For given positive integers $n,k$ prove that there always exists some $x$ for which $2^n \mid \dfrac{x(x+1)}{2}-k.$

My work:
$\dfrac{x(x+1)}{2}$ is the sum of all positive integers upto $x$.
Now, if we can show,
$\dfrac{x(x+1)}{2}\equiv 0,1,\ldots,2^n-1 \mod2^n$ then we are done.
Now, two cases arise,
(i) $x=2j \implies j(2j+1)\equiv 0,1,\ldots,2^n-1$
(ii) $x+1=2j \implies j(2j-1)\equiv 0,1,\ldots,2^n-1$
Now, I have reached a deadlock. Please help.


$$P_{n,k}(x) \equiv 2^n | \frac{x^2 + x}2 - k$$

By mechanical inspection we can see that the $x$ which satisfy $P_{n,k}$ this are all of the form $x_z = z\cdot 2^{n+1} + B$. So it suggests that we should consider this equation this equation in modulo $2^{n+1}$.

Consider possible solutions of the form $x = B \pmod {2^{n+1}}$:

$$2^n = \frac {B^2 + B} 2 - k \pmod {2^{n + 1}}$$

$$B^2 + B - 2k = 0 \pmod {2^{n+1}}$$

So we only have to show that this polynomial always has a solution in $B$. Induct on increasing $n$.

The base case for $n = 1$, $B^2 + B - 2k = 0 \pmod {4}$ finds solutions by a checking the 4 possible values of $k$.

The inductive case follows from Hensel's Lemma.

More detailed information is available from this XKCD thread.


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.