0

Definitions

  • $\gamma$ is the classical Euler-Mascheroni constant.
  • $\zeta(s)$ is the Riemann zeta function.
  • $\Gamma(s)$ is the Gamma function.

Question

From the Claim Results below for $n=2$, I was surprised to find $\gamma_2 = 2\gamma$.

Thus, the Euler-Mascheroni constant $\large \gamma=\frac{\gamma_2}{2}$ can be expressed as

$ \large \bbox[10px,border:2px solid #0000ff]{\gamma =\frac{\pi^{\frac{1}{2}}}{2 \, \Gamma\left(\frac{3}{2}\right)} \lim\limits_{s\to 2}\left(\zeta(s-1)-\dfrac{1}{s-2}\right)} $

Is there a way to numerically verify that expression correctly approximates the classical Euler-Mascheroni constant $\gamma \approx 0.5772156649...$?


Claim

The $n$-dimensional Euler-Mascheroni constant, $\gamma_n$, is given by:

$$ \gamma_n = \begin{cases} \displaystyle \gamma, & \quad n = 1 \\[12pt] \displaystyle \frac{\pi^{\frac{n - 1}{2}}}{\Gamma\left( \dfrac{n + 1}{2} \right)} \lim\limits_{s \to n} \left( \zeta(s - 1) - \frac{1}{s - 2} \right), & \quad n \ge 2 \end{cases} $$

Proof

Case: $n \ge 3$

The volume of an $n$-dimensional hypersphere is

$$ V_n(R) = \frac{\pi^{n/2} R^n}{\Gamma\left(1 + \frac{n}{2}\right)} $$

For $n-1$ dimensions with radius $R=\frac{1}{k}$ is

$$ V_{n-1}(\frac{1}{k}) = \frac{\pi^{\frac{n-1}{2}}}{k^{n-1}\Gamma\left( \frac{n+1}{2} \right)} $$

Total volume of the Stack of Hyperspherical Slices $V_s$

$$ V_s = \sum_{k=1}^\infty V_{n-1}\left( \dfrac{1}{k} \right) = \frac{\pi^{\frac{n - 1}{2}}}{\Gamma\left( \dfrac{n + 1}{2} \right)} \sum_{k=1}^\infty \dfrac{1}{k^{n - 1}} = \frac{\pi^{\frac{n - 1}{2}} \, \zeta(n - 1)}{\Gamma\left( \dfrac{n + 1}{2} \right)} $$

Total volume of the $n$-dimensional Gabriel's Horn $V_h$

$$ V_h = \int_1^\infty V_{n-1}\left( \dfrac{1}{x} \right) \, dx = \frac{\pi^{\frac{n - 1}{2}}}{\Gamma\left( \dfrac{n + 1}{2} \right)} \int_1^\infty \dfrac{1}{x^{n - 1}} \, dx = \frac{\pi^{\frac{n - 1}{2}}}{(n - 2) \, \Gamma\left( \dfrac{n + 1}{2} \right)} $$

Define $\gamma_n$ as

$$ \gamma_n = V_s - V_h = \frac{\pi^{\frac{n - 1}{2}}}{\Gamma\left( \dfrac{n + 1}{2} \right)} \left( \zeta(n - 1) - \dfrac{1}{n - 2} \right) $$

Case: $n = 2$

To avoid the problematic terms $\zeta(1)$ and $\frac{1}{0}$, reformulate $\gamma_n$ as

$$ \gamma_n=\frac{\pi^{\frac{n-1}{2}}}{\Gamma\left(\dfrac{n+1}{2}\right)} \lim\limits_{s\to n}\left(\zeta(s-1)-\dfrac{1}{s-2}\right) $$

Case: $n = 1$

Define the $1$-dimensional Euler-Mascheroni constant as $\gamma_1 = \gamma$.


Claim Results

\begin{array}{|c|c|c|} \hline \textbf{n-dimension} & \mathbf{\gamma_n} & \mathbf{\approx} \\ \hline 1 & \hspace{5pt} \gamma \hspace{5pt} & 0.5772156649... \\ \hline 2 & \hspace{5pt} 2\gamma \hspace{5pt} & 1.1544313298... \\ \hline 3 & \hspace{5pt} \pi \left( \zeta(2) - 1 \right) \hspace{5pt} & 2.0261201264...\\ \hline 4 & \hspace{5pt} \frac{2}{3} \pi \left( 2 \, \zeta(3) - 1 \right) \hspace{5pt} & 2.9407690791... \\ \hline 5 & \hspace{5pt} \frac{1}{2} \pi^2 \left( \zeta(4) - \frac{1}{3} \right) \hspace{5pt} & 3.6961170085... \\ \hline 6 & \hspace{5pt} \frac{8}{15} \pi^2 \left( \zeta(5) - \frac{1}{4} \right) \hspace{5pt} & 4.1422216722... \\ \hline 7 & \hspace{5pt} \frac{1}{6} \pi^3 \left( \zeta(6) - \frac{1}{5} \right) \hspace{5pt} & 4.2237941871... \\ \hline 8 & \hspace{5pt} \frac{16}{105} \pi^3 \left( \zeta(7) - \frac{1}{6} \right) \hspace{5pt} & 3.9767533569... \\ \hline \end{array}

Interestingly, $0$-dimensional $\gamma_0 = \frac{5}{12\pi}$


References

vengy
  • 2,421
  • Your definition of $\gamma_2$ makes no sense because it involves $\zeta(1)$ and $\frac{1}{0}$. – jjagmath Oct 27 '24 at 02:08
  • 1
    @jjagmath It seems to follow from $\underset{n\to 2}{\text{lim}}\left(\zeta (n-1)-\frac{1}{n-2}\right)=\gamma$ and $\frac{\pi ^{\frac{n-1}{2}}}{\Gamma \left(\frac{n+1}{2}\right)}=2$ when $n=2$. – Steven Clark Oct 27 '24 at 02:31
  • 1
    This follows from the identity $\underset{s\to 1}{\text{lim}}\left(\zeta (s)-\frac{1}{s-1}\right)=\gamma$ (see formula 49 at Wolfram MathWorld: Euler-MascheroniConstant for example). – Steven Clark Oct 27 '24 at 02:51
  • @StevenClark Definitions don't follow from something, they are given, not deduced from other facts. – jjagmath Oct 27 '24 at 02:55
  • @jjagmath I think I see your point, the formula should have been written as $$\gamma_n=\lim\limits_{s\to n}\left(\frac{\pi^{\frac{s-1}{2}}}{\Gamma\left(\dfrac{s+1}{2}\right)} \left(\zeta(s-1)-\dfrac{1}{s-2}\right)\right)+\delta_{n,1} \left(\gamma-\dfrac{1}{2}\right).$$ – Steven Clark Oct 27 '24 at 04:05
  • @StevenClark ... or define the case for $n=2$ separately. – jjagmath Oct 27 '24 at 04:10
  • Or perhaps $$\gamma_n=\frac{\pi^{\frac{n-1}{2}}}{\Gamma\left(\dfrac{n+1}{2}\right)} \lim\limits_{s\to n}\left(\zeta(s-1)-\dfrac{1}{s-2}\right)+\delta_{n,1} \left(\gamma-\dfrac{1}{2}\right)$$ would be sufficient, but it bothers me that the $\delta_{n,1} \left(\gamma-\dfrac{1}{2}\right)$ terms seems rather artificial. Assuming your formula is correct, I'm wondering if there's another representation that would seem less contrived. – Steven Clark Oct 27 '24 at 15:58
  • I've removed the delta term and just defined the $1$-dimensional Euler-Mascheroni constant as $\gamma_1 = \gamma$. It seems less artificial now. Thanks for the tip. – vengy Oct 27 '24 at 20:38
  • 1
    It seems to me your dimensions are wrong, i.e. the 1-dimensional case result should be zero and the 2-dimensional case result should be $\gamma$. – Steven Clark Oct 29 '24 at 02:41
  • One dimension is a line, two dimensions is a plane, and the $\gamma$ constant is derived from integration in the x-y plane which you even refer to as 2D in your related question. – Steven Clark Oct 29 '24 at 02:56
  • I think the reason your formula is giving you $2 \gamma$ instead of $\gamma$ for the case $n=2$ is that your formula assumes a symmetric structure about the real axis, and the contribution of the lower-half plane is the same as the contribution of the upper-half plane. – Steven Clark Oct 29 '24 at 05:15

0 Answers0