Teradata will join the growing list of analytics vendors offering software based on the R programming language...
with a new interface for its Aster database.
Teradata Aster R, to be released by the end of this year, will enable customers to use the R language to access and run jobs against data in the Aster database. It will be free for Aster customers.
The new software will integrate with Aster through Teradata's SNAP Framework, which already supports other query languages like SQL and MapReduce.
With the announcement, Teradata joins other vendors such as IBM SPSS, Oracle, SAS and StatSoft, which have all released their own interpretations of the open-source R programming language. These commercial versions must solve two inherent problems with open-source R. First, R runs on a single computer, so it is limited by that machine's processing power and can lead to long job times. Second, it processes jobs in-memory, so it can only handle small data volumes.
Teradata Aster R solves these problems by processing jobs across a distributed database setup. This enhances processing power and removes limits on data volume. This processing happens in-database, rather than in-memory.
Arlene Zaima, a Teradata product marketing manager, said the company took on R because there are so many users. It is one of the primary programming languages statisticians and data scientists learn in school, so there is a large community of users. A recent survey from an online data science community showed that R is the second-most commonly used analytics tool, behind only software from Cambridge Massachusetts-based vendor RapidMiner. A total of 38.5% of the 3,000 data scientists who responded said they use R.
Additionally, Zaima said Teradata was seeing existing customers using open-source R in commercial applications at small scales.
"We're seeing great adoption of open-source R but we all understand that it was designed essentially for research in academia where they're not processing massive amounts of data like we are in the business world," Zaima said.
IBM supports releases new self-service offerings
R programming language goes mainstream
Analytics vendors empower business users