On 23/02/2010 9:35 PM, Ben wrote:
Thanks to everyone for the responses so far. I didn't know about
setBreakpoint, mlocal, or proto before.
I think I'll try using proto a lot more and see if that fixes most of
my problems. It appeals to me for two bad reasons: I had run across
proto when reading the code to ggplot2 earlier but didn't know the
purpose before (and I like ggplot2 and plyr, so Hadley's endorsement
carries some weight with me). Also it uses an explicit passing of the
object to instances of that object---it's similar to Python's self,
and I'm a big fan of Python. Anyway, so much for rationality.
I find it interesting that everyone seems to run into this issue, and
everyone solves it in their own way. I hadn't known that Mark's
mvbutils and Duncan's setBreakpoint (how do I get this function?)
setBreakpoint is in the utils package, as of R 2.10.0. It is just a
front end to trace, which has been in the base package for quite a long
time.
Duncan Murdoch
existed before. I also like Mark Kimpel's trick, which appeals to a
simple-minded person like me.
Either R needs to expand it's standard library, or there should be a
standard guide on what you really need to work productively. I'd have
a hard time in R without ggplot2, plyr, RUnit, perhaps proto now, and
a few other functions and modules, but it's taken a while to figure
out what I needed. It could be in 6 months I'll discover something
else and realize that I've been wasting my time all along.
Anyway, thanks again for all the responses!
______________________________________________
R-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
