Comments for BabelGraph http://www.babelgraph.org/wp health care, medicine, R, social network analysis, random nerdy stuff Wed, 19 Sep 2012 08:48:27 +0000 hourly 1 http://wordpress.org/?v=3.4.2 Comment on Rcpp is smoking fast for agent-based models in data frames by icecoldo http://www.babelgraph.org/wp/?p=358#comment-236 icecoldo Wed, 19 Sep 2012 08:48:27 +0000 http://www.babelgraph.org/wp/?p=358#comment-236 Thanks for a great article, Error in compileCode: I also got this nasty error, but fixed it with proper path's, thanks to this awesome guide: http://tonybreyal.wordpress.com/2011/12/07/installing-rcpp-on-windows-7-for-r-and-c-integration/#comment-3152 Thanks for a great article,

Error in compileCode: I also got this nasty error, but fixed it with proper path’s, thanks to this awesome guide:

http://tonybreyal.wordpress.com/2011/12/07/installing-rcpp-on-windows-7-for-r-and-c-integration/#comment-3152

]]> Comment on Rcpp is smoking fast for agent-based models in data frames by Simono101 http://www.babelgraph.org/wp/?p=358#comment-203 Simono101 Thu, 19 Jul 2012 08:45:56 +0000 http://www.babelgraph.org/wp/?p=358#comment-203 Hi Josh, Thanks. I tried your suggestion and I got a different problem to you! First the compilation error: <code>Error in compileCode(f, code, language = language, verbose = verbose) : Compilation ERROR, function(s)/method(s) not created! cygwin warning: </code> Then some warnings about msdos style paths. Then the relevant error messages, about comparison between signed and unsigned integers: <code>file1fdc365c273.cpp:1:0: sorry, unimplemented: 64-bit mode not compiled in file1fdc365c273.cpp: In function 'SEXPREC* file1fdc365c273(SEXPREC*)': file1fdc365c273.cpp:58:38: warning: comparison between signed and unsigned integer expressions make: *** [file1fdc365c273.o] Error 1 </code> Hi Josh,
Thanks. I tried your suggestion and I got a different problem to you!
First the compilation error:
Error in compileCode(f, code, language = language, verbose = verbose) :
Compilation ERROR, function(s)/method(s) not created! cygwin warning:

Then some warnings about msdos style paths.
Then the relevant error messages, about comparison between signed and unsigned integers:
file1fdc365c273.cpp:1:0: sorry, unimplemented: 64-bit mode not compiled in
file1fdc365c273.cpp: In function 'SEXPREC* file1fdc365c273(SEXPREC*)':
file1fdc365c273.cpp:58:38: warning: comparison between signed and unsigned integer expressions
make: *** [file1fdc365c273.o] Error 1

]]> Comment on Rcpp is smoking fast for agent-based models in data frames by Joshua Ulrich http://www.babelgraph.org/wp/?p=358#comment-202 Joshua Ulrich Mon, 16 Jul 2012 15:35:06 +0000 http://www.babelgraph.org/wp/?p=358#comment-202 Simon, I had to replace <code>p = max(0,p); p = min(1,p);</code> with <code>p = max(0.0,p); p = min(1.0,p);</code>. Even then, I got an error regarding how the C++ dataframe was being constructed and a segfault if I tried to re-run the code that threw the error. FWIW, Dirk couldn't replicate my issue, so you may not have a problem. Simon,
I had to replace p = max(0,p); p = min(1,p); with p = max(0.0,p); p = min(1.0,p);. Even then, I got an error regarding how the C++ dataframe was being constructed and a segfault if I tried to re-run the code that threw the error. FWIW, Dirk couldn’t replicate my issue, so you may not have a problem.

]]> Comment on Rcpp is smoking fast for agent-based models in data frames by Simono101 http://www.babelgraph.org/wp/?p=358#comment-201 Simono101 Sat, 14 Jul 2012 10:55:14 +0000 http://www.babelgraph.org/wp/?p=358#comment-201 Hi Gary, Thanks for the advice. That has changed things a bit. Either of your solutions now throws up a different error (the same in either case), which is: <code>file1f382c842ee5.cpp:28:25: error: no matching function for call to 'max(int, double&)' file1f382c842ee5.cpp:28:44: error: no matching function for call to 'min(int, double&)' </code> My guess is it's something to do with the variable type which is being passed to <code>min</code> and <code>max</code> functions? Cheers, Simon Hi Gary,
Thanks for the advice. That has changed things a bit. Either of your solutions now throws up a different error (the same in either case), which is:
file1f382c842ee5.cpp:28:25: error: no matching function for call to 'max(int, double&)'
file1f382c842ee5.cpp:28:44: error: no matching function for call to 'min(int, double&)'

My guess is it’s something to do with the variable type which is being passed to min and max functions?

Cheers, Simon

]]> Comment on Rcpp is smoking fast for agent-based models in data frames by gary http://www.babelgraph.org/wp/?p=358#comment-200 gary Sat, 14 Jul 2012 00:54:59 +0000 http://www.babelgraph.org/wp/?p=358#comment-200 Exactly! Thanks ;-) Exactly! Thanks ;-)

]]> Comment on Rcpp is smoking fast for agent-based models in data frames by gary http://www.babelgraph.org/wp/?p=358#comment-199 gary Sat, 14 Jul 2012 00:53:42 +0000 http://www.babelgraph.org/wp/?p=358#comment-199 Hi Simon, You might try replacing <code>max</code> with <code>std::max</code> and doing the same for <code>min</code>. Alternatively, try placing <code>using namespace std;</code> at the beginning of your source. Hope that helps! Hi Simon,
You might try replacing max with std::max and doing the same for min. Alternatively, try placing using namespace std; at the beginning of your source. Hope that helps!

]]> Comment on Rcpp is smoking fast for agent-based models in data frames by gary http://www.babelgraph.org/wp/?p=358#comment-198 gary Sat, 14 Jul 2012 00:44:03 +0000 http://www.babelgraph.org/wp/?p=358#comment-198 To actually compare <i>significance</i> it would be interesting if the <code>rbenchmark</code> package also included information about the distribution of the trial times, such as the standard deviation. To actually compare significance it would be interesting if the rbenchmark package also included information about the distribution of the trial times, such as the standard deviation.

]]> Comment on Rcpp is smoking fast for agent-based models in data frames by gary http://www.babelgraph.org/wp/?p=358#comment-197 gary Sat, 14 Jul 2012 00:40:02 +0000 http://www.babelgraph.org/wp/?p=358#comment-197 A great point about vectorization in R that I hadn't considered (also made below many of the posters). The example above does lend itself to vectorization, but many agent-based models do not since there are dependencies or other dynamic computations happening between agents that would not be vectorizable... <code>Rcpp</code> fills that particular niche. A great point about vectorization in R that I hadn’t considered (also made below many of the posters). The example above does lend itself to vectorization, but many agent-based models do not since there are dependencies or other dynamic computations happening between agents that would not be vectorizable… Rcpp fills that particular niche.

]]> Comment on Rcpp is smoking fast for agent-based models in data frames by gary http://www.babelgraph.org/wp/?p=358#comment-196 gary Sat, 14 Jul 2012 00:37:25 +0000 http://www.babelgraph.org/wp/?p=358#comment-196 thanks for writing a great package! thanks for writing a great package!

]]> Comment on Rcpp is smoking fast for agent-based models in data frames by Simono101 http://www.babelgraph.org/wp/?p=358#comment-195 Simono101 Fri, 13 Jul 2012 11:46:52 +0000 http://www.babelgraph.org/wp/?p=358#comment-195 Hi, Thanks for this, I found it very interesting. I should say that I was not able to compile your C++ code (copied and pasted from the site) to test this myself (I wanted to have a look and see if it made any difference in speed sorting the cohort data.frame by sex for a much larger cohort). When I tried to compile the code I got the following errors: <code>Error in compileCode(f, code, language = language, verbose = verbose) : Compilation ERROR, function(s)/method(s) not created! cygwin warning: MS-DOS style path detected: C:/PROGRA~1/R/R-215~1.1/etc/x64/Makeconf Preferred POSIX equivalent is: /cygdrive/c/PROGRA~1/R/R-215~1.1/etc/x64/Makeconf CYGWIN environment variable option "nodosfilewarning" turns off this warning. Consult the user's guide for more details about POSIX paths: http://cygwin.com/cygwin-ug-net/using.html#using-pathnames fileabc1779658e.cpp:1:0: sorry, unimplemented: 64-bit mode not compiled in fileabc1779658e.cpp: In function 'double vaccinate_cxx(double, int, int)': fileabc1779658e.cpp:27:20: error: 'max' was not declared in this scope fileabc1779658e.cpp:27:34: error: 'min' was not declared in this scope fileabc1779658e.cpp: In function 'SEXPREC* fileabc1779658e(SEXPREC*)': fileabc1779658e.cpp:57:38: warning: comparison between signed and unsigned integer expressions make: *** [fileabc1779658e.o] Error 1 In addition: Warning message: running command 'C:/PROGRA~1/R/R-215~1.1/bin/x64/R CMD SHLIB fileabc1779658e.cpp 2> fileabc1779658e.cpp.err.txt' had status 1 </code> Can anyone point me in the direction of why this might be occuring (sorry - I am just starting out in C++ and using Rcpp). Hi,
Thanks for this, I found it very interesting. I should say that I was not able to compile your C++ code (copied and pasted from the site) to test this myself (I wanted to have a look and see if it made any difference in speed sorting the cohort data.frame by sex for a much larger cohort). When I tried to compile the code I got the following errors:
Error in compileCode(f, code, language = language, verbose = verbose) :
Compilation ERROR, function(s)/method(s) not created! cygwin warning:
MS-DOS style path detected: C:/PROGRA~1/R/R-215~1.1/etc/x64/Makeconf
Preferred POSIX equivalent is: /cygdrive/c/PROGRA~1/R/R-215~1.1/etc/x64/Makeconf
CYGWIN environment variable option "nodosfilewarning" turns off this warning.
Consult the user's guide for more details about POSIX paths:
http://cygwin.com/cygwin-ug-net/using.html#using-pathnames
fileabc1779658e.cpp:1:0: sorry, unimplemented: 64-bit mode not compiled in
fileabc1779658e.cpp: In function 'double vaccinate_cxx(double, int, int)':
fileabc1779658e.cpp:27:20: error: 'max' was not declared in this scope
fileabc1779658e.cpp:27:34: error: 'min' was not declared in this scope
fileabc1779658e.cpp: In function 'SEXPREC* fileabc1779658e(SEXPREC*)':
fileabc1779658e.cpp:57:38: warning: comparison between signed and unsigned integer expressions
make: *** [fileabc1779658e.o] Error 1
In addition: Warning message:
running command 'C:/PROGRA~1/R/R-215~1.1/bin/x64/R CMD SHLIB fileabc1779658e.cpp 2> fileabc1779658e.cpp.err.txt' had status 1

Can anyone point me in the direction of why this might be occuring (sorry – I am just starting out in C++ and using Rcpp).

]]>