Re: Sum of factors


"Christian Bau" <christian.bau@xxxxxxxxxxxxxxxxxxxx> escribió en el mensaje
news:christian.bau-A0311F.00430721042005@xxxxxxxxxxxxxxxxxxxxxxxxxxx
>
> You can probably find the factors if M and N are very close together.
> Assume that c = a +/- 1, solve for a, b, c and d, check if the results
> are integers. Same for c = a +/- 2, c = a +/- 3 and so on. If M and N
> are very close then there is probably a low limit how far a and c can be
> apart.


Mmmm interesting... Take a look at this:

a*b=M
c*d=N

a+b=c+d

Then...

|a-c| divides (M-N)

(yet unproved. But seems to be true empiricaly. Somebody could prove this?)


So if it´s true it would be enough to factor M-N and try...

c = a +/- t ; for all t divisor of M-N

(I´m spanish, so please forgive my English)







.

Quantcast