Crazy Gaussian Integral

Calculus Level 5

If the following integral is satisfied for some relationship between the constants a , b a , b and c c -

e ( a x 2 + 2 b x y + c y 2 ) d x d y = 1 \displaystyle\int_{-∞}^{∞}\displaystyle\int_{-∞}^{∞} e^{-(ax^{2} + 2bxy + cy^{2})}dxdy = 1

Find the value of b 2 a c b^{2} -ac


The answer is -9.869.

This section requires Javascript.
You are seeing this because something didn't load right. We suggest you, (a) try refreshing the page, (b) enabling javascript if it is disabled on your browser and, finally, (c) loading the non-javascript version of this page . We're sorry about the hassle.

2 solutions

Ronak Agarwal
Nov 14, 2014

Change to polar co-ordiantes :

x = r c o s ( θ ) x=rcos(\theta)

y = r s i n ( θ ) y=rsin(\theta)

And our area element d x d y dxdy becomes r d θ d r rd\theta dr

Also we are integrating it for the whole space :

Integral becomes :

0 2 π 0 r e r 2 ( a c o s 2 ( θ ) + 2 b c o s ( θ ) s i n ( θ ) + c s i n 2 ( θ ) ) d r d θ \large \int _{ 0 }^{ 2\pi }{ \int _{ 0 }^{ \infty }{ r{ e }^{ -{ r }^{ 2 }(a{ cos }^{ 2 }(\theta )+2bcos(\theta )sin(\theta )+c{ sin }^{ 2 }(\theta )) }drd\theta } }

With respect to dr this can be easily solved by substituting r 2 = t {r}^{2}=t

Integrating first integral we get :

0 2 π 1 2 ( a c o s 2 ( θ ) + 2 b c o s ( θ ) s i n ( θ ) + c s i n 2 ( θ ) ) d θ \int _{ 0 }^{ 2\pi }{ \frac { 1 }{ 2(a{ cos }^{ 2 }(\theta )+2bcos(\theta )sin(\theta )+c{ sin }^{ 2 }(\theta )) } d\theta }

Which can be written as :

I = 0 π 1 ( a c o s 2 ( θ ) + 2 b c o s ( θ ) s i n ( θ ) + c s i n 2 ( θ ) ) d θ I=\int _{ 0 }^{ \pi }{ \frac { 1 }{ (a{ cos }^{ 2 }(\theta )+2bcos(\theta )sin(\theta )+c{ sin }^{ 2 }(\theta )) } d\theta }

Breaking it in parts we get :

I = 0 π / 2 1 ( a c o s 2 ( θ ) + 2 b c o s ( θ ) s i n ( θ ) + c s i n 2 ( θ ) ) d θ + π / 2 π 1 ( a c o s 2 ( θ ) + 2 b c o s ( θ ) s i n ( θ ) + c s i n 2 ( θ ) ) d θ I=\int _{ 0 }^{ \pi /2 }{ \frac { 1 }{ (a{ cos }^{ 2 }(\theta )+2bcos(\theta )sin(\theta )+c{ sin }^{ 2 }(\theta )) } d\theta } +\int _{ \pi /2 }^{ \pi }{ \frac { 1 }{ (a{ cos }^{ 2 }(\theta )+2bcos(\theta )sin(\theta )+c{ sin }^{ 2 }(\theta )) } d\theta }

Solving both integrals (as they are pretty standard and can be solved by first dividing numerator and denominator by c o s 2 ( θ ) {cos}^{2}(\theta) and substituting t a n ( θ ) = t tan(\theta)=t ) we get :

π a c b 2 = I = 1 \frac { \pi }{ \sqrt { ac-{ b }^{ 2 } } } =I=1

a c b 2 = π 2 \Rightarrow ac-{b}^{2}={\pi}^{2}

Humberto Bento
Nov 9, 2014

Just use gaussian distribution: 1 2 π σ 2 e ( x μ ) 2 2 σ 2 d x = 1 \int\limits_{-\infty }^{\infty }{\frac{1}{\sqrt{2\pi {{\sigma }^{2}}}}{{e}^{-\frac{{{(x-\mu )}^{2}}}{2{{\sigma }^{2}}}}}dx}=1

Knowing that: e α ( x μ ) 2 d x = π α \int\limits_{-\infty }^{\infty }{{{e}^{-\alpha {{(x-\mu )}^{2}}}}dx}=\sqrt{\frac{\pi }{\alpha }} The Integral may be rewriteen as: e ( a x 2 + 2 b y x + c y 2 ) d x d y = e a ( x + b a y ) 2 e ( c b 2 a ) y 2 d x d y = \int\limits_{-\infty }^{\infty }{\int\limits_{-\infty }^{\infty }{{{e}^{-(a{{x}^{2}}+2byx+c{{y}^{2}})}}dxdy}}=\int\limits_{-\infty }^{\infty }{\int\limits_{-\infty }^{\infty }{{{e}^{-a{{(x+\frac{b}{a}y)}^{2}}}}{{e}^{-(c-\frac{{{b}^{2}}}{a}){{y}^{2}}}}dxdy}}= π a e ( c b 2 a ) y 2 d y = π a π c b 2 a = π 2 a c b 2 = 1 \sqrt{\frac{\pi }{a}}\int\limits_{-\infty }^{\infty }{{{e}^{-(c-\frac{{{b}^{2}}}{a}){{y}^{2}}}}dy}=\sqrt{\frac{\pi }{a}}\sqrt{\frac{\pi }{c-\frac{{{b}^{2}}}{a}}}=\sqrt{\frac{{{\pi }^{2}}}{ac-{{b}^{2}}}}=1

Therefore: a c b 2 = π 2 ac-{{b}^{2}}={{\pi }^{2}}

Please correct me if I'm wrong, I think we should make restriction here, i.e. < y < x < -\infty<y<x<\infty , so that this process can be justified e a ( x + b a y ) 2 d x = π a \int_{-\infty}^{\infty} e^{-a\left(x+\frac{b}{a}y\right)^2}\,dx=\sqrt{\frac{\pi}{a}} Both variables are not separable here.

Anastasiya Romanova - 6 years, 7 months ago

Log in to reply

Right. But we are integrating over x, and y acts as a constant relative to x. That't the reason why the integral may be calculated!

Humberto Bento - 6 years, 7 months ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...