7
$\begingroup$

I asked myself which primes $p$ can be written as $p=x^3+y^3+z^3-(x^2z+y^2x+z^2y)$ with $x,y,z \in \mathbb{N}$. But for $p \neq 2$ we have the solution $x=y=\frac{p-1}{2}$ and $z=\frac{p+1}{2}$. So the only prime for which I can not find a solution is $p=2$. But I can not prove that there is not a solution. Any ideas?

$\endgroup$
  • $\begingroup$ Just out of curiousity: What is the matrix $M$? $\endgroup$ – Jose Arnaldo Bebita-Dris Jul 29 '17 at 15:12
  • 1
    $\begingroup$ @JoseArnaldoBebitaDris: I edited the question. $\endgroup$ – orgesleka Jul 29 '17 at 15:15
  • $\begingroup$ Look at the equation modulo 4. $\endgroup$ – user940 Jul 29 '17 at 16:57
3
$\begingroup$

EDIT: You're right stackExchangeUser; my proof doesn't work. With a similar tack, we can still salvage this: \begin{align*} &x^3 + y^3 + z^3 - (x^2 z + y^2 x + z^2 y) \\ = ~ &(x + y + z)^3 - 4(x^2 z + y^2 x + z^2 y) - 3(x^2y + y^2 z + z^2 x) - 6xyz \end{align*} So, we are solving, $$(x + y + z)^3 = 2 + 4(x^2 z + y^2 x + z^2 y) + 3(x^2y + y^2 z + z^2 x) + 6xyz$$ Suppose first the left side is divisible by $2$. Again, at least one of $x, y, z$ must be even. If all of them are even, we see that the left side is $0$ mod $8$, but the right side is $2$ mod $8$. Thus, exactly one must be even. But then, the $3(x^2y + y^2 z + z^2 x)$ term is odd, which makes the right hand side odd, and we get a contradiction again.

Thus, the left side is odd. Similarly, either all of $x, y, z$ are odd, or exactly one is. If exactly one of them is odd, then the right hand side is even, hence all $x, y, z$ are odd.

Finally, considering the original formulation, and the fact that $x^2 \equiv 1$ mod $8$ for all odd $x$, we get,

$$x^3 + y^3 + z^3 - (x^2 z + y^2 x + z^2 y) \equiv x + y + z - (z + x + y) \equiv 0$$

mod $8$, which cannot equal $2$.

$\endgroup$
  • 1
    $\begingroup$ sage: expand($(x+y+z)^3-4*(x^2*z+y^2*x+z^2*y)-6*x*y*z$) $x^3 + 3*x^2*y - x*y^2 + y^3 - x^2*z + 3*y^2*z + 3*x*z^2 - y*z^2 + z^3$ . I do not understand your first equality. SAGE tells me something different. $\endgroup$ – orgesleka Jul 29 '17 at 15:21

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.