Anyhow, there is a bug in
min_wt_vec_gap()
in the non-GUAVA part the vector v might never get initialised. The if-block
if dist and dist < dist_min:
dist_min = dist
ans = list(v)
might never gets executed, e.g. in the case
codes.HammingCode(2, GF(2)).minimum_distance()
it does not, as dist never drops below dist_min, and ans is left empty
Start with
dist_min = n+1
and the bug goes away.
I'll open a ticket with a fix.
Dima
On Saturday, 23 May 2015 10:03:19 UTC+1, Dima Pasechnik wrote:
>
> GAP in Sage was upgraded in the meantime.
>
> On Friday, 22 May 2015 13:06:28 UTC+1, vdelecroix wrote:
>>
>>
>>
>> On 22/05/15 13:03, David Joyner wrote:
>> > On Fri, May 22, 2015 at 3:46 AM, Vincent Delecroix
>> > <20100.d...@gmail.com> wrote:
>> >> Hello,
>> >>
>> >>
>> >>
>> >> On 22/05/15 04:33, Rob Beezer wrote:
>> >>>
>> >>> The code below fails on 6.7 with a RuntimeError.
>> >>>
>> >>> r = 2
>> >>> C = codes.HammingCode(r, GF(2))
>> >>> C.minimum_distance()
>> >>
>> >>
>> >> This is a bug in the algorithm since it does not handle the case where
>> all
>> >> your vectors have non zero entries.
>> >
>> > This statement should be interpreted to mean that all the entries of
>> > the generator matrix of C are non-zero.
>>
>> Nope. It will fail if and only if your linear code is 1 dimensional and
>> the generating vector has no zero entry. This works
>>
>> sage: V = LinearCode(matrix(GF(5), 2, 3, [[1,2,3],[1,1,1]]))
>> sage: V.minimum_distance()
>> 2
>>
>> >> It might be because of me in #17452
>> >>
>> >>> For r > 5 it seems to hang interminably, while back in
>> November/December
>> >>> I could do r = 7 in a few seconds, and do r = 10 without waiting
>> too
>> >>> long.
>> >>
>> >
>> >>From your traceback, it seems that there is a problem with calling
>> > some C code in GAP's kernel?
>>
>> No. It is just an explicit RuntimeError raised by the current code. The
>> name of the Python function in which this error is raised is
>> min_wt_vec_gap. Which does not imply that it is a problem with GAP.
>>
>> Vincent
>>
>
--
You received this message because you are subscribed to the Google Groups
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at http://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.
