Composite function and definite integration

Click For Summary

Homework Help Overview

The discussion revolves around the evaluation of the definite integral of a composite function, specifically \(\int_{\frac{1}{4}}^{\frac{3}{4}} f(f(x))dx\), where \(f(x) = x^3 - \frac{3x^2}{2} + x + \frac{1}{4}\). Participants explore the implications of the function's properties and the challenges of expanding the composite function.

Discussion Character

  • Exploratory, Conceptual clarification, Problem interpretation, Assumption checking

Approaches and Questions Raised

  • Participants discuss the complexity of expanding \(f(f(x))\) and whether it is necessary for solving the integral. Some consider estimating the area under the curve using graphical methods or rectangles, while others question the validity of these approaches. The existence of a fixed point and the function's monotonicity are also noted.

Discussion Status

The discussion is ongoing, with various perspectives on how to approach the integral. Some participants suggest graphical estimation methods, while others emphasize the need for a more precise calculation. There is recognition of the limitations imposed by the problem's context, including the requirement for an integer result.

Contextual Notes

Participants note that the integral's result is expected to be an integer between 0 and 9, which raises questions about the nature of the integral itself. There is also mention of a potential typo regarding the expected outcome of the integral.

Saitama
Messages
4,244
Reaction score
93

Homework Statement


[tex]f(x)=x^3-\frac{3x^2}{2}+x+\frac{1}{4}[/tex]
find[tex]\int_{\frac{1}{4}}^{\frac{3}{4}} f(f(x))dx[/tex]


Homework Equations





The Attempt at a Solution


I am clueless here. I started by writing f(f(x)) as
[tex](f(x))^3-\frac{3(f(x))^2}{2}+f(x)+\frac{1}{4}[/tex]
I don't think expanding (f(x))^3 would be of any use because this question is from a test paper and it would take a long time if i expanded (f(x))^3 and the other terms. I don't have any idea how to proceed. I observed that the f(x) is always an increasing function, therefore only one real root for f(x) exist but i don't think that would be of any use here.

Any help is appreciated!
 
Physics news on Phys.org
Hey Pranav-Arora!

Is this about a multiple choice perhaps?

Then you'd be supposed to make an estimate (using inscribed rectangles) and match the answer.
Or alternatively make a drawing and measure the area.

I agree that expanding everything is rather time consuming.
 
Hey ILS! :smile:

I like Serena said:
Is this about a multiple choice perhaps?
Nope, this question was from a section where the answer is an integer from 0 to 9.

Then you'd be supposed to make an estimate (using inscribed rectangles) and match the answer.
Or alternatively make a drawing and measure the area.
I don't understand why are you talking about rectangles here. Maybe i could estimate the area using the graph separately for (f(x))^3 and (f(x))^2.

Isn't there any proper method to get an exact answer?
 
Ah, so you do have at least some limitation on the acceptable answers.

An integral is the area under a graph.
220px-Integral_as_region_under_curve.svg.png


A first estimate is the area of the rectangle that has a height equal to a point on the graph, and a width equal to the boundaries.
Somewhat like this (for an integral with boundaries 0.5 and 1.5):
http://www.livemath.com/hoffnercalculus/preview/Chapter%206/Chpt6_Resources/image37.gif

In your case you could take f(f(1/2)) as the height and (3/4 - 1/4) as the width.
I believe that will suffice to find the proper answer, since that has to be an integer.

The method to get the exact answer is exactly what you were already doing.
I do not see ways to significantly speed up that process.
 
Last edited by a moderator:
  • Like
Likes   Reactions: 1 person
Is there a typo? I do not think that integral is an integer.
Are you familiar with

[tex](b-a)L\le \int_a^b f(x) dx \le (b-a)U[/tex]
where U and L are numbers such that
L<f<U
 
Pranav-Arora said:

Homework Statement

[tex]f(x)=x^3-\frac{3x^2}{2}+x+\frac{1}{4}[/tex]find[tex]\int_{\frac{1}{4}}^{\frac{3}{4}} f(f(x))dx[/tex]

Homework Equations



The Attempt at a Solution


I am clueless here. I started by writing f(f(x)) as[tex](f(x))^3-\frac{3(f(x))^2}{2}+f(x)+\frac{1}{4}[/tex]I don't think expanding (f(x))^3 would be of any use because this question is from a test paper and it would take a long time if i expanded (f(x))^3 and the other terms. I don't have any idea how to proceed. I observed that the f(x) is always an increasing function, therefore only one real root for f(x) exist but i don't think that would be of any use here.

Any help is appreciated!
There are several things to notice here.

The point, (1/2, 1/2), is a fixed point. So, f(1/2) = 1/2, and f(f(1/2)) = 1/2 .

Also, the point, (1/2, 1/2) is a point of inflection for f(x), and f(x) is a cubic function, so the graph, y = f(x) is invariant under a 180° rotation about the fixed point, (1/2, 1/2). In other words, if you shift the graph, y = f(x), to the left by 1/2 and down by 1/2, the resulting graph, y = f(x + 1/2) - 1/2 is symmetric with respect to the origin.

Thus, the function, g(x) = f(x + 1/2) - 1/2 is an odd function, actually an odd cubic function.

The function, g(g(x)), is also an odd function, so [itex]\displaystyle \int_{-1/4}^{1/4}g(g(x))\,dx=0\ .[/itex]

You can also show, graphically or otherwise, that y=g(g(x)) is a shifted version of y=f(f(x)) .

So I like Serena's idea of using a rectangle is relevant.
 
Better yet:

Define [itex]\displaystyle g(x) = f(x+(1/2)) - 1/2[/itex], which is an odd (cubic) function.

Now for the integral [itex]\displaystyle \int_{1/4}^{3/4}f(f(x))\,dx\ :[/itex]

Do the substitution x = u + 1/2, then dx = du. x = 1/4, 3/4 becomes u = -1/4, 1/4 .

[itex]\displaystyle \int_{1/4}^{3/4}f(f(x))\,dx[/itex]
[itex]\displaystyle =\int_{-1/4}^{1/4}f(f(u + 1/2))\,du[/itex]

[itex]\displaystyle =\int_{-1/4}^{1/4}f(f(u + 1/2))-\frac{1}{2}+\frac{1}{2}\,du[/itex]

[itex]\displaystyle =\int_{-1/4}^{1/4}f(f(u + 1/2))-\frac{1}{2}\,du+\int_{-1/4}^{1/4}\frac{1}{2}\,du[/itex]

[itex]\displaystyle =\int_{-1/4}^{1/4}f\left(f(u + 1/2)-1/2 +\frac{1}{2}\right)-\frac{1}{2}\,du+\int_{-1/4}^{1/4}\frac{1}{2}\,du[/itex]

[itex]\displaystyle =\int_{-1/4}^{1/4}f\left(g(u) +\frac{1}{2}\right)-\frac{1}{2}\,du+\int_{-1/4}^{1/4}\frac{1}{2}\,du[/itex]

[itex]\displaystyle =\int_{-1/4}^{1/4}g(g(u))\,du+\int_{-1/4}^{1/4}\frac{1}{2}\,du[/itex]
 
  • Like
Likes   Reactions: 1 person
I did not analyze as far as you did, but if I understand you correctly, the area of the rectangle is equal to the exact solution. ;)
And indeed, that is not a whole number.
 
I like Serena said:
I did not analyze as far as you did, but if I understand you correctly, the area of the rectangle is equal to the exact solution. ;)
And indeed, that is not a whole number.
Yes, that's correct.

It is the reciprocal of a whole number.
 
  • #10
Woops, sorry for a late reply.

lurflurf said:
Is there a typo? I do not think that integral is an integer.
Sorry, there is a typo. The reciprocal of the definite integral is the answer.
lurflurf said:
Are you familiar with

[tex](b-a)L\le \int_a^b f(x) dx \le (b-a)U[/tex]
where U and L are numbers such that
L<f<U

Yes, i am familiar with that. Thanks for that, this might be useful to estimate the answer. :smile:

SammyS said:
There are several things to notice here.

The point, (1/2, 1/2), is a fixed point. So, f(1/2) = 1/2, and f(f(1/2)) = 1/2 .

Also, the point, (1/2, 1/2) is a point of inflection for f(x), and f(x) is a cubic function, so the graph, y = f(x) is invariant under a 180° rotation about the fixed point, (1/2, 1/2). In other words, if you shift the graph, y = f(x), to the left by 1/2 and down by 1/2, the resulting graph, y = f(x + 1/2) - 1/2 is symmetric with respect to the origin.

Thus, the function, g(x) = f(x + 1/2) - 1/2 is an odd function, actually an odd cubic function.

The function, g(g(x)), is also an odd function, so [itex]\displaystyle \int_{-1/4}^{1/4}g(g(x))\,dx=0\ .[/itex]

You can also show, graphically or otherwise, that y=g(g(x)) is a shifted version of y=f(f(x)) .

So I like Serena's idea of using a rectangle is relevant.
Nicely explained SammyS. :smile:

SammyS said:
Better yet:

Define [itex]\displaystyle g(x) = f(x+(1/2)) - 1/2[/itex], which is an odd (cubic) function.

Now for the integral [itex]\displaystyle \int_{1/4}^{3/4}f(f(x))\,dx\ :[/itex]

Do the substitution x = u + 1/2, then dx = du. x = 1/4, 3/4 becomes u = -1/4, 1/4 .

[itex]\displaystyle \int_{1/4}^{3/4}f(f(x))\,dx[/itex]
[itex]\displaystyle =\int_{-1/4}^{1/4}f(f(u + 1/2))\,du[/itex]

[itex]\displaystyle =\int_{-1/4}^{1/4}f(f(u + 1/2))-\frac{1}{2}+\frac{1}{2}\,du[/itex]

[itex]\displaystyle =\int_{-1/4}^{1/4}f(f(u + 1/2))-\frac{1}{2}\,du+\int_{-1/4}^{1/4}\frac{1}{2}\,du[/itex]

[itex]\displaystyle =\int_{-1/4}^{1/4}f\left(f(u + 1/2)-1/2 +\frac{1}{2}\right)-\frac{1}{2}\,du+\int_{-1/4}^{1/4}\frac{1}{2}\,du[/itex]

[itex]\displaystyle =\int_{-1/4}^{1/4}f\left(g(u) +\frac{1}{2}\right)-\frac{1}{2}\,du+\int_{-1/4}^{1/4}\frac{1}{2}\,du[/itex]

[itex]\displaystyle =\int_{-1/4}^{1/4}g(g(u))\,du+\int_{-1/4}^{1/4}\frac{1}{2}\,du[/itex]
Excellent! How do they expect us to think like this and that too in the examination room.
The solution would be take me some time to understand it completely, i will get back to this thread if i get stuck.

I will mention it again, there's a typo. The answer is the reciprocal of the value of the definite integral.
 
Last edited:

Similar threads

  • · Replies 4 ·
Replies
4
Views
1K
  • · Replies 14 ·
Replies
14
Views
2K
Replies
4
Views
3K
  • · Replies 10 ·
Replies
10
Views
3K
  • · Replies 4 ·
Replies
4
Views
2K
  • · Replies 105 ·
4
Replies
105
Views
11K
  • · Replies 9 ·
Replies
9
Views
2K
  • · Replies 8 ·
Replies
8
Views
1K
  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 14 ·
Replies
14
Views
2K