X^x, does this function have a name ?

  • Context: High School 
  • Thread starter Thread starter uart
  • Start date Start date
  • Tags Tags
    Function
Click For Summary

Discussion Overview

The discussion revolves around the function x^x, exploring its properties, differentiation, and whether it has a commonly accepted name. Participants also touch on related concepts such as integrals and alternative representations of the function, as well as its relationship to other mathematical functions.

Discussion Character

  • Exploratory
  • Technical explanation
  • Conceptual clarification
  • Debate/contested
  • Mathematical reasoning

Main Points Raised

  • Some participants note that x^x does not have a widely recognized name and suggest that it can be represented as e^{x ln x} for differentiation purposes.
  • One participant mentions that there is no indefinite integral for x^x, while others confirm that it cannot be expressed in terms of common special functions.
  • A participant refers to x^x as the "power tower of order 2," but another argues that this term is improper due to the definition of power towers involving an infinite number of exponents.
  • Some participants discuss the differentiability and continuity of x^x on certain intervals, asserting that it is Riemann integrable on a subdomain of the real numbers.
  • There are mentions of the Lambert function in relation to the equation x^x = y, with one participant comparing x^x to factorial and Gamma functions in terms of growth.
  • Several participants share anecdotes about their experiences with mathematical software like Mathematica and Maple, discussing instances where these tools failed to compute certain integrals correctly.

Areas of Agreement / Disagreement

Participants generally agree that x^x does not have a commonly accepted name and that its indefinite integral cannot be expressed in terms of standard functions. However, there are competing views on the appropriateness of the term "power tower" and the relationship of x^x to other mathematical functions.

Contextual Notes

Limitations include the lack of consensus on terminology and the unresolved nature of certain mathematical properties, such as the existence of a closed-form antiderivative for x^x.

uart
Science Advisor
Messages
2,797
Reaction score
21
Hi, the other day I was explaining to somebody the difference between a power function like x^a and an exponential function like a^x (it was in relation to differentiation). So they asked me "what about x^x" and I was stumped to give it a name.

BTW, I was able to differentiate x^x ok, I got ( 1+ln(x) ) x^x which I'm pretty sure is correct, I just don't know if x^x has a commonly used name.
 
Physics news on Phys.org
Indeed your differentiation is correct. It doesn't have any sort of name that I know of (I've never seen it used for anything, either). A more useful representation is just

[tex]e^{x\ln x},[/tex]

of course.
 
It doesn't have a name.The exponential form saves you from learning yet another concept:logarthmic differentiation.

Daniel.
 
Data said:
A more useful representation is just [tex]e^{x\ln x},[/tex]

Doh of course. I don't know why I didn't think of that, particularly as I was explaining to this guy how to express a^x as [tex]e^{x\ln a}[/tex] (in order to differentiate it) at that very time. :blushing:
 
unfortunately there is no indefinite integral for x^x
 
Anectodically : the function [tex]x^x[/tex] is called "power tower of order 2"
 
Power tower is defined for an infinity of exponents...So it's a bit improper.

I'm sure that the function [itex]x^{x}[/itex] is Riemann integrable on a subdomain from R...

Daniel.
 
Yes, it is, of course (it's continuous on closed subintervals of [itex]\mathbb{R}^+[/itex], and as demonstrated above, differentiable too!). There's no closed-form antiderivative though.
 
[tex]\int x^{x} \ dx[/tex] cannot be expressed in terms of "common" special functions...It it were,i'm sure Mathematica would have done it...

Daniel.
 
  • #10
I loved it when I first worked out the minimum point of xx, really brought calc alive for me. I didn't have it set as an example I just realized as soon as I had learned all the needed things that I could do it.
 
  • #11
dextercioby said:
[tex]\int x^{x} \ dx[/tex] cannot be expressed in terms of "common" special functions...It it were,i'm sure Mathematica would have done it...

Proof by Mathematica?
j/k
 
  • #12
Yeah,go the integrator's webpage and type that function.The calculations are made with the latest version of the software.

Daniel.
 
  • #13
dextercioby said:
Yeah,go the integrator's webpage and type that function.The calculations are made with the latest version of the software.

Daniel.
Actually not true, I've now reported a series of integrations to Wolfram on things their online integrator should be able to integrate and can't. But things fairly obscure and I don’t think any people other than very bored ones like me will find it.
 
  • #14
Give an example.You can't contradict me with words,not in math,at least.

Daniel.
 
  • #15
lambert function and product i know have something to do with x^x=y
 
  • #16
as to x^x, I usually called it george, but sometimes ozymandias.

I have occasionally wondered what the name of the following number is as well:

0.01001000100001000001...

i also have found simple partial fractions that mathematica would not do, but which i did myself with pencil and paper.

More importantly to me, when I was teaching graduate algebra and galois theory i used mathematica to do some of the discriminant calculations for solutions formulas of cubics and quadratics, and found that sometimes mathematica just gave the wrong answer.

It did not say it couldn't do it, or that I had enetered it wrong (usually the case) but simply gave a false answer.

The errors I made in entering data were subtle, like omitting a space somewhere, but the upshot was that a naive student, who believed everything he got back, would have been misled by these incorrect responses, just as when using any calculator. So you always have to know how to check the validity of what you get from a calculator or computer.
 
Last edited:
  • #17
Reminds of my Maple.I had to reverse a fraction in an ODE so he could be able to solve it...

A software is made by a human.Computers beat human mind for speed/efficiency.

Daniel.
 
  • #18
dextercioby said:
Give an example.You can't contradict me with words,not in math,at least.

Daniel.
It can compute the two following:

[tex]\int \frac{1}{e^x + 2 e^{2 x} + 3 e^{3 x}} dx[/tex]

[tex]\int \frac{x^2}{e^x + 2 e^{2 x} + 3 e^{3 x}} dx[/tex]

But not:

[tex]\int \frac{1 + x^2}{e^x + 2 e^{2 x} + 3 e^{3 x}} dx[/tex]
 
  • #19
:smile: Yeah,it's a big black bug there...

Mathematica latest version won't do this

[tex]\int \frac{x+1}{e^{x}+2e^{2x}+3e^{3x}} \ dx[/tex]

A 10 y.o. Maple returns proudly

[tex]\int \frac{x+1}{e^x+2e^{2x}+3e^{3x}}dx=\allowbreak -\frac {1}{e^x}\ln \left( e^x\right) -\frac {2}{e^x}-\ln ^2\left( e^x\right)+\allowbreak i\ln \left( -\frac {1}{3}-\frac {1}{3}i\sqrt{2}\right) \arctan \frac {1}{2}\sqrt{2}[/tex]

[tex]+\frac {1}{8}i\sqrt{2}\ln \left( -\frac {1}{3}+\frac {1}{3}i\sqrt{2}\right) \ln \left( 1+2e^x+3e^{2x}\right) -\allowbreak \frac {1}{4}\sqrt{2}\ln \left( -\frac {1}{3}+\frac {1}{3}i\sqrt{2}\right) \arctan \left( \frac {3}{2}e^x\sqrt{2}+\frac {1}2\sqrt{2}\right)[/tex]

[tex]+\frac {1}{4}\sqrt{2}\ln \left( -\frac {1}{3}+\frac {1}{3}i\sqrt{2}\right) \arctan \frac {1}{2}\sqrt{2}+\allowbreak \frac {1}{2}\ln \left( -\frac {1}{3}+\frac {1}{3}i\sqrt{2}\right) \ln \left( 1+2e^x+3e^{2x}\right)[/tex]

[tex]-\allowbreak i\ln \left( -\frac {1}{3}+\frac {1}{3}i\sqrt{2}\right) \arctan \frac {1}{2}\sqrt{2}-\frac {1}{8}i\sqrt{2}\ln \left( -\frac {1}{3}-\frac {1}{3}i\sqrt{2}\right) \ln \left( 1+2e^x+3e^{2x}\right)[/tex]

[tex]-\allowbreak \frac {1}{4}\sqrt{2}\ln \left( -\frac {1}{3}-\frac {1}{3}i\sqrt{2}\right) \arctan \left( \frac {3}{2}e^x\sqrt{2}+\frac {1}{2}\sqrt{2}\right) +\frac {1}{4}\sqrt{2}\ln \left( -\frac {1}{3}-\frac {1}{3}i\sqrt{2}\right) \arctan \frac {1}{2}\sqrt{2}[/tex]

[tex]+\allowbreak \frac {1}{2}\ln \left( -\frac {1}{3}-\frac {1}{3}i\sqrt{2}\right) \ln \left( 1+2e^x+3e^{2x}\right) -i\ln \left( -\frac {1}{3}-\frac {1}{3}i\sqrt{2}\right) \arctan \left( \frac {3}{2}e^x\sqrt{2}+\frac {1}{2}\sqrt{2}\right)[/tex]

[tex]-\frac {1}{4}i\sqrt{2}\ \mbox{dilog}\left( 3\frac{e^x}{-1+i\sqrt{2}}\right) - \ \mbox{dilog}\left( 3\frac{e^x}{-1+i\sqrt{2}}\right) +\allowbreak \frac {1}{4}i\sqrt{2} \ \mbox{dilog}\left( -3\frac{e^x}{1+i\sqrt{2}}\right) - \ \mbox{dilog}\left( -3\frac{e^x}{1+i\sqrt{2}}\right)[/tex]

[tex]-2\ln \left( e^x\right) +\allowbreak \ln \left( 1+2e^x+3e^{2x}\right) -\frac {1}{2}\sqrt{2}\arctan \frac {1}{4}\left( 2+6e^x\right) \sqrt{2} <br /> +i\ln \left( -\frac {1}{3}+\frac {1}{3}i\sqrt{2}\right) \arctan \left( \frac {3}{2}e^x\sqrt{2}+\frac {1}{2}\sqrt{2}\right)+ C'[/tex]


:approve:

Daniel.
 
Last edited:
  • #20
dextercioby said:
...
A 10 y.o. Maple returns proudly
...
[the spawn of Satan himself]
...
:approve:

Daniel.

Quick, someone take the derivative and find out whether it's right. :-p
 
Last edited:
  • #21
hypermorphism said:
Quick, someone take the derivative and find out whether its right. :-p

Learn how to spell words and then be a wise-ass...

Daniel.
 
  • #22
dextercioby said:
Learn how to spell words and then be a wise-ass...

Daniel.

w-o-r-d-s ? :rolleyes:

Seriously, that would be tedious to differentiate by hand (or even to enter into Mathematica).
 
  • #23
[tex]\frac {d}{dx}\ \mbox{dilog} x =\allowbreak \frac{\ln x}{1-x}[/tex]

is all u need to know.There are 22 terms,outta which a few are constant...

Daniel.
 
  • #24
dextercioby said:
:smile: Yeah,it's a big black bug there...

Mathematica latest version won't do this

[tex]\int \frac{x+1}{e^{x}+2e^{2x}+3e^{3x}} \ dx[/tex]

A 10 y.o. Maple returns proudly
Oi!

Don't bad mouth the actual Mathematica, it can integrate that fine, even if the web version can't. Same with the other problem, I was just pointing out that the web version doesn't use the latest software because it can't do things it should be able to, I have found other things as well it can't do.
 
  • #25
tongos said:
lambert function and product i know have something to do with x^x=y

I sometimes think of x^x as being kind of similar to factorial or Gamma. What I mean is it seems to be usful if you're trying to think of a function that is bigger than factorial but not too much bigger.
 
  • #26
[tex]x^x=x\uparrow\uparrow2[/tex] in Knuth's up-arrow terminology; it's [tex]x\rightarrow2\rightarrow2[/tex], as I recall, in Conway's chained arrow notation.
 

Similar threads

  • · Replies 12 ·
Replies
12
Views
2K
  • · Replies 3 ·
Replies
3
Views
4K
  • · Replies 36 ·
2
Replies
36
Views
7K
  • · Replies 10 ·
Replies
10
Views
6K
  • · Replies 5 ·
Replies
5
Views
3K
  • · Replies 3 ·
Replies
3
Views
3K
  • · Replies 4 ·
Replies
4
Views
1K
  • · Replies 3 ·
Replies
3
Views
3K
  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 16 ·
Replies
16
Views
3K