Take the 2-minute tour ×
Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

I read on Wikipedia that growth rate of a function can sometimes be greater than exponential. Can you give me some examples of such functions (preferably continuous ones)?

Obviously $x^x$ grows faster than normal exponential, and $x^{x^x}$ even more so - does this concept have a name, and can an arbitrary/infinite amount of such "exponentiality" be expressed with a mathematical expression?

Any other interesting functions to be aware of?

share|improve this question
1  
For the name of the concept: look for "powertower" in mathworld or wikipedia, and for attempts to the extension to continuous "heights" (the number of x in your question as the variable parameter) look for the term "tetration" in wikipedia (note, that there is not yet an accepted extension for general powertowers to interpolated "heights") –  Gottfried Helms Jan 25 '12 at 20:40

4 Answers 4

up vote 3 down vote accepted

The Gamma function defined by $\Gamma(x) = \int_0^\infty e^{-t}t^{x-1}dt$ is a continuous function (and further, an analytic function) for which $\Gamma(n) = (n-1)!$ for all $n \in \mathbb{N}$. In particular, it grows faster than any exponent.

Asymptotically,

$\Gamma(z) \cong \sqrt{z} \cdot (\frac{z}{e})^z $

share|improve this answer

You can compose the exponential function any number of times with itself, say $f_0(x)=x$ and $f_{n+1}(x)=\exp(f_n(x))$ for all $n\in\mathbf N$, to get ever faster growing functions. Every $f_n$ is an analytic function, so everywhere (in $\mathbf C$) indefinitely differentiable. For increasing arguments $x\in\mathbf R$, these functions still grow much slower than even $2\uparrow\uparrow m$ does as a function of $m\in\mathbf N$, because the latter composes $x\mapsto 2^x$ un unbounded number of times (namely $m$ times) with itself (and then applies to $1$), whereas each $f_n$ only has a fixed number $n$ compositions of $\exp$. Lacking a continuous version of function composition (composing a function $x$ times with itself), I'm not sure one can match the growth of $2\uparrow\uparrow m$ with an analytic function.

share|improve this answer
2  
$@$Marc: There is a standard result in complex analysis (in the neighborhood of Weierstrass Factorization and Mittag-Leffler) that given any closed discrete subset $\Omega \subset \mathbb{C}$ and any function $f: \Omega \rightarrow \mathbb{C}$, $f$ can be extended to an entire function. Taking $\Omega = \mathbb{N}$, we see that we can interpolate the up-arrow stuff by an analytic function. –  Pete L. Clark Jan 25 '12 at 19:41

There is a standardized system of writing very large numbers, to be found here. The example from the page looks like this:

$$ 2 \uparrow\uparrow\uparrow 4 = \begin{matrix} \underbrace{2_{}^{2^{{}^{.\,^{.\,^{.\,^2}}}}}}\\ \qquad\quad\ \ \ 65,536\mbox{ copies of }2 \end{matrix} \approx (10\uparrow)^{65,531}(6.0 \times 10^{19,728}) \approx (10\uparrow)^{65,533} 4.3 , $$ where $(10\uparrow)^n$ denotes a functional power of the function $f(n) = 10n$.

The $\uparrow$ is Knuth's up-arrow notation. With it $x^x$ translates to $x\uparrow\uparrow2$ and $x^{x^x}$ to $x\uparrow\uparrow3$.

share|improve this answer

Aside from the fact that that it seems like you're trolling because I already provided examples of continuous functions that grow faster, you're also wrong to imply descrete operators can't produce continous functions. x+x^2+x^3... may require transcendental operations to be described in closed form as a descrete summation, but it can still be applied to any infinitesimally small number to produce a continuous curve. In fact, every taylor series is a descrete summation of polynomials, yet, they all produce continuous curves that approximate continuous functions. Also you can apply a single summation operation to each value as I mentioned before. You could use the sum of n=1 of k/n^2 to n=infinity, where k has a domain of all real numbers, so you could replace k one by one with all possible valus and get a continuous curve from that summation operator. Such an operation is commonly used as and reffered to by anyone who has taken first semester calculus as an "integral."

share|improve this answer

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.