We need to fond the right compromise between bloating Rcpp (which is
already quite huge:
wc src/* inst/include/**/** inst/include/* 2> /dev/null | tail -n1
66180 784183 6425152 total
and support generic enough things.
I can see things like union and setdiff being generic enough (we already
Hi all,
I'm working on a description of how missing values work in Rcpp
(expanding on FAQ 3.4). I'd really appreciate any comments,
corrections or suggestions on the text below.
Thanks!
Hadley
# Missing values
If you're working with missing values, you need to know two things:
* what happen
> > I mean, the .h and *.R files generated with Rcpp.package.skeleton seem to
> > have a simple structure that I find easy to replicate and modify as I keep
> > adding stuff to my *.cpp file. (For instance, the placement of RccpExport
> > before functions intended to be exported, the usage of S
I think it is just I did something wrong with installing Rcpp from
source, which I cannot reproduce now.
Thanks very much.
--
Jiqiang
On Thu, Nov 15, 2012 at 1:55 PM, Dirk Eddelbuettel wrote:
> Possibly a non-standard Rcpp installation?
___
Rcpp-devel
> 1. Suggestion: it might be appropriate to say (probably before 5.1) that
> Makevars, etc, are still needed. And, thus, point the reader to
> vignette("Rcpp-package").
>
Very good suggestion -- will do!
> I mean, the .h and *.R files generated with Rcpp.package.skeleton seem to
> have a simple
Dear list
[>>] I am not sure if Hadleys remark below was an invitation to make a
"wishlish", but I'll take the risk:
1) I have made several packages related to graphical models for multivariate
data. Much of these packages deals with "book keeping": operations on sets of
subsets of a finite s
Dear Dirk, Romain, and JJ,
Version 0.10 is really amazing. In particular, the Rcpp Attributes are
awesome, as well as some of the additions to sugar!!
If I may, I'd like to make a suggestion and a question about the
documentation for Rcpp Attributes.
1. Suggestion: it might be appropriate to
On 15 November 2012 at 13:42, JJ Allaire wrote:
| The error appears to be occurring in the call to RcppLdPath (made by both
| sourceCpp and RStan). Here's the source code for that function:
|
| RcppLdPath <- function() {
| if (nzchar(.Platform$r_arch)) { ## eg amd64, ia64, mips
| path
On 15 November 2012 at 13:40, Jiqiang Guo wrote:
| Thanks.
|
| The same error if I use cppFunction thought the example in sourceCpp
| use a string of c++ code passed using code:
Oh, sorry, my bad -- was short of time and couldn;t test, and don't even have
a mac. I will let the mac squad get to
I see. For some reason, my installing Rcpp from source last time did
not install the 64 bits lib, so
Rcpp.system.file("lib", .Platform$r_arch) for me returns "".
I reinstalled Rcpp 0.10.0 from source and made sure lib/x86_64 exists, then
everything is fine.
Thanks.
--
Jiqiang
On Thu, Nov 15, 2
The error appears to be occurring in the call to RcppLdPath (made by both
sourceCpp and RStan). Here's the source code for that function:
RcppLdPath <- function() {
if (nzchar(.Platform$r_arch)) { ## eg amd64, ia64, mips
path <- Rcpp.system.file("lib",.Platform$r_arch)
} else {
Thanks.
The same error if I use cppFunction thought the example in sourceCpp
use a string of c++ code passed using code:
> require(Rcpp)
>
> cppFunction(code='
+#include
+
+// [[Rcpp::export]]
+int fibonacci(const int x) {
+ if (x == 0) return(0);
+
You can make the differences more profound if you choose examples
where range.default does really badly (assuming your implementation is
similar to my attached implementation)
expr min lq median uq max neval
range(x, na.rm = TRUE) 1920 2488 2951 3349 23306 100
min
On 15 November 2012 at 13:20, Jiqiang Guo wrote:
| I tried the example of sourceCpp on Mac mountain lion with R 2.15.1
| with the following error (I hit the similar error in using RStan as
| well).
|
| > library(Rcpp)
| > require(Rcpp)
| >
| > sourceCpp(code='
Operator error: sourceCpp() t
Thanks.
But I am using R from command line, and I think it is set correctly:
> Sys.getenv("R_ARCH")
[1] "/x86_64"
--
Jiqiang
On Thu, Nov 15, 2012 at 1:30 PM, JJ Allaire wrote:
>
>
> On Thu, Nov 15, 2012 at 1:20 PM, Jiqiang Guo wrote:
>>
>> I tried the example of sourceCpp on Mac mountain lion
On Thu, Nov 15, 2012 at 1:20 PM, Jiqiang Guo wrote:
> I tried the example of sourceCpp on Mac mountain lion with R 2.15.1
> with the following error (I hit the similar error in using RStan as
> well).
>
I think the common problem here might be whether R_ARCH is set correctly.
On the Mac R_ARCH n
I tried the example of sourceCpp on Mac mountain lion with R 2.15.1
with the following error (I hit the similar error in using RStan as
well).
> library(Rcpp)
> require(Rcpp)
>
> sourceCpp(code='
+#include
+
+// [[Rcpp::export]]
+int fibonacci(const int x) {
+
On 15 November 2012 at 08:40, Karl Millar wrote:
| Yes. I think only fairly recent versions of clang pick up on this though.
Cool. Thanks for the feedback. At the top-secret "Rcpp R+D labs" someone is
already plotting to get some of the clang goodness into Rcpp and its toolchain.
Dirk
--
Dir
Le 15/11/12 17:52, Romain Francois a écrit :
Hello,
I've commited an Rcpp version of %in%.
For example:
require(Rcpp)
require(microbenchmark)
sourceCpp( code = '
#include
using namespace Rcpp ;
// [[Rcpp::export]]
LogicalVector in_( CharacterVector x, CharacterVector table){
return in(
Hello,
I've commited an Rcpp version of %in%.
For example:
require(Rcpp)
require(microbenchmark)
sourceCpp( code = '
#include
using namespace Rcpp ;
// [[Rcpp::export]]
LogicalVector in_( CharacterVector x, CharacterVector table){
return in( x, table ) ;
}
' )
`%in++%` <- in_
> c("a",
Yes. I think only fairly recent versions of clang pick up on this though.
On Thu, Nov 15, 2012 at 4:34 AM, Dirk Eddelbuettel wrote:
>
> On 14 November 2012 at 21:57, Karl Millar wrote:
> | In the Rcpp 0.10.0,
> inst/include/Rcpp/sugar/logical/SingleLogicalResult.h there
> | is the code:
> |
>
On 15 November 2012 at 16:45, Willem Ligtenberg wrote:
| No need to jump on me.
| I didn't know that Rcpp lived on R-Forge, but that is mostly because all of
the
| info is found on your site:
| http://dirk.eddelbuettel.com/code/rcpp.html
Well the problem is that
a) we have too much document
No need to jump on me.
I didn't know that Rcpp lived on R-Forge, but that is mostly because all of
the info is found on your site:
http://dirk.eddelbuettel.com/code/rcpp.html
I will gladly use R-forge for this.
(for the list etiquette I am blaiming gmail, it truncated the other
messages...)
On T
On 15 November 2012 at 16:02, Willem Ligtenberg wrote:
| and now to the list...
|
| Nothing specific, but some way of sharing would be nice. :)
Funny how we managed to do that "sharing thing" for almost 4000 commits
without using github...
Seriously, this has been on R-Forge for 4 1/2 years, an
and now to the list...
Nothing specific, but some way of sharing would be nice. :)
On Thu, Nov 15, 2012 at 3:58 PM, Romain Francois
wrote:
> Help is very welcome. Why github though ?
>
> Le 15/11/12 15:42, Willem Ligtenberg a écrit :
>
>> Would be nice to keep the list and the progress on githu
Help is very welcome. Why github though ?
Le 15/11/12 15:42, Willem Ligtenberg a écrit :
Would be nice to keep the list and the progress on github.
Other might pitch in and help out, I am also interested, just don't know
how much time I can spend on it.
Kind regards,
Willem
On Thu, Nov 15, 2
Would be nice to keep the list and the progress on github.
Other might pitch in and help out, I am also interested, just don't know
how much time I can spend on it.
Kind regards,
Willem
On Thu, Nov 15, 2012 at 3:24 PM, Romain Francois
wrote:
> Le 15/11/12 14:52, Hadley Wickham a écrit :
>
> R
On 15 November 2012 at 07:49, Hadley Wickham wrote:
| I didn't, but I'm looking at it now, thanks. The main problem is that
| many of the examples use C++ idioms I don't understand - and as I
That is *the whole point* -- C++, not unlike R, is a version of nethack:
Enter the dungeon, and find the
>> PS. Would you be interested in a set of R functions that from a quick
>> skim of the R sources that I think could be much much faster if
>> implemented in Rcpp?
>
> Sure. Time I have available will decide what I do of the list ^^.
Here's the list:
rowsum.default (C)?
findInterval (C)?
I'm not sure what compliers you are using out there, but gcc does have a way to
enable warnings for this type of implicit conversion.
-Wconversion
Warn for implicit conversions that may alter a value. This includes conversions
between real and integer...
http://gcc.gnu.org/onlinedocs/gcc/Warnin
Le 15/11/12 14:52, Hadley Wickham a écrit :
R
utilisateur système écoulé
0.338 0.024 0.362
R (second time)
utilisateur système écoulé
0.262 0.000 0.262
match_ manual Rcpp
utilisateur système écoulé
0.144 0.000 0.1
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Just to add one more test here:
Unit: microseconds
exprmin lq median uq max
1 sum1(x) 19.835 20.0445 20.184 20.3585 43.651
2 sum2a(x) 19.835 20.0095 20.184 20.3240 48.958
3 sum2(x) 93.797 94.0760 94.146 94.4250 204.355
4 s
Le 15/11/12 15:14, Hadley Wickham a écrit :
On Thu, Nov 15, 2012 at 8:11 AM, Romain Francois
wrote:
I'm not getting the same results, so I might not make the same conclusions:
Unit: microseconds
exprmin lq median uqmax
1 sum(x) 16.072 16.161 16.2090 16.3110 21.721
2
On Thu, Nov 15, 2012 at 8:11 AM, Romain Francois
wrote:
> I'm not getting the same results, so I might not make the same conclusions:
>
>
> Unit: microseconds
> exprmin lq median uqmax
> 1 sum(x) 16.072 16.161 16.2090 16.3110 21.721
> 2 sum1(x) 10.771 10.954 11.0445 11.1
I'm not getting the same results, so I might not make the same conclusions:
Unit: microseconds
exprmin lq median uqmax
1 sum(x) 16.072 16.161 16.2090 16.3110 21.721
2 sum1(x) 10.771 10.954 11.0445 11.1910 16.163
3 sum2(x) 48.438 48.567 48.6650 48.8425 61.116
4 sum2a(x)
> R
> utilisateur système écoulé
> 0.338 0.024 0.362
> R (second time)
> utilisateur système écoulé
> 0.262 0.000 0.262
> match_ manual Rcpp
> utilisateur système écoulé
> 0.144 0.000 0.144
> match__ internal Rcpp
> ut
> | I'm trying to get a better handle on writing efficient numerical code
> | with C++. I've attached five implementations of a simple sum function
> | (mostly ignoring NAs) and timed them below:
>
> Did you see the (fairly extensive) comparison in
> examples/ConvolveBenchmarks/
> which comp
On 15 November 2012 at 07:24, Hadley Wickham wrote:
| Hi all,
|
| I'm trying to get a better handle on writing efficient numerical code
| with C++. I've attached five implementations of a simple sum function
| (mostly ignoring NAs) and timed them below:
Did you see the (fairly extensive) compar
Commited in revision 3973. I added a few abstraction that let me
eventually write the code in terms of stl algorithms.
See the implementation in match.h or the curious.
With that:
require(Rcpp)
sourceCpp( code = '
#include
using namespace Rcpp ;
// [[Rcpp::export]]
IntegerVector match__( Ch
Hi all,
I'm trying to get a better handle on writing efficient numerical code
with C++. I've attached five implementations of a simple sum function
(mostly ignoring NAs) and timed them below:
> source("sum.r")
Unit: microseconds
expr minlq medianuq max neval
sum(x) 11.67 11.72
On 15 November 2012 at 04:00, Christian Gunning wrote:
| | Also, you may need to add RNGScope() if you use R's RNG.
| |
| |
| |
| | Yes, I should have mentioned that in my post. Calling RNGScope() is
left
| to
| | the user, since multiple calls to sample() can con
On 14 November 2012 at 21:57, Karl Millar wrote:
| In the Rcpp 0.10.0, inst/include/Rcpp/sugar/logical/SingleLogicalResult.h
there
| is the code:
|
| template
| class conversion_to_bool_is_forbidden :
| conversion_to_bool_is_forbidden{
|
| Which obviously can't be instantiated.
|
| Presumabl
>
> | Also, you may need to add RNGScope() if you use R's RNG.
> |
> |
> |
> | Yes, I should have mentioned that in my post. Calling RNGScope() is
> left to
> | the user, since multiple calls to sample() can conceivably happen inside
> a
> | single RNGScope().
>
> With that, maybe best to be i
43 matches
Mail list logo
