Zsbán Ambrus wrote: > > On Wed, Dec 8, 2010 at 9:02 AM, bill lam <[email protected]> wrote: >> Is the following expected behavior? >> 256 256#: 255 >> 0 255 >> 256 256#: (0.2-0.2)+255 >> 0 255 >> 256 256#: (0j1-0j1)+255 >> 1 _1 > > On Wed, Dec 8, 2010 at 9:53 AM, Roger Hui <[email protected]> wrote: >> Looks like a bug. > > It's not a bug. Tolerant <. behaves inconsistently on numbers > represented as real versus complex floats. In addition, the > dictionary never really defines how tolerant <. works (on either reals > or complexes). The | and #: dyads are just defined using <. and > inherit the inconsistencies from it, but are themselves implemented > correctly as far as I see. > > Ambrus > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > > All these arguments notwithstanding, in this case the relevant part is that (0j1 - 0j1) should give zero, and therefore (0j1-0j1)+255 should give 255. So I'd definitely say it's a bug. -- View this message in context: http://old.nabble.com/-%3A-on-integer-of-internal-complex-type-tp30403330s24193p30406759.html Sent from the J Programming mailing list archive at Nabble.com.
---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
