If you missed this morning's webinar, Revolution R Enterprise, 100% R and More, I've embedded the slides below. Interestingly, about half of today's participants were SAS users, and the remainder R users. The first section introduces open-source R, and the second describes the additional features of Revolution R Enterprise.
Unfortunately we had a technical hiccup with the recording (a dropped internet connection in the producer booth), so we can't make a replay available as we usually do with our webinars. But you can see a recording of the last time I have this presentation on YouTube.
Revolution Analytics webinars: Revolution R Enterprise, 100% R and More
Hi David,
Can we extend the functionality of revoscaler package i.e can we use function of other package in revoscaler for doing faster data analysis.
Posted by: sunny | February 23, 2012 at 21:01
@Sunny, yes, you can implement your own Big Data algorithms with the RxExec function, which you can use to run arbitrary R code in parallel across cores or across machines in a cluster. The result will be a list of computation results on parts of the data, which you'll need to consolidate with more R programming. The technique lends itself well to problems where partitioning the data is natural (e.g. random forests), but with programming you can implement just about any statistical method (or at least approximate it).
Posted by: David Smith | February 24, 2012 at 10:29
Is it possible to do scheduling e.g i want to run a particular function after specified time interval
Posted by: sunny | April 11, 2012 at 02:39