[seqfan] Re: A180248

[Klaus Brockhaus]

The original definition of A180248 can be transformed (also replacing 
Sindelar's N, P, S with n, d, r) to

"Odd composite squarefree numbers n such that r = 2*(d-2+n/d)/(d-1) is 
an integer for each prime divisor d of n".

Thereby the computation is speeded up considerably; the terms < 3800000 
can be obtained in less than a minute.

Perhaps with this definition someone can prove the conjecture

"n is a Carmichael number if and only if n is a term of A180248 and all 
r-values of n are even."

KB

Klaus Brockhaus:
> Hans Havermann:
>   
>> Am I reading the definition incorrectly? Why isn't 25 (p=5, s=4
>> included?
>>     
> W. F. Sindelar's definition is somewhat unclear to me; from the list of 
> terms given I concluded that they have to be squarefree. Perhaps this 
> follows from the definition, but I can't tell how. Rewording seems 
> appropriate.
>
>