Boost logo

Ublas :

Subject: [ublas] error in reducing types with vector_scalar_binary_traits
From: Michael Norel (minorlogic_at_[hidden])
Date: 2009-08-14 04:51:26


Hi All!

I use "1_35_0" boost version and try to use ublas with "fadbad" library.

"http://www.fadbad.com/fadbad.html"
Automatic differentiation using templates and operator overloading in C++

I have problm with "reducing" operators from "fadbad::F<double>". The ublas assume that result of reduce is "double" type, but correct is "fadbad::F<double>".

simple pseudocode:
//=============================
#include "externalLibs/fadbad/fadiff.h"
void test()
{
   typedef fadbad::F<double> fadDouble;
   boost::numeric::ublas::vector<fadDouble> vvv(777);
   boost::numeric::ublas::inner_prod(vvv, vvv);
}
//=============================

Please help me, how can i say to ublas reduce fadbad::F<double> to fadbad::F<double> ? This will help me a lot.

Thanks!