On Sat, Aug 20, 2016 at 3:47 PM, leif wrote:
> Erik Bray wrote:
>> On Fri, Aug 19, 2016 at 11:00 AM, Erik Bray wrote:
>>> On Wed, Aug 10, 2016 at 1:38 PM, Erik Bray wrote:
Hi all,
Sorry if this has been discussed ad-infinitum before--I looked around
a bit but didn't find a de
Erik Bray wrote:
> On Fri, Aug 19, 2016 at 11:00 AM, Erik Bray wrote:
>> On Wed, Aug 10, 2016 at 1:38 PM, Erik Bray wrote:
>>> Hi all,
>>>
>>> Sorry if this has been discussed ad-infinitum before--I looked around
>>> a bit but didn't find a definitive answer.
>>>
>>> I have one (well at least one
On Fri, Aug 19, 2016 at 11:00 AM, Erik Bray wrote:
> On Wed, Aug 10, 2016 at 1:38 PM, Erik Bray wrote:
>> Hi all,
>>
>> Sorry if this has been discussed ad-infinitum before--I looked around
>> a bit but didn't find a definitive answer.
>>
>> I have one (well at least one) test that's failing on C
On Wed, Aug 10, 2016 at 1:38 PM, Erik Bray wrote:
> Hi all,
>
> Sorry if this has been discussed ad-infinitum before--I looked around
> a bit but didn't find a definitive answer.
>
> I have one (well at least one) test that's failing on Cygwin due to
> tiny difference in the last digit of the resu
Jeroen Demeyer wrote:
> On 2016-08-12 11:24, leif wrote:
>> Yeah, the doctest framework should automatically base the tolerance on
>> such comparisons... ;-) :P
>
> I don't know how serious you use, but you have to keep in mind that the
> doctest framework works with strings. The main problem is t
On 2016-08-12 11:24, leif wrote:
Yeah, the doctest framework should automatically base the tolerance on
such comparisons... ;-) :P
I don't know how serious you use, but you have to keep in mind that the
doctest framework works with strings. The main problem is that you would
need to guess the
Erik Bray wrote:
> On Fri, Aug 12, 2016 at 9:42 AM, Jeroen Demeyer
> wrote:
>> On 2016-08-11 12:24, Erik Bray wrote:
>>>
>>> In this case the difference is just one ulp.
>>
>>
>> To be more precise: the difference is actually about 0.59 ulp on Cygwin:
>>
>> sage: R = RealField(256)
>> sage: log3_
Erik Bray wrote:
> On Thu, Aug 11, 2016 at 2:17 PM, Dima Pasechnik wrote:
>>> FWIW, on Cygwin we used to use the Cephes library for a while. (No idea
>>> whether it now replaced libm or parts of it got merged into libm
>>> upstream, or none of that.)
>>>
>>> But the result you get presumably also
On Thu, Aug 11, 2016 at 2:17 PM, Dima Pasechnik wrote:
>> FWIW, on Cygwin we used to use the Cephes library for a while. (No idea
>> whether it now replaced libm or parts of it got merged into libm
>> upstream, or none of that.)
>>
>> But the result you get presumably also depends on whether libm
On Thursday, August 11, 2016 at 12:15:32 PM UTC+1, leif wrote:
>
> Erik Bray wrote:
> > On Wed, Aug 10, 2016 at 2:45 PM, Thierry Dumont
> > > wrote:
> >> Le 10/08/2016 à 13:38, Erik Bray a écrit :
> >>> Hi all,
> >>>
> >>> Sorry if this has been discussed ad-infinitum before--I looked aroun
Erik Bray wrote:
> On Wed, Aug 10, 2016 at 2:45 PM, Thierry Dumont
> wrote:
>> Le 10/08/2016 à 13:38, Erik Bray a écrit :
>>> Hi all,
>>>
>>> Sorry if this has been discussed ad-infinitum before--I looked around
>>> a bit but didn't find a definitive answer.
>>>
>>> I have one (well at least one)
Jeroen Demeyer wrote:
> If the error is 1 ulp or more for a basic function (like log) on a
> reasonable non-pathological input (like 3.0), I would consider that
> an upstream bug.
Quite a few implementations only provide accuracies of 3-4 ulp for speed
reasons (it may make it possible to use doub
Erik Bray wrote:
> Hi all,
>
> Sorry if this has been discussed ad-infinitum before--I looked around
> a bit but didn't find a definitive answer.
>
> I have one (well at least one) test that's failing on Cygwin due to
> tiny difference in the last digit of the result of log(3).
>
> This leads to
Hey Erik,
It seems like what you want is # rel tol XYZ or # abs tol XYZ. For an
example, see rings/real_double.pyx.
Best,
Travis
On Wednesday, August 10, 2016 at 6:38:59 AM UTC-5, Erik Bray wrote:
>
> Hi all,
>
> Sorry if this has been discussed ad-infinitum before--I looked around
> a bit
14 matches
Mail list logo
