d3po: Fast and Beautiful Interactive Visualization for ‘Markdown’ and ‘Shiny’

R-CMD-check

The Problem

R already features excellent visualization libraries such as ‘D3’ (via the ‘r2d3’ package) or ‘plotly’, however though those enable the creation of great looking visualisations they have very steep learning curves and even sometimes require some understanding of ‘JavaScript’.

There is also the ‘Highcharter’ package but the underlying ‘JavaScript’ library (‘Highcharts’) is not free for commercial and governmental use. A single user license costs around 2,500 USD, something that NGOs cannot always afford.

Another relevant aspect is the lack of easy alternatives to provide internalization, something that can be addressed easily in a community project due to the variety of languages.

Our intention is to solve those problems by releasing a complete ‘R’ package under an open source license (Apache 2.0).

The proposal

Overview

The name ‘d3po’ is inspired after ‘r2d3’ but also ‘d3po’ is something that a chilean will read as “dee three pooh”, and “pooh” is a very typical chilean slang. That slang reflects the true chilean spirit that copes with earthquakes but in the end they build the technology and adapt to the medium constraints.

Rather than starting from scratch, ‘d3po’ takes many ideas and adapts codes from ‘highcharts’ (just inspiration, not code), ‘echats’ (excellent library), ‘d3plus’ 1.9.8 (unmaintained, which is very unfortunate) and many posts from Stackoverflow that were adapted since 2016 to the date.

It is worth mentioning that the most important goal is to provide out-of-the-box beautiful visualizations with minimum time and coding effort from the final user.

At the end of the project ‘d3po’ will be a new software that can be perceived as an intermediate layer between the user and D3 by providing “templates”, enabling high quality interactive visualizations oriented to and designed to be used with ‘Shiny’.

The focus will remain on the integration with ‘Shiny’ and ‘R Markdown’, the javascript outcome from this project could nonetheless be used with other tools.

The ultimate aim of the project is to produce a package that:

Details

The Minimum Viable Product includes these visualization methods and features:

The goal is to provide a package that becomes a gold standard for interactive visualization and facilitates using Shiny.

As an R package, the idea is to obtain a high quality result to be accepted on rOpenSci and then CRAN. It should run on any computer with ‘RStudio’ and any server running ‘Shiny’.

We have worked under three goals:

  1. Minimum number of dependencies
  2. Use %>% to provide a ‘Tidyverse’-alike experience
  3. Make the internals “hard” and the externals “easy”