Python) to create and render Vega-Lite specifications.
• The Vega-Lite developers, for providing a foundation upon which the rest of this is built.

Contributing

Contributions are welcome, please see this guide. Please note that this project is released with a Contributor Code of Conduct. By participating in this project you agree to abide by its terms.