The Gamma function satisfies the relation $z\Gamma(z)=\Gamma(z+1)$, whence $|\Gamma(z+1)|>|\Gamma(z)|$ whenever $|z|>1$ (and $z$ is not a non-positive integer). This naturally leads us to the idea that for a fixed $y>1$ the function $|\Gamma(x+iy)|$ of the variable $x$ could be increasing. Is this true? I would much appreciate a reference or a proof. The area $\{x+iy: \ x>1, \ y<1\}$ is of interest as well.

  • $\begingroup$ What is $G$? Is it supposed to be $\Gamma$? $\endgroup$ – Alex Ortiz Oct 29 '18 at 18:34
  • $\begingroup$ Yes. Edited. Thanks. $\endgroup$ – Durac Oct 29 '18 at 18:36
  • $\begingroup$ $$|\Gamma(x+iy)| \sim \sqrt{2 \pi} \, |y|^{x-1/2} \exp{(-\frac{\pi}{2}|y|)} , \quad y \to \infty$$ can be found in Gradshteyn and Ryzhik, eq. 8.328, and a reference given therein. Eq. 8.326.2 might be of interest to you as well. $\endgroup$ – skbmoore Oct 29 '18 at 18:43
  • $\begingroup$ @skbmoore: IMO your hint does not apply here, because $y$ should be fix. $\endgroup$ – gammatester Oct 29 '18 at 18:48
  • 2
    $\begingroup$ Real and imaginary part of Gamma function. $\endgroup$ – AccidentalFourierTransform Oct 29 '18 at 18:57

This is wrong. Here a plot for $|\Gamma(x+1.01 \cdot i)|$

enter image description here

$$|\Gamma(0.2+1.01\cdot i)|= 0.516403048428670750862809525289$$ $$|\Gamma(0.6+1.01\cdot i)|= 0.510624247985962860913065928364$$

  • $\begingroup$ Thanks, you are right. But in fact I am interested in this property for sufficiently large $y$. What should I do then? Accept this answer and open a new question? $\endgroup$ – Durac Oct 29 '18 at 19:08
  • $\begingroup$ It is your decision. (The critical $|y|$ seems to be very close to $1$). $\endgroup$ – gammatester Oct 29 '18 at 19:14
  • $\begingroup$ math.stackexchange.com/questions/2976553/… $\endgroup$ – Durac Oct 29 '18 at 19:24
  • $\begingroup$ I'd say there is a convex curve $x+i\, f(x)$ with $f(-\infty) = +\infty, f(+\infty) = 0$ such that for every $z, \Im(z) \ge f(\Re(z))$ then $|\Gamma(z+x)|$ is strictly increasing in $x > 0$. Thus a question would be the asymptotic of the optimal $f$. @Durac $\endgroup$ – reuns Oct 29 '18 at 22:07

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.