Some define the Dirac Delta Function as:

$$\int_{-\infty}^{\infty}\delta(x)f(x)\ dx=f(0)$$

For every continuous function $f$. In some books, I've noticed a different definition of $\delta(x)$ as an operation that satisfies the following two conditions:

$$\int_{-\infty}^{\infty}\delta(x)\ dx=1\quad\text{and}\quad\forall x\neq0:\delta(x)=0$$

Are the two definitions the same?


First of all, $\delta $ is not a function ! There are no function s.t. $$\int_{\mathbb R}\delta (x)=1\quad \text{and}\quad \forall x\neq 0, \delta (x)=0.$$

The second definition is rather not correct strictly speaking, but should be understood as $$\int_{\mathbb R}\delta (x)\,\mathrm d x=1\quad \text{and}\quad \int_A \delta (x)\,\mathrm d x=\begin{cases}1&0\in A\\ 0&\text{otherwise}\end{cases}.$$

So, no they are not the same. But they are indeed equivalents.

  • $\begingroup$ I have managed to show that the first definition implies the two conditions that the second definition requires. Unfortunately I couldn't do the opposite, could you please explain the second direction? $\endgroup$
    – Amit Zach
    Apr 2 '20 at 13:42
  • $\begingroup$ You can see $$\int_A\delta (x)\,\mathrm d x:=\int_{\mathbb R}\delta (x)\boldsymbol 1_A(x)\,\mathrm d x.$$ Then use the standard approximation theorem to prove that $$\int_{\mathbb R}f(x)\delta (x)dx=f(0).$$ Notice that the writing $\delta (x)dx$ is a bit confusing, and should rather be written either $\mathrm d \delta (x)$ or even better $\delta (\mathrm d x).$. And thus $(\mathbb R,\mathcal M,\delta )$ can be seen as a measure space. And all theorem you know as DCT, MCT, Fatou , approximations by simple function... works. $\endgroup$
    – Surb
    Apr 2 '20 at 13:55

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.