Hi, hmm that’s a bug. There has been a change in Gecode 6 for FloatVal (if used as coefficients in linear constraints which expr creates, they are not allowed to extend from negative to positive and include zero) but not for FloatVar. I’ll have a look sometime this week and get back to you.
Cheers Christian -- Christian Schulte, https://chschulte.github.io/ Professor of Computer Science, KTH, [email protected]<mailto:[email protected]> Expert Researcher, RISE SICS, [email protected]<mailto:[email protected]> From: [email protected] <[email protected]> On Behalf Of Kurt Van Den Branden Sent: 24 April 2018 17:03 To: [email protected] Subject: [gecode-users] gecode 6 exception hi, here is some simple example code: FloatVar x (*this, -100, 100); FloatVar t1 = expr (*this, abs (x * 100)); // OK FloatVar t2 = expr (*this, abs (x) * 100); // NOK variable t1 doesn't cause a problem when creating variable t2, an exception is thrown (OutOfLimits("Float::linear")) the same code compiled with gecode 4.3.2 doesn't cause problems. am I doing something wrong? Kind Regards, Kurt Van Den Branden | Agfa Graphics
_______________________________________________ Gecode users mailing list [email protected] https://www.gecode.org/mailman/listinfo/gecode-users
