0

here is my attempt.is it correct? $$I = \int e^{-x^2}dx$$ let $u = e^{-x^2}$ and $v=x$ then : $$I = xe^{-x^2} - \int -2x^2e^{-x^2} dx ; J = \int -x^2e^{-x^2}dx$$ here is where I'm in doubt am I alowed to do this? let $x^2 = i\theta \Rightarrow x = \sqrt{i\theta}$ $$dx = \frac{\sqrt{i}}{2\sqrt{\theta}}$$ $$J = -\int \frac{\sqrt{i}}{2\sqrt{\theta}} i\theta e^{-i\theta}d\theta$$ $$ = $$ $$- \frac{i\sqrt{i}}{2}\int\sqrt{\theta}e^{-i\theta} d\theta = - \frac{i\sqrt{i}}{2} K$$ $$K = \int\sqrt{\theta}e^{-i\theta} d\theta = \int\sqrt{\theta}(\cos\theta - i\sin\theta) d\theta$$ with $$\Lambda (\theta)= \int\sqrt{\theta}\cos\theta d\theta = \frac{\sqrt{-i\theta} \Gamma(3/2, -i \theta) + \sqrt{i \theta} \Gamma(3/2, i \theta)}{2 \sqrt{\theta}} + C$$ And : $$\int\sqrt{\theta} (- i\sin\theta) d\theta = \Lambda (-\theta)$$ we have: $$K = \int\sqrt{\theta}(\cos\theta - i\sin\theta) d\theta = \Lambda (\theta) + i\Lambda (-\theta)$$ then: $$J = - \frac{i\sqrt{i}}{2} K = - \frac{i\sqrt{i}}{2}\Lambda (\theta) + \frac{\sqrt{i}}{2}\Lambda (-\theta)$$ we finally have : $$I = \int e^{-x^2}dx = xe^{-x^2} - 2J = xe^{-x^2} +i\sqrt{i}\Lambda (\theta) - \sqrt{i}\Lambda (-\theta) + \kappa$$

for $x^2 = i\theta$

by the way I am in high school so please refrain from technical talk!

  • What does $\int$ mean for you? Indefinite integral or $\int_{-\infty}^\infty$? –  May 26 '22 at 17:09
  • indefinite integral. perhaps it was not wise of me to use the term gaussian integral! – Ilia Varnaseri May 26 '22 at 17:11
  • There is no elementary indefinite integral for $e^{-x^2}$. –  May 26 '22 at 17:12
  • well i was trying to find one! is it correct? or is it proven that there is no elementary integral ? – Ilia Varnaseri May 26 '22 at 17:13
  • See https://math.stackexchange.com/questions/265780/how-to-determine-with-certainty-that-a-function-has-no-elementary-antiderivative –  May 26 '22 at 17:15
  • And this: https://www.quora.com/Why-is-there-no-anti-derivative-for-e-x-2-in-terms-of-elementary-functions –  May 26 '22 at 17:16
  • thanks for your information. then where did i go wrong? – Ilia Varnaseri May 26 '22 at 17:18
  • @user1046533 but then again gamma function is not an elementary function. is it? – Ilia Varnaseri May 26 '22 at 17:30
  • When you set $x^2=i\theta$ your $\theta$ can't be real. What is then all that integration w.r.t. $\theta$ ? We cannot just manipulate symbols. – Kurt G. May 26 '22 at 18:37
  • @KurtG. but when i evaluate it from minus inf to plus inf (its simplified form) i get $\sqrt{\pi} \approx 1.77$ here is the link : https://www.wolframalpha.com/input?i2d=true&i=Limit%5B-Divide%5B%5C%2840%29Sqrt%5BPi%5D+Erfc%5C%2891%29x%5C%2893%29%5C%2841%29%2C2%5D+-+I+Gamma%5C%2891%29Divide%5B3%2C2%5D%5C%2844%29+Power%5Bx%2C2%5D%5C%2893%29%2Cx-%3E%E2%88%9E%5D+-+Limit%5B-Divide%5B%5C%2840%29Sqrt%5BPi%5D+Erfc%5C%2891%29x%5C%2893%29%5C%2841%29%2C2%5D+-+I+Gamma%5C%2891%29Divide%5B3%2C2%5D%5C%2844%29+Power%5Bx%2C2%5D%5C%2893%29%2Cx-%3E-%E2%88%9E%5D – Ilia Varnaseri May 26 '22 at 19:13
  • $\theta=-ix^2$ is purely imaginary. What is that "it" that you evaluate from $-\infty$ to $+\infty$ ? – Kurt G. May 26 '22 at 19:15
  • well when i gave the result that i found in the detail of the question to the wolfram alpha machine ; the computer simplified the expression and gave me some other expression(it is visible in the link) then i gave it to the machine to evaluate the expression using limits (like for ordinary integrals) and i got the result. as you can see in the link – Ilia Varnaseri May 26 '22 at 19:19
  • It looks like what you rediscovered is a fomula akin to $\gamma(\frac{1}{2},x)=\sqrt{\pi}\operatorname{erf}(\sqrt{x})$ where $\gamma(s,x)$ is the lower incomplete Gamma function. – Kurt G. May 26 '22 at 19:24
  • well is this method that was rediscovered useful in any way? – Ilia Varnaseri May 26 '22 at 19:29
  • Certainly useful for both of us. For the rest of the world I cannot tell. – Kurt G. May 26 '22 at 19:30
  • well i sincerely am thankful for the time you spend on this question! – Ilia Varnaseri May 26 '22 at 19:35
  • Never loose your enthusiasm for mathematics ! – Kurt G. May 26 '22 at 19:40
  • Also posted to MO, https://mathoverflow.net/questions/423372/gaussian-indeterminate-integral-without-the-error-function – Gerry Myerson May 27 '22 at 01:32

0 Answers0