bug#29387: Guile-2.2.2 complex numbers edge case

bug#29387: Guile-2.2.2 complex numbers edge case

[Jan Wedekind]

Hi,
I think I encountered a bug in the numerical stack.
i times i should equal -1:

     (equal? -1 (* 0+i 0+i))
     ; #f

i times i plus one is zero (which is correct):
     (zero? (+ (* 0+i 0+i) 1))
     ; #t

Regards
Jan

bug#29387: Guile-2.2.2 complex numbers edge case

[Daniel Llorens]

[-- Attachment #1: Type: text/plain, Size: 960 bytes --]


> From: Jan Wedekind <jan@wedesoft.de>
> Subject: bug#29387: Guile-2.2.2 complex numbers edge case
> Date: 21 Nov 2017 23:09:57 GMT+1
> To: 29387@debbugs.gnu.org
> Reply-To: Jan Wedekind <jan@wedesoft.de>
> 
> 
> Hi,
> I think I encountered a bug in the numerical stack.
> i times i should equal -1:
> 
>    (equal? -1 (* 0+i 0+i))
>    ; #f
> 
> i times i plus one is zero (which is correct):
>    (zero? (+ (* 0+i 0+i) 1))
>    ; #t
> 
> Regards
> Jan

In Guile 0+i and (* 0+i 0+i) are inexact numbers, but -1 is exact. That's why equal? fails.

You can check that either of

(equal? -1.+0. (* 0+i 0+i))

or

 (= -1 (* 0+i 0+i))

return true, as expected.

I don't know if the fact that Guile doesn't have exact complex numbers could be considered a bug. AFAIR none of the standards require them to be implemented. On the other hand I just checked and both Racket and Chez seem to have them, so...

regards

	Daniel


[-- Attachment #2: Type: text/html, Size: 2800 bytes --]

bug#29387: Guile-2.2.2 complex numbers edge case

[Jan Wedekind]

[-- Attachment #1: Type: text/plain, Size: 1299 bytes --]

Thanks Daniel,
I see. So I need to use "=" for numerical comparisons. "equal?" is only true if the types are the same, too:

    (equal? 1.0 1)
    ; #f
    (= 1.0 1)
    ; #t

Sorry for filing a bug report. Feel free to close the bug.

Regards
Jan

On November 22, 2017 3:53:00 PM GMT, Daniel Llorens <daniel.llorens@bluewin.ch> wrote:
>
>> From: Jan Wedekind <jan@wedesoft.de>
>> Subject: bug#29387: Guile-2.2.2 complex numbers edge case
>> Date: 21 Nov 2017 23:09:57 GMT+1
>> To: 29387@debbugs.gnu.org
>> Reply-To: Jan Wedekind <jan@wedesoft.de>
>> 
>> 
>> Hi,
>> I think I encountered a bug in the numerical stack.
>> i times i should equal -1:
>> 
>>    (equal? -1 (* 0+i 0+i))
>>    ; #f
>> 
>> i times i plus one is zero (which is correct):
>>    (zero? (+ (* 0+i 0+i) 1))
>>    ; #t
>> 
>> Regards
>> Jan
>
>In Guile 0+i and (* 0+i 0+i) are inexact numbers, but -1 is exact.
>That's why equal? fails.
>
>You can check that either of
>
>(equal? -1.+0. (* 0+i 0+i))
>
>or
>
> (= -1 (* 0+i 0+i))
>
>return true, as expected.
>
>I don't know if the fact that Guile doesn't have exact complex numbers
>could be considered a bug. AFAIR none of the standards require them to
>be implemented. On the other hand I just checked and both Racket and
>Chez seem to have them, so...
>
>regards
>
>	Daniel

[-- Attachment #2: Type: text/html, Size: 3454 bytes --]

bug#29387: Guile-2.2.2 complex numbers edge case

[Daniel Llorens]

On 22 Nov 2017, at 19:32, Jan Wedekind <jan@wedesoft.de> wrote:

> Thanks Daniel,
> I see. So I need to use "=" for numerical comparisons. "equal?" is only true if the types are the same, too:
> 
> (equal? 1.0 1)
> ; #f
> (= 1.0 1)
> ; #t
> 
> Sorry for filing a bug report. Feel free to close the bug.

no problem! closing the bug (right this time)