[R] flexible processing

[Antje]

Hello,

unfortunately, I don't know a better subject. I would like to be very flexible 
in how to process my data.
Assume the following dataset:

par1 <- seq(0,1,length.out = 100)
par2 <- seq(1,100)
fac1 <- factor(rep(c("group1", "group2"), each = 50))
fac2 <- factor(rep(c("group3", "group4", "group5", "group6"), each = 25))

df <- data.frame(par1, par2, fac1, fac2)

Now, I would like to calculate e.g. the "sum" for par1 grouping by "fac1" and 
the "mean" and "sd" for par2 grouping by "fac1".

I would like to determine this method as a string.

In the end, I would like to have something like this:

fac1.analysis:

       par1.sum    par2.mean     par2.sd
group1    xxx           xxx         xxx
group2    xxx           xxx         xxx

fac2.analysis:

...

Does anybody have any idea how to realize it?

Antje