[Rcpp-commits] r3306 - pkg/Rcpp/inst/doc/Rcpp-FAQ

[noreply at r-forge.r-project.org]

Author: romain
Date: 2011-11-07 13:22:39 +0100 (Mon, 07 Nov 2011)
New Revision: 3306

Modified:
   pkg/Rcpp/inst/doc/Rcpp-FAQ/Rcpp-FAQ.Rnw
Log:
update the vignette to show int64_t usage

Modified: pkg/Rcpp/inst/doc/Rcpp-FAQ/Rcpp-FAQ.Rnw
===================================================================
--- pkg/Rcpp/inst/doc/Rcpp-FAQ/Rcpp-FAQ.Rnw	2011-11-07 12:11:23 UTC (rev 3305)
+++ pkg/Rcpp/inst/doc/Rcpp-FAQ/Rcpp-FAQ.Rnw	2011-11-07 12:22:39 UTC (rev 3306)
@@ -583,40 +583,33 @@
 @
 %
 
-\subsection{Why can long long types not be cast correctly?}
+\subsection{Can I use 64 bits integer types ?}
 
-That is a good and open question. We rely on the basic \proglang{R} types,
-notably \code{integer} and \code{numeric}.  These can be cast to and from
-\proglang{C++} types without problems.  But there are corner cases.  The
-following example, contributed by a user, shows that we cannot reliably cast
-\code{long} types (on a 64-bit machines).
+Yes. Since version 0.9.8, \pkg{Rcpp} depends on the \pkg{int64}
+package. \pkg{int64} brings 64 bit integer vectors to R, with the 
+two S4 classes \code{int64} and \code{uint64} that internally 
+handle C++ types \code{int64\_t} and \code{uint64\_t}. 
 
-<<eval=FALSE>>=
+\pkg{Rcpp}'s \code{wrap} function is able to handle the 
+primitives \code{int64\_t} and \code{uint64\_t} as well as containers
+of these: 
+
+<<>>=
 BigInts <- cxxfunction(signature(),
-  'std::vector<long> bigints;
+  'std::vector<int64_t> bigints;
    bigints.push_back(12345678901234567LL);
    bigints.push_back(12345678901234568LL);
    Rprintf("Difference of %ld\\n", 12345678901234568LL - 12345678901234567LL);
-   return wrap(bigints);', plugin="Rcpp", includes="#include <vector>")
+   return wrap(bigints);', 
+plugin="Rcpp" )
 
 retval<-BigInts()
-
-# Unique 64-bit integers were cast to identical lower precision numerics
-# behind my back with no warnings or errors whatsoever.  Error.
-
-stopifnot(length(unique(retval)) == 2)
+retval
 @
-%
 
-While the difference of one is evident at the \proglang{C++} level, it is no
-longer present once cast to \proglang{R}. The 64-bit integer values get cast
-to a floating point types with a 53-bit mantissa. We do not have a good
-suggestion or fix for casting 64-bit integer values: 32-bit integer values
-fit into \code{integer} types, up to 53 bit precision fits into
-\code{numeric} and beyond that truly large integers may have to converted
-(rather crudely) to text and re-parsed. Using a different representation as
-for example from the \href{http://gmplib.org/}{GNU Multiple Precision Arithmetic
-  Library} may be an alternative.
+The vignette in the \pkg{int64} package has the details on how 64 bit 
+integer types are handled. In short, we split the higher bits and the lower
+into two integers. 
 
 \section{Support}