0
$\begingroup$

I'm curious about the following diophantine equation from my invention: I don't know if this is in the literature, I wrote it using creativity in an attempt to write a variant of the equation in Fermat's Last Theorem.

Problem. We consider for positive integers $x,y,z\geq 1$ and being pairwise coprime, $(x,y)=(x,z)=(y,z)=1$, and for a fixed integer $n\geq 3$ the diophantine equation that involves partial sums of exponentials $$\sum_{k=0}^n\frac{x^k}{k!}+\sum_{k=0}^n\frac{y^k}{k!}=\sum_{k=0}^n\frac{z^k}{k!}.\tag{1}$$ The problem asks if it is possible to determine if $(1)$ for our given integer $n>2$ has solutions.

Question. I don't know if previous problem is in the literature. Is it possible to determine for a given integer $n\geq 3$ if the diophantine equation (the problem) have solutions $(x,y,z)$ in coprime positive integers $x,y,z$? Many thanks.

If this problem is in the literature please refer it in your comments of answer.

Example (Case $n=2$). One can to get easily solutions for the equation with $n=2$ that isn't considered in our Problem. For example $(x,y,z)=(17,5,18)$.

Remark. I know that the polynomial $P(x)=1+x+\frac{x^2}{2!}+\ldots+\frac{x^n}{n!}$ is in the literature and enjoys of certain properties.

$\endgroup$
2
  • 2
    $\begingroup$ For $n=4$, you have the solution $(x,y,z) = (32, 51, 53)$. $\endgroup$ Sep 13, 2021 at 10:13
  • $\begingroup$ Many thanks, this is excellent @Carl-FredrikNybergBrodda , I hope that if it is feasible you or other professor can to add a discussion, this is the purpose of the question, about what work can be done for the Question. $\endgroup$
    – user142929
    Sep 13, 2021 at 15:13

0

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.