I've looked on Math.SE and Wikipedia so far and found nothing. I thought I found what I was looking for, but it turned out to be Euler's Totient function.

Taking the canonical representation of an integer $n$:

$$n=p_0^{a_0}\cdot p_1^{a_1}\cdots p_r^{a_r}$$

Is there a standard notation for $\sum_{i=0}^r a_r$? Are there any known theorems dealing directly with this quantity aside from those dealing with the Totient function? Thank you!

  • 1
    $\begingroup$ It's $\Omega(n)$, the number of all prime factors of $n$, in contrast to $\omega(n)$, which is the number of distinct prime divisors of $n$. $\endgroup$ – Daniel Fischer Sep 25 '13 at 17:51
  • $\begingroup$ @DanielFischer: perfect, thank you! $\endgroup$ – abiessu Sep 25 '13 at 17:53
  • $\begingroup$ @DanielFischer: wait a moment though, how does this usage relate to the one shown here? $\endgroup$ – abiessu Sep 25 '13 at 17:56
  • 1
    $\begingroup$ As for a theorem dealing with it, the Riemann hypothesis is equivalent to the statement $\sum_{k \le n}{(-1)^{\Omega(k)}} = O(n^{\frac{1}{2}+\epsilon})$ for all $\epsilon \gt 0$. $\endgroup$ – Dan Brumleve Sep 25 '13 at 17:57
  • $\begingroup$ @abiessu Not related. The same symbol used for different things. $\endgroup$ – Daniel Fischer Sep 25 '13 at 17:58

The comments provided the requested notation accurately, namely, that if we have

$$n=p_0^{a_0}\cdot p_1^{a_1}\cdots p_r^{a_r}\tag 1$$

then we use $\Omega(n)=\sum_{i=0}^ra_i$ to denote the count of prime factors of $n$ including multiplicity. I believe I have even found a reasonable arithmetic function composition that can be used to "calculate" $\Omega(n)$ for any given $n$, although I introduce a "new" function (which can probably best be considered an "anti-derivative" of an existing function) in order to arrive at this information.

To begin, note that the Mangoldt function is defined as

$$\Lambda(n)=\left\{\begin{array}{ll}\log p & \text{if $n=p^k$ for prime $p$ and integer $k$}\\0 & \text{otherwise}\end{array}\right.$$

Now branch our "new" function from here as the arithmetic "anti-derivative" of the Mangoldt function, which is to say that the Mangoldt function will be the arithmetic derivative of our function:

$$\eth(n)=\left\{\begin{array}{ll}\log_np=\frac 1k & \text{if $n=p^k$ for prime $p$ and integer $k$}\\0 & \text{otherwise}\end{array}\right.$$

Now the sum over the distinct prime divisors of $n$ works like this:

$$\omega(n)=\sum_{d\mid n}\mu(d)\eth(d)$$

Getting $\Omega(n)$ is a little trickier, since we have $\eth(p^k)=\frac 1k$ for primes $p$ when we actually want $1$ for each of these values. But this can be solved with multiplication by another sum:

$$\Omega(n)=\sum_{d\mid n}\left(\mu(d)\eth(d)\sum_{d^k\mid n}1\right)$$

Using our definition, we can ask some questions about our function; for instance:

$$\sum_{d\mid n}\eth(d)=\sum_{i=0}^rH_{a_i}$$

where $H_n$ is the $n$th Harmonic Number, being the sum $\sum_{i=1}^n\frac 1i$. This is slightly counter-intuitive when we consider that $\log n=\sum_{d\mid n}\Lambda(d)$ and that $\eth(n)={\Lambda(n) \over \log n}$.

| cite | improve this answer | |

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.