Silly question about sampling a function (distribution vs integral representation)

69 Views Asked by At

I have a stupid question... if i want to sample a continuous function i can use the Dirac delta distribution

$$f(x_0) = \int_{-\infty}^{+\infty} \delta(x - x_0) f(x)dx,$$

which indeed involves the distribution theory, however i've been thinking that the same thing (with some assumption) could be written as:

$$ f(x_0) = \frac{1}{2 \pi j} \oint_{\gamma} \frac{f(z)}{z - x_0} dz$$

where $\gamma$ is a closed curve that contains $x_0$. Is there some relevant relention/difference between the two representations? Could the Cauchy integral formula be seen as a linear functional (like the one usually studied in functional analysis)?