Yes, I believe it would.
Whether that's desirable is a different question.
Henry Rich
On 11/30/2014 12:48 AM, robert therriault wrote:
What I don't quite understand at this point is why we can't set the comparison
to intolerant using ~:!.0
Wouldn't that cause both definitions to return the same result for a=. 1j1e_45
I will spend a bit more time with Raul's answer, as i think that the answer may
be lurking in there.
Cheers, bob (too)
On Nov 29, 2014, at 9:32 PM, Henry Rich <[email protected]> wrote:
Right. (Bob too)
I think tolerant equality to 0 is the same as intolerant.
Henry Rich
On 11/30/2014 12:19 AM, Raul Miller wrote:
#I.0=(iscomplex=isnotreal)1 j. -:^:a:1
1030
So, for example:
(iscomplex=isnotreal)1j1e_99
0
Basically, the epsilon used for tolerant equality is much coarser than
the ratio of real to imaginary which is possible in J's complex
representation.
Thanks,
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
