In the last Utah R Users group meeting I gave a presentation on data manipulations on R, and today I found through the plyr mailing list two commands that I was previously unaware of that should definitely be made mention of.

arrange

I was very pleased to find arrange because it fills the nagging hole for sorting data frames.  Calling

arrange(df, var1, var2)

is much better than calling

df[order(df$var1, df$var2),]

because it’s understandable by practically anyone, and when your code is understandable there is less chance of mistakes.

mutate

mutate is not that different from transform, but I have to make the confession that when I was setting things up for my presentation I tried to see if transform could do the things that mutate does.  mutate can include previously defined variables in later defined variables.  I quote from the mutate help file,

# Things transform can't do
mutate(airquality, Temp = (Temp - 32) / 1.8, OzT = Ozone / Temp)

Notice that temp is first defined then used.  Usually when I need to do something like that I resort to using within, but hopefully I will have to do that less now.