« Training students on mega-scale data | Main | Because it's Friday: If lottery tickets told the truth »

October 15, 2009


Feed You can follow this conversation by subscribing to the comment feed for this post.

David, you gave a great talk yourself. Also, the reason there were absolutely no debugging hints in my list is: prior to seeing your demo I was fairly certain there was no point in using R's debugging facilities (the impossible task (the nearly impossible task of "just getting it right" seemed easier than dealing with the standard R debugging tools). After seeing your demo and product I am pretty excited about adding interactive debugging to my R programming repertoire.

I'm pleased to hear about more and better debugging utilities making their way towards R's core, and I think that overall REvolution Computing's dual-level model of using paid support to fund open source development is a great one. A similar system has driven postgresql's development rapidly forwards, for example.

At present, my most effective bug-fighting ally is setting error and warning options in .Rprofile. For identifying typos and simple confusions, an inspection of the traceback is usually sufficient:

options(error=recover) ## requires utils package
options(warn=2) ## sometimes this is overkill, but often invaluable

Your post is really interesting but I enjoyed reading comments on your post more than your post. It's true. I really appreciate your work and appreciate you.

The comments to this entry are closed.

Search Revolutions Blog

Got comments or suggestions for the blog editor?
Email David Smith.
Follow revodavid on Twitter Follow David on Twitter: @revodavid
Get this blog via email with Blogtrottr