Return Styles: Pseud0ch, Terminal, Valhalla, NES, Geocities, Blue Moon. Entire thread

X^X

Name: Anonymous 2007-11-06 18:17

integrate x^x
not possible?

Name: Anonymous 2007-11-06 20:04

It should be possible, but I have no idea.

Name: Anonymous 2007-11-06 20:17

(x^x)*(1 + ln(x))

Name: Anonymous 2007-11-06 20:20

The indefinite integral x^x dy is x^x * y + c, dummy.

Name: Meranda, Matt 2007-11-06 20:21

ya'll are a buncho losersss

Name: Anonymous 2007-11-06 20:25

>>3

This is >>2
you've differentiated, douche-bag, everyone knows how to do that one.

Name: Anonymous 2007-11-06 23:45

>>1
Not possible as the problem stands.  I need to know what variable I'm integrating with respect to.

Name: Anonymous 2007-11-07 3:25

bump for great justice

Name: Anonymous 2007-11-07 11:38

X^X
=1^X+1, or am i retarded?

Name: Anonymous 2007-11-07 11:50

>>9
Yes. Yes you are.

Name: Anonymous 2007-11-07 12:26

>>7
lets say it's (x^x) dx
is it possible?

Name: Anonymous 2007-11-07 13:52

INT[ x^x . dx ] = f(x)
where, f(x) = INT[ x^x . dx ]

Name: Anonymous 2007-11-07 13:55

>>9
To clarify: 1^(X+1) is always 1, obviously.
I'm guessing your reasoning went like this:

1. Given (X*X)^(X-1)
2. Can be simplified to X^X
3. Can be simplified to 1^(X+1), as per step 2.

The problem is that step 2 is false. 1 can be rewritten as (X^2)^(X-1), or X^((X-1)*2), not X^((X-1)+1). Rewriting X^X in the same way gives X^(X*1), not 1^(X+1).

This is middle school material.

Name: Anonymous 2007-11-07 14:19

let F(x) = f(x)x^x
    F'(x) = x^x

gives us a first order ODE to find f(x).

=> f'(x) + (1 + ln x ) f(x) = 1

couldn't be fucked to do anything actually interesting with it like fourier transform, trying to solve by an integrating factor led to the rather annoying integrating factor x^-x
Fuck yeah.

Might give it some more though later, have a degree to do though

Name: Anonymous 2007-11-07 16:20

>>1

x^x = exp(x*ln(x))

int(exp(f(x))*dx) = exp(f(x)/f'(x) (if I'm not a complete idiot now :?)
=(x^x)/(1+ln(x))

Name: Anonymous 2007-11-07 19:55

>>15
almost entirely incorrect.

try differentiating the expression you just got exp(f(x))/f'(x)


= exp(f(x))[1 + d/dx(1/f'(x))]  != exp(f(x))

Name: Anonymous 2007-11-07 21:03

int(exp(f(x))*dx) = exp(f(x)/f'(x)) Only if f(x) is linear in x. In which case f'(x) is just a constant.

>>14
Seems to have the best idea.

Name: Anonymous 2007-11-08 0:54

The function is Riemann-Stieltjes integrable since it is monotone, so an integral does exist. However, the antiderivative can't be expressed as a finite composition of the traditional elementary functions as a result of differential Galois theory.

Name: Anonymous 2007-11-08 2:17

i do all this numerically, since i have a computer.
lol at all you losers who dont have computers

Name: Anonymous 2007-11-08 2:28

>>19
lol finite precision

Name: Anonymous 2007-11-08 12:31

>>18

Care to elaborate?

Name: Anonymous 2007-11-08 15:41

FROSTED BUTTS
lol 4chan cant do math

Name: Anonymous 2007-11-08 18:47

erm

e^( .5 x^2 lnx - x^2 / 4 + c)=int(x^x)dx

Name: Anonymous 2007-11-08 18:52

expand in a taylor series and integrate that?

Name: Anonymous 2007-11-08 19:07

>>24
The Taylor terms have x^x in them.

Name: Anonymous 2007-11-08 19:25

>>25
That isn't possible you nitwit.


>>23
e^(0.5x^2*ln(x) - x^2/4 + c) = e^(0.5x^2*ln(x) - x^2/4) * e^c. In other words, the derivative of the "integral" you gave depends on c, which means (until you at least specify a value of c) it cannot be the integral of x^x.

Name: Anonymous 2007-11-08 21:48

>>25
no they dont

Name: Anonymous 2007-11-08 23:27

>>1
Maxima doesn't know how to do it.  Probably not possible.

Name: Anonymous 2007-11-10 0:49

there is no analytical solution you asstards.

Name: Anonymous 2007-11-10 2:18

LOL I GOT YOU STUPID NIGGERS.
YOU FUCKERS ALL THINK YOU KNOW SO MUCH SHIT ABOUT EVERYTHING BUT FAIL TO RECOGNIZE ONE OF THE MOST BASIC NONINTEGRABLE FUNCTIONS.
FUCKING RETARDED LOSERS LOL

Name: Anonymous 2007-11-10 2:45

>>30
It is integrable, you just can't express it as a closed form solution. gb2elementary analysis

Name: Anonymous 2007-11-10 5:39

>>30
What the fuck, is your caps lock key stuck down? Turn it fucking off!

Name: Anonymous 2007-11-10 6:27

>>32
Mommy, the big kids are shouting at me!

Name: ANONYMOUS 2007-11-10 21:11

DID YOU KNOW "SINE" MEANS BREAST? ITS A MIS TRANSLATION ON THE GREEKS PART. SINE!!!! BREAST!!!! HAHAHAHA 300BREAST30=150

Name: Anonymous 2007-11-10 21:53

>>34
What are you using a Commodore 64? Even that has lower case if you press two keys.

Name: Anonymous 2007-11-11 3:39

>>35
Whining about capslock is not a meme.

Name: Anonymous 2007-11-11 3:59

>>36
Don't be a cunt.

Name: Anonymous 2007-11-11 5:49

>>34
Mistranslation FROM Greek, retard.  Suck a fuck.

Name: Anonymous 2007-11-11 7:34

>>37
I'll be a cunt if I want to, cockpouch. Quit shitting up /sci/ with your whining.

Name: Anonymous 2007-11-11 8:48

Am I doin it rite?

int(x^x)dx
=int(exp(xln[x]))dx
=int(sum_n=0^infty{xln[x])^n/n!})dx
=sum_n=0^infty{int(xln[x])^n)dx/n!}
=sum_n=0^infty{((Gamma[1 + n, -((1 + n)*Log[x])]* Log[x]*(-((1 + n)*Log[x]))^(-1 - n)* (x*Log[x])^n)/x^n)/n!}

There you go.

Newer Posts
Don't change these.
Name: Email:
Entire Thread Thread List