R without Hadley Wickham

Tim Hopper asked on Twitter today:

#rstats programming without @hadleywickham’s libraries is like ________ without _________.

Some of the replies were:

  • (skydiving, a parachute)
  • (gouging your own eyes out, NULL)
  • (dentistry, anesthesia)
  • (shaving, a razor)
  • (Internet shopping, credit card)

Clearly there’s a lot of love out there for Hadley Wickham’s R packages. I’m familiar with his ggplot2 graphics package, and it’s quite impressive. I need to look at his other packages as well.

6 thoughts on “R without Hadley Wickham

  1. plyr is an essential; it improves the basic ‘apply’ functions by adding progress bars and parallelisation (via another package…)

  2. I don’t know about the devtools since I don’t write packages. Otherwise, I do a lot of R coding completely without plyr and ggplot2. I’ve generally found them slow (plyr) and inflexible (ggplot2). I understand ggplot2 has become more flexible but I’ve already written my own plot functions that match templates for publication so going back and learning ggplot2 won’t save me any time at this point.

    That said, I do now use qplot() instead of lattice functions when exploring data.

    Also, they’re difficult to understand for the absolute beginner, not necessarily on their own, but because the dot notation syntax and other features are fairly unique to those packages. This is true of the ez package as well, which is modelled on them. I avoid telling new learners about them unless they really are excited about watching me do exploratory analysis. It takes a lot of motivation to get them to add that syntax to the varied R dialects across packages.

    (oh, and I know many individual who just couldn’t survive without ez)

  3. Oh, just because you don’t write packages doesn’t mean you shouldn’t use devtools! It lets you do this: stick all your R files for a project in an R directory, then you edit those files, run load_all(), and it loads just the ones that have changed. And into a separate environment so that ls() is kept clear.

    The same thing works for C and Fortran in a src directory. There’s a tiny overhead (you need a five-line DESCRIPTION file and a one-line NAMESPACE file if you have C code).

    Wrote about this here:

    http://rpubs.com/geospacedman/lazydevtools

    As a pleasant side effect, you end up with a workspace that’s much easier to turn into a working package if you want.

  4. Hadley has done a lot for the platform. plyr is essential. ggplot2 is great for making graphics like a statistician.

Comments are closed.