R help - May 2008 - question on marginal effects

I have a question regarding calculating marginal effects (change in  
the probability due to a one unit change in x)  for a probit or for  
that matter any discrete choice model. Does R have a function that  
will essentially do this:
     dE[y|x]
     ------- = f(x'beta)*beta
     dx

Stata has a command called "mfx" that calculates the marginal effects
at the mean (of the x values). Is there a R equivalent to mfx?

Thanks,
Renuka.