[R] problem applying .C function in outer product

[Dirk Eddelbuettel]

On Wed, Mar 12, 2003 at 03:02:15PM -0800, Jim McLoughlin wrote:
> The function is loaded, and it works for when given specific (scalar) 
> arguments, but does not work when passed Vector arguments, and 
> therefore fails in the outer product I need.
[...] 
> blackS <- function(S,K,r,T,vol){
> 	tempx <- .C("blackScholesR",
> 	as.double(S),
> 	as.double(K),
> 	as.double(r),
> 	as.double(T),
> 	as.double(vol),
> 	as.double(0.0))
> 	as.double(tempx[6])
> }
> 
> This function returns the appropriate value for constant vals, example:
> 
> > blackS(52,50,0,0.5,0.43)
> [1] 7.213387
> 
> but fails if I use vector arguments for S or T.  I ultimately want to 
> create an outer product varying both S and T and graph the surface.

Allin already pointed out that your function is scalar-valued -- yet you
want a vector context.  No real problem there -- you simply need to provide
it.  

It so happens that I have an example in my small RQuantLib package (which is
essentially a wrapper around some functions of the larger QuantLib library)
that does just that. 

>From example(EuropeanOptionArrays):

   # define two vectors for the underlying and the volatility
   und.seq <- seq(10,180,by=2)
   vol.seq <- seq(0.1,0.9,by=0.1)
   # evaluate them along with three scalar parameters
   EOarr <- EuropeanOptionArrays("call", underlying=und.seq,
                                 strike=100, dividendYield=0.01,
                                 riskFreeRate=0.03, 
                                 maturity=1, volatility=vol.seq)
   [... more code to plot this ...]	   

and you can look how EuropeanOptionArrays is essentially looping over all
possible vector-form arguments to pass them to the scalar function that
calls the corresponding (closed-form) QuantLib pricer.  

Hope this helps, Dirk

-- 
Prediction is very difficult, especially about the future. 
				             -- Niels Bohr