On 6/2/12 4:32 AM, Jaime Villate wrote:
> On 06/02/2012 05:28 AM, Raymond Toy wrote:
>> On 6/1/12 6:20 PM, Robert Dodier wrote:
>>> On 2012-06-01, ?????? ???????<bankov_dmitrii at mail.ru> wrote:
>>>
>>>> integrate(sin(a*x^3)/x, x, 0, +inf);
>>>> maxima result: 0;
>>>> correct result: %pi * abs(a) / 6 / a;
>>> Thanks for this info. I confirm that it happens w/ Maxima 5.27.0.
>> How did you get this? I tried it with the current git version and I get
>>
>> gamma(1/3)/6/a^(1/3)
>>
>> when a> 0.
>>
>> Looking at the code in defint.lisp, I see maxima has a special routine
>> (scaxn) to handle this type of integral. integrate(sin(a*x^n),x,0,inf)
>> -> gamma(1/n)/a^(1/n)/n*sin(%pi/2/n).
>>
>>
> ??
> I got the most recent git master snapshot yesterday and compiled it with
> SBCL, CCL, GCL, CLisp and ECL
> and I also get the wrong 0 result
>
> Maxima branch_5_27_base_65_g08552c3_dirty http://maxima.sourceforge.net
> using Lisp CLISP 2.49 (2010-07-07)
> Distributed under the GNU Public License. See the file COPYING.
> Dedicated to the memory of William Schelter.
> The function bug_report() provides bug reporting information.
> (%i1) integrate(sin(a*x^3)/x, x, 0, +inf);
> (%o1) 0
Well that's interesting. On my mac with sources from just now, cmucl,
ccl, and clisp don't give 0. Maxima asks for the sign of a (pos) and
returns gamma(1/3)/6/a^(1/3), which I think is right. Which differs
from what the original poster said should be the correct value.
Someone else is going to have to debug this.
Could you check to see if scaxn is called? It is on my systems. If it
isn't, then that would be the problem.
Ray