I am wondering, why is it so hard to take the derivative or integral of a factorial function defined as $f(x) = x!$.
I know that approximations do exist, and it wouldn't be so hard to subject them to derivation or integration. I doubt that the reason is that factorials are defined only for whole numbers…
Is my thought right? What other things make this kind of a derivation seemingly hard (or impossible) to get ?
One thing I did is like this:
$$ (n+1)! – n! = n\times n!$$
This is a commonly known thing.
Now, derivatives are defined as (I am very much of a noob in calculus, even if I know a little bit of it):
$$\lim_{h\to 0}\dfrac{f(x+h) – f(x)}{h}$$
Taking $f(x) = x!$, when we try to take the derivative:
$$\lim_{h \to 0} \dfrac{(x+h)! – x!}h$$
In this case, simple substitution won't work, and since I don't know many other rules well, maybe I won't be able to work it out well.
But if I try to use Lagrange's mean value theorem to see if there is a place where the $f(x)$ here has a minima or a maxima (I only have a small knowledge about this) here, I'd have $f'(c) = \dfrac{a! – b!}{a-b}$ ($a$ and $b$ are distinct) and this can't be zero anywhere, since factorials are defined only for positive integers and $a = b$ means that the $f'(c)$ we get enters indeterminate forms.
Anything more that I can go through to understand why $\dfrac d{dx}{x!}$ is hard to evaluate ?
Best Answer
That's exactly the reason. Calculus tools are made to handle functions defined on (most of) the real number line. That's their raison d'ĂȘtre. They can't really handle a function that is defined only on the integers, without either a massive reworking of calculus itself, or redefining the function to have a value everywhere.