# A triangular representation for the divisor summatory function, $D(x)$

Let $d(n)$ represent the divisor function as

$d(n)=\displaystyle\sum\limits_{k|n}1$

and the divisor summatory function as

$D(x)=\displaystyle\sum\limits_{n \leq x}d(n)$

I found the following triangular representation for the values of $D(n)$

$$\begin{array}{ccccccccc} D(1)=&&&&&&&&& 1 &&&&&&&&&&=1\\ &\\ D(2)=&&&&&&&& 2 &+& 1 &&&&&&&&&=3\\ &\\ D(3)=&&&&&&& 3 &+& 1 &+& 1 &&&&&&&&=5\\ &\\ D(4)=&&&&&& 4 &+& 2 &+& 1 &+& 1 &&&&&&&=8\\ &\\ D(5)=&&&&&5 &+& 2 &+& 1 &+& 1 &+& 1&&&&&&=10\\ &\\ D(6)=&&&&6 &+& 3 &+& 2 &+& 1 &+& 1 &+& 1&&&&&=14\\ &\\ D(7)=&&&7 &+& 3 &+& 2 &+& 1 &+& 1 &+& 1&+& 1 &&&&=16\\ &\\ D(8)=&&8 &+& 4 &+& 2 &+& 2 &+& 1 &+& 1&+& 1&+&1&&&=20\\ &\\ \end{array}$$

The values on the right are the sum of all elements in a row.

EDIT 1:

The above picture is the result of the following observation:

Let $v_{m}(n)$ be the greatest power of $m$ that divides $n$ with $m,n \in \mathbb{N}$ , so we get that

$D(n)=\displaystyle\sum\limits_{m=2}^{\infty}v_{m}(p^{n}), p \in \mathbb{P}$ where $p$ is a fixed prime number.

I didn't try to prove this. I don't know how to do it, but hopefully some one will have some idea on how to prove or disprove this conjecture.

I'd like to know if this is a known fact. I don't have a proof but I've tested lots of values and woks all the time.

Thanks.

-
Am I missing an obvious patter in your column on the right? –  BBischof Sep 27 '10 at 15:22
@BBischof, the values on the right are $D(n)$, e.g. $D(1)=1$, $D(2)=3$, $D(3)=5$, $D(4)=8$, etc. –  Neves Sep 27 '10 at 15:26
I've swapped n and x in the divisor summatory function. –  anon Sep 27 '10 at 15:32
The values in the triangle are just the quotient of dividing the row number by the column number, so it's no surprise that this gives the sum of divisors. –  anon Sep 27 '10 at 15:36
When voting down, please, explain why. –  Neves Sep 27 '10 at 16:23

Yes, this is true. Write $D(x) = \sum_{n \le x} d(n) = \sum_{n \le x} \sum_{d | n} 1 = \sum_{d \le x} \lfloor \frac{x}{d} \rfloor$; this is equivalent to the pattern you observe. The last step is exchanging the order of summation together with the observation that the number of times a number $d$ appears in the double sum is the number of numbers less than or equal to $x$ it divides.
thats true, I already knew that way of calculating $D(x)$, but as I came to this observation from another path I never looked at the triangle as a result of $\sum_{d \leq x}\lfloor\frac{x}{d}\rfloor$, but thats true, the values in the triangle are "just the quotient of dividing the row number by the column number". –  Neves Sep 27 '10 at 17:10
$$D(n)=\sum_{k=1}^{\lfloor\sqrt{n}\rfloor}\left(2 \cdot \left\lfloor \frac{n-k^2}{k}\right\rfloor+1\right)$$