R help - Jan 2002 - trouble using R Mathlib as standalone

Dear People,

I am trying to use R's Math library as standalone, as documented in
/src/nmath/standalone. I am using C++ in Debian testing, and the versions
are as follows:

ii  g++-3.0        3.0.3-1        The GNU C++ compiler.
ii  r-mathlib      1.4.0-1        `GNU S'  - Standalone R math library

I have a file (rand.cc) as follows. I don't think that lattice.hh or mh.hh
are very relevant. The class Cube is defined in lattice.hh and I don't
think mh.hh are used at all. I get the following errors when trying to
compile this. Of course these cannot be reproduced without all my files,
but I am hoping these errors have nothing to do with my code and that the
answer will be obvious to those in the know.

I get two different errors.

The first one I am completely mystified by. It seems to have something to
do with a double declaration of log1p which I have not used and not
knowingly declared twice in any way. Neither lattice.hh nor mh.hh contain
any math headers.

The second one clearly has to do with the usage of GetRNGstate() and
PutRNGstate() which are correct for use as entry points into the R
executable but no longer correct for standalone usage. The README in the
standalone directory said something about using the uniform random number
generator, but gave no examples so I didn't understand. Can someone tell
me what the correct syntax should be in this case?

Thanks in advance. Please cc me; I'm not on the list.

                                            Sincerely, Faheem Mitha.

************************************************************************
#include "lattice.hh"
#include "mh.hh"
#define MATHLIB_STANDALONE
#include <Rmath.h>

/* random sample from proposal distribution q-gibbs */
double q_gibbs_rand(Cube& cube, double& beta,
vector<double>::size_type&
t)
{
  vector<double> ctr = ((0.5)*(cube.down + cube.up)).coord;
  vector<double>::size_type hsize = ctr.size();
  double a, u, y;

  if(t == 0)
    a = beta*ctr[t+1];
  else if(t == hsize - 1)
    a = beta*ctr[t-1];
  else
    a = beta*(ctr[t-1] + ctr[t+1]);

  GetRNGstate();
  u = unif_rand();
  PutRNGstate();

  y = (1/a)*log( ( exp(a) - exp(-a) )*u + exp(-a) );
  return y;
}
****************************************************************************

g++-3.0 -c -Wall -pedantic -Werror -g  rand.cc -lRmath
In file included from rand.cc:4:
/usr/include/Rmath.h:520: declaration of `double log1p(double)' throws
   different exceptions
/usr/include/bits/mathcalls.h:125: than previous declaration `double
   log1p(double) throw ()'
rand.cc: In function `double q_gibbs_rand(Cube&, double&,
size_t&)':
rand.cc:23: `GetRNGstate' undeclared (first use this function)
rand.cc:23: (Each undeclared identifier is reported only once for each
function
   it appears in.)
rand.cc:25: `PutRNGstate' undeclared (first use this function)
make: *** [rand.o] Error 1

-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at
stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._

"Faheem" == Faheem Mitha <faheem at email.unc.edu> writes:
  Faheem> I am trying to use R's Math library as standalone, as
documented in
  Faheem> /src/nmath/standalone. I am using C++ in Debian testing, and the
  Faheem> versions are as follows:
  Faheem> 
  Faheem> ii  g++-3.0        3.0.3-1        The GNU C++ compiler.
  Faheem> ii  r-mathlib      1.4.0-1        `GNU S'  - Standalone R math
library
[...]
  Faheem> I get two different errors.
  Faheem> 
  Faheem> The first one I am completely mystified by.

It is just a warning which goes away if you omit -pedantic. It seems to
collide with the header from glibc:

edd at homebud:/tmp> grep log1p /usr/include/Rmath.h
/usr/include/bits/mathcalls.h
/usr/include/Rmath.h:#define log1p              Rf_log1p
/usr/include/Rmath.h:double  log1p(double); /* = log(1+x) {care for small x} */
/usr/include/bits/mathcalls.h:__MATHCALL (log1p,, (_Mdouble_ __x));

  Faheem> The second one clearly has to do with the usage of GetRNGstate()
and
  Faheem> PutRNGstate() which are correct for use as entry points into the R
  Faheem> executable but no longer correct for standalone usage. 

They do not exist in Rmath.h -- meaning that you cannot call them.

  Faheem> The README in the standalone directory said something about using
  Faheem> the uniform random number generator, but gave no examples so I
  Faheem> didn't understand. Can someone tell me what the correct syntax
  Faheem> should be in this case?

Not sure I understand your question.  The set_seed() function allows to fix a
seed so that that subsequent calls will always get the same sequence of
random deviates.  So a simple example which works is

edd at homebud:/tmp> cat fm.cc
#include <iostream>
#include <cstdlib>
#include <cmath>

#define MATHLIB_STANDALONE 1
#include <Rmath.h>

int main(void)
{
  for (int i=0; i<3; i++) {
    set_seed(123,456);          // no claims that these are good seeds
    double u = unif_rand();
    cout << "Got " << u << endl;
  }
}
edd at homebud:/tmp> g++ -Wall -o fm fm.cc -lRmath
edd at homebud:/tmp> ./fm
Got 0.384571
Got 0.384571
Got 0.384571

Hope this helps,  Dirk

-- 
Good judgment comes from experience; experience comes from bad judgment. 
							    -- F. Brooks
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at
stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._