Find all functions of this integral equation, very tricky

Click For Summary

Homework Help Overview

The discussion revolves around an integral equation involving an unknown function f(x) defined by the equation \(\int{f(x)dx}\int{\frac{1}{f(x)}dx} = -1\). Participants explore potential solutions and the implications of various manipulations of the equation.

Discussion Character

  • Exploratory, Mathematical reasoning, Assumption checking

Approaches and Questions Raised

  • Participants discuss various attempts to manipulate the integral equation, including taking derivatives and exploring relationships between f(x) and its integrals. Some suggest specific functions like \(e^x\) and \(e^{-x}\) as potential solutions, while others express uncertainty about their validity.

Discussion Status

The conversation is ongoing, with participants providing hints and suggestions for further exploration. There is a mix of agreement and questioning regarding the validity of certain functions as solutions, and some participants are attempting to clarify the implications of their manipulations.

Contextual Notes

Participants note the complexity of the integral and the challenges associated with finding a formal solution. There are indications of confusion regarding the use of integrals and derivatives in the context of the problem.

DrummingAtom
Messages
657
Reaction score
2

Homework Statement



\int{f(x)dx}\int{\frac{1}{f(x)}dx} = -1

Homework Equations


The Attempt at a Solution



At first glance, I thought ex is a solution. But I'm not convinced yet because I didn't formally go through it. When I did it turns into some nasty integral. Here's a couple things I've tried:

From here I took the derivative of each side and then cleaned it up:

\int{f(x)dx}\int{\frac{1}{f(x)}dx} = -1

Evaluating the left side integral(if this is allowed) and some algebra,.. gives:

f(x)\int{\frac{1}{f(x)}dx} + \int{f(x)dx}\frac{1}{f(x)} = 0

I then took the derivative of each side again and then some algebra gives:

\int{f(x)} = -f(x)^2*ln(f(x))f(x) = -2f(x)*f ' (x)*ln(f(x)) - f(x)*f ' (x)

Cleaning it up and canceling gives finally:

f '(x) = \frac{1}{-2}*\frac{1}{ln(f(x))-1}

Which then I would just need to integrate both sides and there's my final function. That integral is nasty (wolfram gave me some strange function as the solution). Which makes me think that it must be wrong. Here's a second try at the solution.
The hint was: \int{\frac{1}{f(x)}dx} = -\int{f(x)}^{-2}*f(x)} by the chain rule.

If the hint is used then:

\int{f(x)dx}\int{\frac{1}{f(x)}dx} = -1

After integrating both sides:

f(x) = \int{f(x)}dx^{-2}*f(x)

Then:

\int{f(x)}^{2} = 1

From here I can square root and give:

\int{f(x)} = 1

If this is the case I have no clue as to what integrated function gives 1 as the solution.

Any help would be greatly appreciated. Thank you.
 
Physics news on Phys.org
Hi DrummingAtom! :smile:

Be simple! :rolleyes:

start "if g' = f then (1/g)' = … " :wink:
 
Thank tiny-tim,

By that, g' = f then (1/g)' = f '

So then plugging into the original problem would give:

∫g' * ∫(1/g)' = -1

or

∫f * ∫f` = -1

By that it gives:

g = -1/f

From this point do I just guess functions that fit this?
 
DrummingAtom said:
Thank tiny-tim,

By that, g' = f then (1/g)' = f '

So then plugging into the original problem would give:

∫g' * ∫(1/g)' = -1

or

∫f * ∫f` = -1

By that it gives:

g = -1/f

From this point do I just guess functions that fit this?

But if g' = f then (1/g)' is 1/g^2 * g', not what you gave.
 
Why isn't f(x) = ex a solution?
 
Char. Limit said:
But if g' = f then (1/g)' is 1/g^2 * g', not what you gave.

Ooopps, I was thinking inverse functions. Thanks for clearing that up.

Ok so using that instead would give:

∫g` = ∫f or g = ∫f

and

∫(1/g)` = (1/g^2 * g')


Which would all give:

g' = -g

Which the only function that would fit that is e-x, right?
 
SammyS said:
Why isn't f(x) = ex a solution?

I think it is too, but I don't know how to formally show that it is. I think that ex and e-x would have to work.
 
DrummingAtom said:
I think it is too, but I don't know how to formally show that it is. I think that ex and e-x would have to work.

If 1/e^(x) is a solution, as you proved that it is, then e^(x) must be a solution as well, because the original equation involves multiplying the integrals of f(x) and 1/f(x).
 
I got that f '(x) = ± f(x) .
 
  • #10
(just got up :zzz: …)
DrummingAtom said:
∫g` = ∫f or g = ∫f

and

∫(1/g)` = (1/g^2 * g')

DrummingAtom, stop using ∫s

the whole point of introducing g is to simplify by converting everything to differential equations, with no ∫s
Which would all give:

g' = -g

No, I get something with squareds in
 
  • #11
No wonder this is giving me trouble, I know next to nothing about differential equations.

I'll try this again:

If g' = f and (1/g)' = 1/g2*g'

then the original equation would say:

g' * (1/g2)*g' = -1

or

(g')2 = -g2

At this point I can't take the square root of a negative function, does this mean there are no real solutions?
 
  • #12
Hi DrummingAtom! :smile:

Yes, that's fine except …
DrummingAtom said:
If g' = f and (1/g)' = 1/g2*g' …

… it should be (1/g)' = -1/g2*g' :wink:

ok, now take the square-root of your last equation (and remember to include a ±). :smile:
 
  • #13
tiny-tim said:
(just got up :zzz: …)


DrummingAtom, stop using ∫s

the whole point of introducing g is to simplify by converting everything to differential equations, with no ∫s



You set f = g', which gives (1/g)' = -g'/g2 .

I don't see how these help with ∫ (1/f) dx .
 
  • #14
What I suggest:

\int{f(x)dx}\int{\frac{1}{f(x)}dx} = -1\quad\quad\to\quad\quad\int{\frac{1}{f(x)}dx} = \frac{-1}{\displaystyle \int{f(x)dx}}

Take the derivative w.r.t. x of both sides of the equation on the right.
 
  • #15
SammyS said:
What I suggest:

\int{f(x)dx}\int{\frac{1}{f(x)}dx} = -1\quad\quad\to\quad\quad\int{\frac{1}{f(x)}dx} = \frac{-1}{\displaystyle \int{f(x)dx}}

Take the derivative w.r.t. x of both sides of the equation on the right.

I can see that working for the left side because it's a set integral. But wouldn't you have to use the quotient rule for the \frac{-1}{\displaystyle \int{f(x)dx}}, doing so would give:

\frac{1}{f(x)} = \frac{f(x)}{\int{f(x)}^{2}dx}
 
  • #16
DrummingAtom said:
I can see that working for the left side because it's a set integral. But wouldn't you have to use the quotient rule for the \frac{-1}{\displaystyle \int{f(x)dx}}, doing so would give:

\frac{1}{f(x)} = \frac{f(x)}{\int{f(x)}^{2}dx}

Not quite!

\text{Let }u=\int{f(x)\,dx}, \text{ then } \frac{d}{dx}\left(\frac{1}{u}\right)=-\frac{1}{u^2}\,\frac{du}{dx}

So you should have: \frac{1}{f(x)}=\frac{f(x)}{\displaystyle \left(\int{f(x)}dx}\right)^{2}

Solve for f(x) and differentiate, or let g' = f & solve for g.
 
Last edited:
  • #17
Haha, yeah, my bad I should have finished that. Would that cause the e-x solution to disappear though? Or do you have to \pm on both sides?
 
  • #18
You will have ±. (You only need it on one side!)
 
  • #19
i stil don't understand why you're using all those integrals …

i got g'2 = g2 :confused:
 

Similar threads

  • · Replies 105 ·
4
Replies
105
Views
10K
Replies
9
Views
2K
  • · Replies 6 ·
Replies
6
Views
2K
Replies
3
Views
2K
  • · Replies 6 ·
Replies
6
Views
2K
  • · Replies 10 ·
Replies
10
Views
3K
Replies
8
Views
2K
  • · Replies 7 ·
Replies
7
Views
2K
Replies
3
Views
2K
Replies
2
Views
2K