Doubt about the classical way to solve the gaussian integral

58 Views Asked by At

Probably the most known way to solve $$I=\int_{-\infty}^{\infty} e^{-x^2} dx$$ Is considering $$I^2=\int_{\mathbb{R}^2}e^{-x^2-y^2}dxdy$$ My doubt is really basic and somehow I understand why it makes no sense, but I would like to know more expert opinions. The doubt is why we use another variable, the variable $y$, and not another variable $x$ so we could get the (wrong) integrand $e^{-x^2-x^2}=e^{-2x^2}$.

My thought is that in general in double integrals the variables are referred to two different sets $A$ and $B$ and so, even if $A=B$, we can't sum them because they are distinct objects of two sets; but I know too that integration variable are dummy variables and it is the same to consider $\int_a^b f(x)dx$ or $\int_a^b f(t)dt$ and sum them, even if in that case they are the same number.

Why is this wrong? Thanks.

1

There are 1 best solutions below

1
On BEST ANSWER

Perhaps you would easily understand this if you compare with sums. Note that $ \sum\limits_{k=1}^{\infty} \frac 1 {2^{k}}=1$. Hence $( \sum\limits_{k=1}^{\infty} \frac 1 {2^{k}})^{2}=1$. But if you write $( \sum\limits_{k=1}^{\infty} \frac 1 {2^{k}})^{2}$ as $ \sum\limits_{k=1}^{\infty} \frac 1 {4^{k}}$ you would end up with $\frac {1/4} {1-1/4}=\frac 1 3$. So $ \sum\limits_{k=1}^{\infty} a_k^{2}=\sum_k a_k \sum_j a_j =\sum_{j,k} a_ja_k$ and not $\sum_k a_k a_k$.