Rcpp class in Sat 9 Mar in NYC

Join Dirk Eddelbuettel for six hours of detailed and hands-on instructions and discussions around Rcpp, RInside, RcppArmadillo, RcppGSL and other packages . . . Rcpp has become the most widely-used language extension for R. Currently deployed by 103 CRAN packages and a further 10 BioConductor packages, it permits users and developers to pass “whole R objects” with ease between R and C++ . . . Morning session: “A Hands-on Introduction to R and C++” . . . Afternoon session: “Advanced R and C++ Topics” . . .

10 thoughts on “Rcpp class in Sat 9 Mar in NYC

    • Joerg:

      Yes, it’s great to see people making money off free software. I hope this sort of success encourages others to make contributions. It’s a great model.

      • I understand the sentiment, but I paid for the workshop so maybe I can contribute an answer. I am a biologist with some decent C++/R background and I’ve used that knowledge to implement MCMC algorithms with R as the glue language. I am using Rcpp for the interface to R and while it’s generally well documented (and easier on me than the C interface) there are parts which are not documented in a way I can understand yet. I think that getting good information on some best practices for using Rcpp, getting my technical questions answered, getting an orientation to the code/documentation, and getting the mental model for how R and Rcpp/C++ code interact right are well worth the fee. If I were a professional C++ programmer and knew R… it might be a different story, but as it is, I’ll take the workshop at the current fee. I’m not sure about the motivation for the other participants… maybe some of them read this blog as well?

        • I see. Thanks for elaborating on the motivation.

          The one part that does seem most useful is the “getting my technical questions answered”. That’s often hard and with sufficient homework done might make good use of six hours with an expert.

    • Yes, of course, but:

      a) Don Rubin’s hourly rate also may look less impressive if you factor in all the years that he took to build his expertise and reputation.

      b) There could be other people than the package author him/herself (Dirk Eddelbuettel) who could give that workshop and people would probably still cough up that money (or so I believe). In other words, hours needed for development might not even be a useful offset here.

      Many people develop software for free, often very time-intensive stuff, without ever making any money with it. And after all, it’s not the primary goal of the usual package developer (speaking mainly about the Stata and R community here) to make money with it, right? Programming is fun, helps with everyday tasks, and may provide some symbolic rewards. If then, as a bonus, you can even get some material rewards, so much the better. I merely found it surprising that the rewards could be that high that they compare (well, sort of) to what an industry with an essentially infinite amount of money pays for the works of a truly outstanding scientist.

      • Joerg:

        I agree. I’d guess that people will continue to work on freeware for the fun of it, or because they need it for their own research, or as a way of participating in a larger community, or to supply a public good, or even to get the exposure and get their name out there. If people can make some money teaching it, that’s one more incentive, and I think that’s a good thing.

  1. It is a lot of bucks. But all the invites I ever got for workshops on doing cool stuff with mathematica, matlab etc. were in that range or even (much) higher.

    Chances are that you learn A LOT in that one day. After all the guy is the expert on the topic. But probably you need to have the appropriate profile to get the most out if it e.g. I would not get much out of it I guess…

    Moreover, can’t we be sympathetic for a guy spending his time developing open source stuff to earn a bit?

    • I’ll add a dissenting voice (just personal opinion though!): I’ve often been tempted by these similar software workshops but the end result has mostly been underwhelming. I’ve reached the conclusion that learning software is a curve needing lots of tinkering, man page reading etc. that’s best achieved asynchronously. Luckily the net offers tons of tutorials, howtos, FAQs etc.

      Perhaps it is a personal learning style issue. The two places a live workshop does seem to work for me are: (1) A teaser session that gets me hooked enough to explore a new package or class of packages. (2) QA with an expert for code I’ve mastered but have more advanced / specific questions and bottlenecks in.

      No offence to Mr. Eddelbuettel!

Comments are closed.