Introducing R-Brain: A New Data Science Platform
R-Brain is a next generation platform for data science built on top of Jupyterlab with Docker. It was recently unveiled at JupyterCon in late August. Don’t let the name fool you. R-Brain currently supports R, Python, Structured Query Language (SQL), and more. It has integrated intellisense, debugging, packaging, and publishing capabilities. This cool new solution also has analytics workspace collaboration and marketplace features for personal, professional and enterprise use cases.
R-Brain applications developed by experts can speed up private group development in an enterprise or be shared in the public R-Brain marketplace. The R-Brain marketplace is ideal for teaching or sharing analytics assets for free or for a fee. Currently Golden Gate University is using R-Brain in their analytics programs.
R-Brain IDE Overview
R-Brain improves multi-language data analysis productivity. It offers all the familiar building blocks of the classic Jupyter Notebook (interactive notebook, terminal, text editor, file browser, rich outputs, etc.) in a flexible, powerful user interface. Since it uses popular Docker container technology, this solution can be deployed on-premises or on your preferred cloud platform. Data gurus can develop, package, share and publish analytics workspaces, data sets and applications that use R, Python, Structured Query Language (SQL) scripts. R-Brain also makes it easy to interactively navigate database schemas, view table content and export data.
R-Brain Key Capabilities
Today most data analysts and data scientists have a toolbox filled with a mix of utilities for Python, R, SQL, and other script languages. For example, you might use RStudio for R, Jupyter (iPython notebooks), Anaconda, PyCharm, Spyder or Apache Zeppelin for Python, a SQL IDE like TOAD, SQL Database Studio, MySQL Workbench, SQL Server Management Studio (SMSS) or Visual Studio for SQL, and a text editor IDE such as Notepad++ or Emacs.
With R-Brain, analytics projects across different languages can be easily organized and managed together in one common workspace. Although a few data science providers have similar features, R-Brain’s IDE provides quite a bit more functionality. Here is an IDE comparison provided by R-Brain. Additional data science platform and kernel comparison matrices are also available.
Comparison of R-Brain and Other Data Science Platforms
Getting Started with R-Brain
Getting started with R-Brain is quick and easy. It took me three minutes to set up my environment. I simply registered to create an account and then selected options for my workspace. If someone had shared an analytics workspace with me, it would have been listed as an available workspace in my Inventory. Customized workspace distribution is quite useful for teachers or enterprises that want to standardize analytics environments.
R-Brain currently offers four base workspace types: Intrinisic, Data Science, TensorFlow, and Spark. For my hands-on evaluation, I created a new Data Science workspace.
SQL, R and Python 3. Essential packages to have all you need for running R, Python, Shiny, ggplot and RMarkdown.
- Data Science
SQL, R, Python 3 and Julia with Gadfly, RDatasets and HDF5. Intrinsic package plus pandas, matplotlib, scipy, seaborn, scikit-learn, scikit- image, sympy, cython, patsy, statsmodel, cloudpickle, dill, numba and bokeh for Python, and forecast, caret and randomforest for R.
SQL, R and Python 3. Data Science package plus TensorFlow.
SQL, R and Python 3. Data Science package plus SparkR library and PySpark package.
R-Brain Workspace Configuration
After creating my workspace, I launched it in R-Brain IDE (RIDE). I then added new files for development using my favorite analytics kernels. Peeking through the user interface, I found it to be intuitive. The online documentation, chat feature and YouTube video librarywas also helpful.
RIDE has a float and flexible layout which allowed me to setup my environment by dragging and dropping widgets. On the left side, I noted an area with files and directories. It also showed me all active sessions in Running and a handy list of all commands and shortcuts under Command. On the right side, Help was shown along with my Environment status.
In my first hands-on test of R-Brain, I uploaded Python, R and delimited .csv files. R-Brain allows users to upload any file up to 15MB into a workspace through upload button in the file system. For larger files, inventory options are available. Here is an example from the Python Scikit-Learn Cookbook files. Alternatively, I could have just started coding in a Notebook, Console, Terminal or Text Editor.
Python Development in R-Brain
Just like desktop IDEs, RIDE provides intellisense, interactive debugging for inspecting variables, a call stack, breakpoints and data views for stepping through my analytics code. Unlike other analytics IDE solutions, I could have also connected to a database, explored database schemas, viewed database table content and ran SQL code.
For data frames and matrixes, RIDE data viewer shows the content of the variable. It can be sorted or filtered based on columns. For large datasets, RIDE’s data view does not overload the web browser. It intelligently pulls in data as the user scrolls up or down.
Although most of RIDE’s features are exposed through the user interface, there are a couple “hidden gems” that can be accessed via key combinations.
- Ctrl+Space: Commands can be used to start content assistant.
- Command + S (Windows: Ctrl+S): Commands can be used to save files.
- Command+F (Windows: Ctrl+F): Commands can be used to start a search/replace operation.
When I was ready to share my analytics project, I could publish to the R-Brain Marketplace. R-Brain provides options for sharing interactive Shiny apps, Notebooks, Markdown files or raw Datasets.
R-Brain Publishing Options
One of my favorite R-Brain features is the ability to share public or private persistent links for analytics embedding into web sites or third-party apps.
R-Brain Persistent Embedding Link
For More Information
The R-Brain data science platform is a powerful offering for collaborative, actionable data science when and where you need it. The intuitive web browser experience powered by Docker is simple, flexible, and fun. The workspace capabilities are robust for small to large scale development. R-Brain’s publishing, embedding and marketplace features are fabulous for sharing, integrating or standardizing analytics projects across different languages, groups of developers and users.
If you’d like to try or learn more about R-Brain, check out the following recommended resources.