cynkra

Introducing blockr: Data Apps Without Code

From the Blog
Shiny
R
Mike Page

Author

Mike Page

Published

We are excited to announce the first stable release of blockr, a tool that lets you build data apps in minutes, using a point-and-click user interface.

Each blockr application is composed of blocks connected together in a graph, where each block performs a single part of a data analysis workflow (e.g., reading, transforming, visualising, etc.). Use blockr to perform data analyses and create dashboards, all without writing a single line of code.

a screenshot of a blockr application showing a linear modelling workflow

Built on R

blockr can be thought of as a visual programming interface to R, the key technology it is built upon. It was designed with extensibility in mind, meaning that most tasks in R can be transformed into a corresponding workflow in blockr.

To achieve this extensibility, blockr has been designed to use many lower-level packages that each perform specific roles. This first stable release of blockr sees the publication of five of these fundamental packages to CRAN:

Collectively, these five packages cover most of the basic needs required to start building a range of different applications.

Start Building Today

blockr is an open-source initiative funded by Bristol Myers Squibb. Born from the need to solve real-world problems in the pharmaceutical sector, it is continually growing to support a new range of projects and use-cases.

If you too are interested in taking blockr for a spin, you can get started by either:

  1. Visiting our online playground at blockr.cloud for free, with no installation required.
  2. Installing the framework directly from GitHub and running it locally with just two commands:
# Install blockr (meta-package bundling the lower-level packages)
pak::pak("BristolMyersSquibb/blockr")

# Launch an empty application
blockr::run_app()

Looking Ahead

In the coming year, we will continue our work enhancing blockr, including AI-assisted support as well as implementing a range of different blocks to cover many more use cases. We will also release a suite of blog posts diving into some of the more technical details as well as guides on how to make blockr work for you across different domains.

In the meantime, we hope you have fun trying blockr, and welcome any feedback you may have.