Alexander Burger <[email protected]> writes:
Hi Alex,
> On Sun, Aug 12, 2012 at 08:19:41AM +0200, Alexander Burger wrote:
>> I would recommend in general to call (load "@lib/math.l") before any
>> floating point math, which sets the scale to 6 by default (unless it is
>> set to another value beforehand).
ok
> And, one more point: 'sqrt' expects a double, so you must pass a cons
> pair with the scale. In total, this should work (I didn't try, though):
>
> (load "@lib/math.l")
done
> (native "libR.so" "sqrt" 1.0 (16.0 . 1.0))
this works, thanks.
,---------------------------------------------
| : (native "libR.so" "sqrt" 1.0 (16.0 . 1.0))
| -> 4
| : (native "libR.so" "sqrt" 1.0 (4.0 . 1.0))
| -> 2
`---------------------------------------------
> This will probably return 4000000 (4.0 as a scaled fixpoint number). To
> get more readable results, try
>
> (round (native "libR.so" "sqrt" 1.0 (16.0 . 1.0)))
.. it looks good even without that.
As a non-C programmer I look at the sources of R and wonder:
1. are .h files (optional) interfaces to .c files, and if they exist, all
I would need to define the native calls?
2. there are many functions in R that look like this:
,--------------------------------
| SEXP Invtrans(SEXP pG, SEXP x);
`--------------------------------
taking two lists, returning one list?
3. what about the many functions of R written in R itself, e.g.
,---------------------------------------------------------------
| mean.default <- function(x, trim = 0, na.rm = FALSE, ...){}
`---------------------------------------------------------------
There is a system of generic functions a bit similar to the CLOS system
(dispatching on all arguments). Can these functions be called via
'native'?
4. I guess for operators like '+' (1 + 2) there must be a
function call like this somewhere +(1,2) that must be used by 'native'?
--
cheers,
Thorsten
--
UNSUBSCRIBE: mailto:[email protected]?subject=Unsubscribe
