Follow Datanami:
November 21, 2013

Tableau Embraces R with Version 8.1 Update

Alex Woodie

Tableau Software today shipped a new release of its data analysis tool that includes support for the open source R statistical package, thereby allowing non-technical users to incorporate thousands of powerful statistical models and algorithms into their data visualization and exploration tasks. Tableau 8.1 also includes 64-bit enhancements, new forecasting tools, and enhancements to the mobile interface.

By some counts, the R package has become the most popular statistical language in the IT world, with more users than SPSS and SAS combined. There are an estimated 20 million programmers out there who know R, which is a strong reflection of the power of free and open source software. It’s no wonder that R integration has been a hot commodity sought by Hadoop distributors and analytic database vendors of late.

Tableau sits a little further down the big data totem pole, closer to the users themselves. With Tableau version 8.1, which the company formally released into GA today, R is now integrated directly into Tableau’s calculation engine, which should make it easier for users to tap into the power of R.

“The beauty of R is it has 3,000-plus functions and algorithms and models that are available, so it really adds a lot of analytical depth” to the product, says Francois Ajenstat, senior director of product management for the Seattle, Washington, software company.

Tableau is a big proponent of the Keep It Simple, Stupid (KISS) school of software design. It’s easy for users to get overwhelmed by technology when they’re analyzing large amounts of data, and so Tableau has put a strong emphasis on making software that’s simple to use.

Ajenstat argues that it abided by that precept with the R integration in version 8.1, which he says works within the confines of the regular Tableau experience. “As you’re doing an analysis, you can take an R function, drag that onto the view, and as you interact with the data, seamlessly we’ll be talking to R in the background and getting results dynamically in the view,” he says. “R is baked in so you can really interact and add the full power R with the simplicity and ease of use of Tableau.”

That doesn’t mean you can take a high-school dropout, put him in front of a Tableau screen, and start generating million-dollar insights from petabyte’s worth of raw data. There’s a little bit more to it than that, and it might take some programming or technical setup work. (Also, raw data often needs to be run through the cleansing mill to eliminate dirt, smudges, and the occasional rock.)

Box and whiskers graphs also debut with 8.1.

For example, if a user wants to determine high values or low values from a set of data, there’s an R function that will do that for them. But before an end-user can start applying that algorithm to the data from the comfort of Tableau’s GUI, somebody must configure the values and the arguments that the function will use. “That’s pretty standard in the R world,” Ajenstat says. “There’s lot of ways of driving the algorithms. That’s where a little bit of programming or setup is required.”

Version 8.1 also marks the completion of the migration to 64-bit components for the software. In previous releases, only the data engine (which runs on servers or desktops) ran in 64-bit mode. Everything else, including the desktop visualization layer, ran in 32-bit mode, and was relegated to a maximum memory capacity of 3GB.

That didn’t often pose a problem for data visualization, which is Tableau’s bread and butter. To hit the 3GB barrier, a user would need a visualization with millions of data points loaded on the screen, which Ajenstat says isn’t really feasible.

Rather, the big benefit having both desktop and server components of Tableau 64-bit is overall performance and scalability. This will allow Tableau to support more users on each Tableau server instance. Also, users should see a pretty sizable performance benefit; Ajenstat says preliminary tests show about a 30 percent in performance, although that number is not official.

Tableau will also ship a 32-bit version of the desktop software, Ajenstat says–a big relief for all those organizations that want to keep their Windows XP desktops running for the next 15 years.

The company has made small tweaks to its mobile clients, including giving mobile users more customization options. The company supports dashboard authoring capabilities from tablets, and supports read-only viewing of dashboards from smartphones.

Users working from full PCs gain the capability to copy and paste sheets and dashboards across copybooks. There’s also a new presentation mode that turns a Tableau screen into something like PowerPoint.  The company also added a new forecasting function with version 8.1. Instead of using just one function for a forecast, the product will run several forecast functions in the background, and then select the best one it thinks will be the best fit for the user.

It’s all about helping users find trends hidden in their data sets, Ajenstat says. “Whether it’s small data or big data, we just make it easy to bring that forward,” he says. “When you have big data in Hadoop or NoSQL, how do you really visualize millions and millions of rows? It’s really hard to do. If you saw that it in a traditional report, it probably wouldn’t tell you the insights. But with Tableau, because it’s so easy to use and so visual, you can easily drag and drop and see the data come to life.”

Tableau added 1,500 new customers in the last quarter ended September 30, bringing its total customer count to more than 15,000. The fast-growing company is on track to bring in about $220 million for the year ending December 31, and has a market capitalization that, at $3.6 billion, is a significant multiple of revenue.

Related Items:

Tableau Throws a Brick at Traditional BI

Hitching a Big Data Ride on Amazon’s Cloud

MicroStrategy Tries Out Free Software

 

Datanami