BARUG Meetup at the Googleplex

via Google Open Source Blog by Stephanie Taylor on 9/20/12

What is R? R is an open-source, statistical programming language that is increasingly becoming the lingua franca of modern data analysis. R developers have created over 4,000 packages to run in the R environment, including several dozen developed by Googlers. We continue to work with the R community through conferences, meetups like this one, Faculty Research Awards, and the annual Google Summer of Code program, which funded 16 students to work with the R Project this summer.

Last week, Google hosted the Bay Area useR Group (BARUG) at our Mountain View HQ. Over 130 R users and developers met for pizza, networking, and presentations by Googlers.

This is interesting since everyone I know at Stanford (not that far from the Googleplex) functions with SAS and Stata. I suspect that this hoopla about R is like the hoopla three decades ago about C, when Bells Labs was proclaiming it to be the next big language, the successor to pl1, fortran, algol, pascal, cobol (if that doesn't date a comment…), and a variety of other languages of the time. C has spun off a number of derivative languages, and it has emerged as the language for systems programming, especially on the net. But much like S and other such languages, it is not used–at least not directly (though SAS may have been rewritten in C by now)–it isn't used for that much statistical analysis. That's not to say one couldn't do so, just that isn't nearly as compelling a language for such purposes as it is for systems work.

SAS if the IBM of statistical software, and everyone else are the dwarves. I don't think that changes until there is some major disruption in the current models of information technology and data analysis. Of course, such disruptions are visible only in hindsight, and I'm hard pressed to see how software along could lead to such disruption. It took rashy personal computers, the TRS-80, the IMSAI 8800 and the Altair 8080 (I speak from personal experience as to how little they could do) to kick off the PC disruption, which itself took a decade to manifest. That was a hardware-led change, much as the 360/370 mainframes and the VAX mini disruptions were hardware-based. I don't see the hardware change leading to a disruption here. Perhaps it's the net itself, but I have my doubts that that's what's unfolding with R. For the net, though, if it is disruptive in software, I would expect to see it in C–and I think SAS is already there.





Speak your mind

1 Comment so far

  1. Alice Allen on September 22, 2012 9:38 pm

    I completely agree that SAS is the “IBM of statistical software, and everyone else are the dwarves,” and I don’t see that position changing either for many reasons, primarily their unlimited dedication to producing quality products that I have seen first-hand.
    But the barrier to SAS use becoming widespread on the free internet is the higher cost of entry (unless you are a collete student under an academic license). As a private company, SAS has to charge to continue to develop and expand their unexcelled software applications.
    Of course, Google is going to be more interested in promoting R and working with R users because the software is free to adapt to Google’s applications and there are no barriers to its adoption: new users can download it any day for free.
    A disclosure note: I had the opportunity to start and lead the technical writing department at SAS Institute from 1980-1988.


Resources & Links