Oracle moves on R statistics market with Advanced Analytics

Oracle is hoping to carve out a prominent place in the world of R, the open source statistical modelling language with roots in academia but an increasingly high profile in enterprise IT shops. It has announced a new Advanced Analytics product that ties R to its database and family of software/hardware appliances.

Oracle Advanced Analytics consists of Oracle R Enterprise, along with the vendor's existing Data Mining module. It's available as an Oracle 11g database option and costs $23,000 (£14,000) per processor licence. Data Mining will fall off the price list and be supplanted by Advanced Analytics.

Oracle R Enterprise allows R users to apply their existing scripts and statistical models to information held in Oracle 11g, according to the company. Oracle data mining algorithms, OLAP (online analytical processing) cubes and other assets are also integrated with R, and R scripts can be called with standard SQL.

Meanwhile, a cross-integration between R, 11g and the Hadoop framework for large scale data processing enables R programmers to write one script and deploy it in multiple environments, including Oracle's Big Data appliance, Oracle said.

Result sets from the Advanced Analytics module can also be accessed through Oracle's BI (business intelligence) suite as well as the Exalytics In-Memory appliance, according to Oracle.

In addition, Oracle is redistributing the R language and providing support for the entire stack, said George Lumpkin, vice president of data warehousing product management.

Oracle has simplified administration and management for R users, Lumpkin said. "Instead of moving data in from multiple places, it's in one server, there's one security policy for the users, there's all of the good things about having a consolidated data source."

Other statistical modelling vendors, such as SAS Institute, have made moves to support R. Customers shouldn't necessarily view Oracle's announcement as a heightening of competition between the two companies, according to Lumpkin.

"Oracle does and continues to have a strong relationship with SAS," Lumpkin said. "We recognise there are large Oracle customers that have investments in SAS, and we intend to support those customers properly," he said. "Oracle has not had an extensive set of products in [the R market] before. One of the reasons we are focusing on R, we do see R growing into commercial entities and into the enterprise space."

The vendor plans to contribute to the R open source project, and in fact has already taken ownership of an R package that had been providing ODBC-like linkage to Oracle's database, albeit not with the same sophistication as Oracle R Enterprise provides, Lumpkin said.

It's not surprising that Oracle has moved to support R, which "has become a real ubiquitous force in advanced analytics," said Forrester Research analyst James Kobielus. "It's everywhere. Enterprise adoption of it has been growing steadily. When we ask our customers what they're using for statistical modelling they'll say SAS or [IBM's] SPSS, but they increasingly say R in the same breath."

SAS "has everything in the kitchen sink as far as statistical modelling" goes, as does SPSS, Kobielus said. You can also find pretty much everything you need out in the R community, he added. However, save for offerings from startups like Revolution Analytics, R has lacked a unified development platform, and Oracle's announcement is an attempt to fill that gap.

It's likely that the next year will see other database and data warehousing vendors, such as IBM, Teradata and EMC Greenplum, start to build out their R tooling, Kobielus said.