[ADMB Users] ADMB with Mavericks

[dave fournier]

On 13-11-14 04:31 PM, Leo Polansky wrote:

my version of the code looks like

void function_minimizer::trust_region_update(int nvar,int _crit,
   independent_variables& x,const dvector& _g,const double& _f)
{
   double & f= (double&)_f;
   dvector & g= (dvector&)_g;
   int unvar=1;   // unvar defined here
   fmm fmc(nvar);
   if (random_effects_flag)
   {
     initial_params::set_active_only_random_effects();
     //cout << nvar << endl;
     unvar=initial_params::nvarcalc();                 //unvar used here
     initial_params::restore_start_phase();

So I have no idea why it thinks that unvar is undefined here.

What does your version of the code look like?

> Dear ADMB Developers,
> Thank you for making this software; I'm looking forward to using it to 
> move beyond shoehorning my problems into things solvable in R. I 
> attempted to install admb on Mac with Mavericks (OS X 10.9) following 
> the instructions from Johnoel and Steve on earlier posting on this 
> thread, i.e.
>
> $ svn checkout http://www.admb-project.org/svn/trunk admb-trunk
> $ cd admb-trunk
> $ make
>
> However,  besides the occasional warning, this did not work. In the 
> end the make command produced
>
> *df1b2-separable/f1b2trst.cpp:31:5: **error: **use of undeclared 
> identifier 'unvar'; did you mean 'nvar'?*
>
> unvar=initial_params::nvarcalc(); // get the number of active
>
> *  ^~~~~*
>
>   nvar
>
> *df1b2-separable/f1b2trst.cpp:21:50: note: *'nvar' declared here
>
> void function_minimizer::trust_region_update(int nvar,int _crit,
>
> *                                              ^*
>
> 1 error generated.
>
> make[1]: *** 
> [../build/objects/dist/optlp-df1b2-separable-f1b2trst.obj] Error 1
>
> make: *** [c++-all] Error 2
>
>
> It appears I didn't get as far as Alejandro who seemed to get it to 
> the point of being able to have an 'admb' command available. I'm not 
> sure how to proceed, what to try, etc. Trying to download admb 
> directly from the ADMB webpage and following the unix installation 
> instructions did not work. Also, should I be nervous about Dave's 
> earlier comment that an unsafe version was installed in some earlier 
> cases? Any suggestions would be appreciated. In case it's helpful here 
> is the C compiler version info.
>
>
> gcc -v
>
> Configured with: 
> --prefix=/Applications/Xcode.app/Contents/Developer/usr 
> --with-gxx-include-dir=/usr/include/c++/4.2.1
>
> Apple LLVM version 5.0 (clang-500.2.76) (based on LLVM 3.3svn)
>
> Target: x86_64-apple-darwin13.0.0
>
> Thread model: posix
>
>
> g++ -v
>
> Configured with: 
> --prefix=/Applications/Xcode.app/Contents/Developer/usr 
> --with-gxx-include-dir=/usr/include/c++/4.2.1
>
> Apple LLVM version 5.0 (clang-500.2.76) (based on LLVM 3.3svn)
>
> Target: x86_64-apple-darwin13.0.0
>
> Thread model: posix
>
>
>
>
>
> On Mon, Nov 11, 2013 at 9:03 AM, dave fournier <davef at otter-rsch.com 
> <mailto:davef at otter-rsch.com>> wrote:
>
>     Your first version appears to make the optimized version of the
>     program which is intended to run fast and does not do bounds checks.
>
>
>     *** tpl2cpp mixsan
>
>     *** adcomp mixsan
>     g++ -c -O3 -Wno-deprecated -D__GNUDOS__ -Dlinux -DOPT_LIB
>     -DUSE_LAPLACE -fpermissive -I. -I"/Applications/ADMB/ADMB-11
>     Terminal.app/Contents/admb-11"/include -
>
>     b
>
>
>     Your second version
>
>
>     *** tpl2cpp -bounds mixsan
>
>     *** adcomp -s mixsan.cpp
>     c++ -c -O3 -D__GNUDOS__ -Dlinux -DSAFE_ALL -DUSE_LAPLACE -I.
>     -I"/Users/alejandro/admb-trunk/build/dist"/include
>     -I"/Users/alejandro/admb-trunk/build/dist"/contrib/include mixsan.cpp
>
>
>     makes the slower "safe" version of the program which does bounds
>     checking.  You should always do this first
>     to make sure that a new program does not walk out of any arrays.
>
>     So it appears that your program never really worked, but the first
>     version did not report the problem.
>
>
>
>
>
>     _______________________________________________
>     Users mailing list
>     Users at admb-project.org <mailto:Users at admb-project.org>
>     http://lists.admb-project.org/mailman/listinfo/users
>
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.admb-project.org/pipermail/users/attachments/20131114/d385ea3e/attachment.html>