I know that $\delta(g(x))=\sum_i\dfrac{\delta(x-x_i)}{g'(x)}$, but I want it's proof, am not getting how to proof this.
-
As the roots of $x^2-x-6$ are $-2$ and $3$, isn't the integral equal to : $(-2)^2+1 + 3^2+1$? – Wyllich Sep 14 '17 at 09:33
-
No the answer is not what you have written, see the detail section. – Vivekanand Mohapatra Sep 14 '17 at 11:00
-
https://math.stackexchange.com/questions/2389100/composition-of-the-derivative-of-dirac-delta-with-a-function/2389360#2389360 – md2perpe Sep 14 '17 at 16:16
2 Answers
First note that,
$$\delta_a(x) = 1 \qquad \text{if $x= a$ and $\delta_a(x) = 0$ if $x\neq a$}$$ and for any measurable function $f$
$$\int_{\mathbb R}f(x)\delta_a(dx) = f(a).$$
$$\delta_0\equiv \delta.$$ But $$ x^2 -x-6 = 0 \Longleftrightarrow x=3 ~~or ~~x=-2$$
Hence
$$\delta(x^2 -x-6) = 1 \qquad \text{if $ x=3 $ or $x=-2$ and $\delta(x^2 -x-6) = 0 $ if else}$$
Therefore $$\delta(x^2 -x-6) = \delta_3+\delta_{-2}$$
Hence for any measurable function $f$ we have
$$\int_{\mathbb R}f(x)\delta(x^2 -x-6)dx = \int_{\mathbb R}f(x) (\delta_3+\delta_{-2})dx= \int_{\mathbb R}f(x) \delta_3(dx)+\int_{\mathbb R}f(x)\delta_{-2}(dx)= f(3)+f(-2)$$
Thus
$$\int_{\mathbb R}(x^2+1)\delta(x^2 -x-6)dx = 3^2+1+2^2+1 = 15$$

- 23,903
-
$\delta(0)$ is not equal to $1$, but rather some kind of infinity. Moreover wolfram|alpha disagrees with the result. However I don't remember anything about dirac distribution-related integrals, so I can't say which is right. Edit: maybe change variables to get $\delta(u)$ in the integral? – Mariuslp Sep 14 '17 at 10:36
-
You need to divide through by $|d/dx (x^2 - x - 6) |$ evaluated at each of the roots, i.e. $5$, which will then agree with wolfram – Nadiels Sep 14 '17 at 10:50
-
Bro the answer is wrong, see the detail section and plz give a proof – Vivekanand Mohapatra Sep 14 '17 at 11:01
I know the answer, the integration will give
$\int_{-\infty}^{+\infty}(x^2+1)\delta(x^2-x-6)\,dx$
$\implies\displaystyle\delta(x^2-x-6)=\delta\left[(x+2)(x-3)\right]=\frac{\delta(x+2)}{-5}+\frac{\delta(x-3)}{5}$
So, the integration becomes
$\int_{-\infty}^{+\infty}(x^2+1)\left[\frac{\delta(x+2)}{-5}+\frac{\delta(x-3)}{5}\right]\,dx$
$\implies\dfrac{(x^2+1)}{-5}|_{x=-2}+\dfrac{(x^2+1)}{5}|_{x=3}=1$
But what’s my doubt is , How to proof this?
$\implies\displaystyle\delta(x^2-x-6)=\delta\left[(x+2)(x-3)\right]=\frac{\delta(x+2)}{-5}+\frac{\delta(x-3)}{5}$
Help me Guys
EDIT
This is how I tried the proof
$\displaystyle\int_{-\infty}^{+\infty}f(x)\delta[(x-a)(x-b)]\,dx$
But we know that $\displaystyle\boxed{\int_{-\infty}^{+\infty}\delta[(x-a)(x-b)]\,d[(x-a)(x-b)]=1}$
Right?
$\implies\int_{-\infty}^{+\infty}(2x-(a+b))\delta[(x-a)(x-b)]\,dx=1$
So,
$\displaystyle\int_{-\infty}^{+\infty}f(x)\delta[(x-a)(x-b)]\,\frac{d[(x-a)(x-b)]}{2x-(a+b)}$
$\implies\lim_{\epsilon\to 0}\left(\int_{0}^{a+\epsilon}+\int_{a+\epsilon}^{b}\right)f(x)\delta[(x-a)(x-b)]\,\frac{d[(x-a)(x-b)]}{2x-(a+b)}$
$\implies\left(\dfrac{f(a)}{2a-(a+b)}+\dfrac{f(b)}{2b-(a+b)}\right)$
Hence, $\displaystyle\boxed{\delta[g(x)]=\sum_i\frac{\delta(x-x_i)}{g'(x)}}\tag*{}$
REGARDS!
VM

- 109
- 2