#247: mln_min() of a floating type is not the effective minimum of the range value -----------------------------------------------------+---------------------- Reporter: David Coeurjolly and Julien Marquegnies | Owner: Olena Team Type: defect | Status: new Priority: critical | Milestone: Component: Milena | Version: 2.0 Keywords: | -----------------------------------------------------+---------------------- According to David Coeurjolly {{{ int main() { mln::accu::stat::max<double> mm; mm.init(); mm.take(-0.02); mm.take(-0.92); mm.take(-0.03); std::cout<<" max= "<<mm.to_result()<<std::endl; } }}} In many situations (w/o init, w/o 'to_result() method in the ostream), I always get a wrong value (2.22-308, which seems to be min_val for positive double). Julien Marquegnies confirmed that mln_min(double) or mln_min(float) returns DBL_MIN instead of -DBL_MAX. According to the meaning of mln_min/max which are supposed to give the lower and upper bounds of the range values, returning -DBL_MAX '''should''' fix the issue. -- Ticket URL: <https://trac.lrde.org/olena/ticket/247> Olena <http://olena.lrde.epita.fr> Olena, a software platform dedicated to image processing.