| Author |
Message |
![[Post New]](/templates/default/images/icon_minipost_new.gif) 4 Jul 2008 22:59:31 IST
|
|
|
Find all tame solutions of:::::
1) f^2(x)=f(x+y)f(x-y)
2) f(x+y)+f(x-y)=2f(x)
3) f(x+y)=[f(x)f(y)]/[f(x)+f(y)]
rates assured for complete solution
|
"You need a perfect strategy to clear JEE more than knowledge " |
|
|
|
![[Post New]](/templates/default/images/icon_minipost_new.gif) 11 Jul 2008 00:06:00 IST
|
|
|
Helpppppppppp
|
"You need a perfect strategy to clear JEE more than knowledge " |
this reply: 0 points
(with 0 
in 0 votes ) [?]
|
|
You have to be logged on to rate
|
|
|
![[Post New]](/templates/default/images/icon_minipost_new.gif) 11 Jul 2008 13:11:47 IST
|
|
|

2) (c1)x + c
|
nobody is perfect......i m nobody.............. |
this reply: 0 points
(with 0 
in 0 votes ) [?]
|
|
You have to be logged on to rate
|
|
|
![[Post New]](/templates/default/images/icon_minipost_new.gif) 11 Jul 2008 18:07:32 IST
|
|
|
Ok, the 1st one:
If f(0) = 0, then f(x) = 0 for all x.
Else:
It is easy to see that the function is either positive for all x or negative for all x. Also, if f(x) is a solution, so is -f(x).
So, without loss of generality, assume

(I have dropped the base a for now, I will bring it back only in the end as typing out the base is a pain!)
Call g(x) = log f(x)
Then 
This means the function is neither convex or concave on any interval. This means the function is linear
i.e. g(x) = px+q where p and q are constants.

|
Time wounds all heels |
this reply: 10 points
(with 2 
in 2 votes ) [?]
|
|
You have to be logged on to rate
|
|
|
![[Post New]](/templates/default/images/icon_minipost_new.gif) 11 Jul 2008 18:10:40 IST
|
|
|
2nd one:
The solution is partly contained in the above post.
Again, we have,

|
Time wounds all heels |
this reply: 10 points
(with 2 
in 2 votes ) [?]
|
|
You have to be logged on to rate
|
|
|
![[Post New]](/templates/default/images/icon_minipost_new.gif) 11 Jul 2008 18:20:11 IST
|
|
|
I suspect the last one has not been answered till now only because of the choice of colour.

Now, it's straightforward as 
and that is one of the Cauchy Functional Equations (see http://www.goiit.com/posts/list/algebra-sir-please-reply-me-71982.htm#354815) with the general solution g(x) = kx
Hence the solution is 
|
Time wounds all heels |
this reply: 5 points
(with 1 
in 1 votes ) [?]
|
|
You have to be logged on to rate
|
|
|
![[Post New]](/templates/default/images/icon_minipost_new.gif) 11 Jul 2008 18:25:28 IST
|
|
|
edit: In the first solution I wrote:
"It is easy to see that the function is either positive for all x or negative for all x. Also, if f(x) is a solution, so is -f(x).
So, without loss of generality, assume ..."
The assumption was meant to be: "f(x) > 0 for all x"
|
Time wounds all heels |
this reply: 0 points
(with 0 
in 0 votes ) [?]
|
|
You have to be logged on to rate
|
|
|
![[Post New]](/templates/default/images/icon_minipost_new.gif) 11 Jul 2008 18:54:03 IST
|
|
|
Ok to the list of standard functional equations, add the Jensen's equation:

with the general solution f(x) = ax+b as indicated above.
The proof, if you dont want to invoke the arguments of convexity and concavity is as follows:
Setting f(0) = a and then y = 0, we get

So, now
![\frac{f(x)+f(y)}{2} = f \left( \frac{x+y}{2} \right) = \frac{f(x+y)+a}{2} \\ \\<br/>\Rightarrow f(x) + f(y) = f(x+y)+a \ \text{or} \ [f(x)-a]+[f(y)-a] = [f(x+y) - a]](http://alt1.artofproblemsolving.com/Forum/latexrender/pictures/0/2/3/023ae54c255901b54c3e0c136f2cec563f612d7f.gif)
Setting g(x) = f(x) - a,
we get g(x)+g(y) = g(x+y)
Now we have reduced it to the standard Cauchy functional equation with solution g(x) = mx
Hence f(x) = mx+a.
|
Time wounds all heels |
this reply: 10 points
(with 2 
in 2 votes ) [?]
|
|
You have to be logged on to rate
|
|
|
![[Post New]](/templates/default/images/icon_minipost_new.gif) 11 Jul 2008 21:50:13 IST
|
|
|
U don't need at all that mumbo-jumbo ( concavity -convexity or the above proof ) to solve the functional eqn
It is very simple to solve .
first differentiate partially both sides wrt 'y' , yielding
....................(1)
Now put y=0 in both sides , yielding f'(x/2)=f'(0) = constant .......(2)
from (2) it follows that f(x) =mx +c ( proved)
|
this reply: 7 points
(with 1 
in 2 votes ) [?]
|
|
You have to be logged on to rate
|
|
|
![[Post New]](/templates/default/images/icon_minipost_new.gif) 11 Jul 2008 21:52:16 IST
|
|
|
Now somebody may object that it( my proof) requires the differentiability of f(x) .
My answer is : Yes , it does require that .
|
this reply: 0 points
(with 0 
in 0 votes ) [?]
|
|
You have to be logged on to rate
|
|
|
![[Post New]](/templates/default/images/icon_minipost_new.gif) 12 Jul 2008 09:22:39 IST
|
|
|
Generally, while solving functional equations, unless specifically stated, it is not a done thing to assume differentiability.
That is why, the definition of convexity is given in the form of the following inequality
instead of the more compact f"(x)>0, so that even functions that are differentiable not everywhere (but countably many points) also qualify.
|
Time wounds all heels |
this reply: 0 points
(with 0 
in 0 votes ) [?]
|
|
You have to be logged on to rate
|
|
|
![[Post New]](/templates/default/images/icon_minipost_new.gif) 12 Jul 2008 13:23:58 IST
|
|
|
I don't require differentiability at every point . If it is differentiable at a single point ( let that be zero point )then as my proof shows it is differentiable at all other points .
|
this reply: 0 points
(with 0 
in 0 votes ) [?]
|
|
You have to be logged on to rate
|
|
|
![[Post New]](/templates/default/images/icon_minipost_new.gif) 12 Jul 2008 13:28:14 IST
|
|
|
It only proves that where it is differentiable, the derivative is a constant
|
Time wounds all heels |
this reply: 0 points
(with 0 
in 0 votes ) [?]
|
|
You have to be logged on to rate
|
|
|
![[Post New]](/templates/default/images/icon_minipost_new.gif) 12 Jul 2008 13:42:33 IST
|
|
|