Bill Allombert on Mon, 30 Sep 2002 15:42:46 +0200


[Date Prev] [Date Next] [Thread Prev] [Thread Next] [Date Index] [Thread Index]

Re: apparent sigma() problem


On Sun, Sep 29, 2002 at 11:52:55PM -0400, Igor Schein wrote:
> Hi,
> 
> ? sigma(2,-1)
> 6
> 
> doesn't make sense.  There's nothing in the manual that prevents 2nd
> argument from being negative, so I'd expect the answer to be 3/2.
Both you and the manual are correct, only the code is at fault:

? sigma(2,-1)
%1 = 6 /*wrong*/
? sigma(2,-2)
%2 = 5/4 /*correct*/

Just an oversight, -1 was special cased, but we forgot to switch
the signe so the correct formula
sigma(n,-k)=sigma(n,k)*n^(-k)
became the wrong one
sigma(n,-1)=sigma(n,k)*n
sigh...

Thanks for spotting it!
Cheers,
Bill