A function f : R → R satisfies
f ( x − f ( y ) ) = f ( f ( y ) ) + x f ( y ) + f ( x ) − 1
for all x , y ∈ R .
Find f ( 2 0 1 6 ) .
This section requires Javascript.
You are seeing this because something didn't load right. We suggest you, (a) try
refreshing the page, (b) enabling javascript if it is disabled on your browser and,
finally, (c)
loading the
non-javascript version of this page
. We're sorry about the hassle.
@Chew-Seong Cheong @Shaun Leong @Priyanshu Mishra This is how to deal with the "What values of f ( y ) can we use?" In particular, note that I'm explicit with "Set x = . . . , y = . . . ", since we can only set those values.
f ( x − f ( y ) ) f ( 0 ) ⟹ 2 f ( f ( y ) ) 2 f ( x ) 2 f ( 0 ) ⟹ f ( 0 ) ⟹ f ( x ) ⟹ f ( 2 0 1 6 ) = f ( f ( y ) ) + x f ( y ) + f ( x ) − 1 = f ( f ( y ) ) + ( f ( y ) ) 2 + f ( f ( y ) ) − 1 = f ( 0 ) + 1 − ( f ( y ) ) 2 = f ( 0 ) + 1 − x 2 = f ( 0 ) + 1 = 1 = 1 − 2 x 2 = 1 − 2 2 0 1 6 2 = − 2 0 3 2 1 2 7 Putting x = f ( y ) Replacing f ( y ) with x Putting x = 0
Thanks for putting in the effort to write up the solution. Unfortunately, it is currently incorrect because:
At the IMO, this solution would be worth at most 1 point. You can see the Q6 point breakdown .
Practice makes perfect!
Log in to reply
Thanks, I get the points. I have changed the solution.
Log in to reply
Unfortunately, the solution is still incorrect, due to the same error.
In the first step, when you "Set f ( y ) = x ", then the only valid values of x that we can subsequently substitute into the equation, is the range of f . Thus, you still have to justify that the range of f contains 0 (similar to Shaun's solution below).
In my previous comment, the step of "Set x = f ( y ) " is valid without any restrictions on the potential values. This is because x is the variable here.
The proper solution to this problem requires an analysis of the range of f ( x ) , and possibly the range of f ( x ) − f ( y ) . This allows us to subsequently do the substitution of "Set f ( y ) = x " with the proper justification.
Log in to reply
@Calvin Lin – I have changed it again. It should be valid to put x = f ( y ) in the first step and then replace it back to x .
Log in to reply
@Chew-Seong Cheong – No, When you "replace it back to x ", remember that we are still restricted to the universe where " f ( y ) = x ", IE that x is in the range of f .
Specifically, if you check your final solution, then there is no value of y such that f ( y ) = 2 0 1 6 . And thus, we cannot do a substitution of x = 2 0 1 6 = f ( y ) , which means that we do not have 2 f ( 2 0 1 6 ) = f ( 0 ) + 1 − 2 0 1 6 2 .
I know that this is a somewhat subtle point of rigour. Please try and understand this in more detail, before adjusting the solution. AFAIK, there isn't a simple fix to this approach. All solutions that I know of put a lot of effort in studying the range properly. There's a reason why it's problem 6 of the IMO.
Log in to reply
@Calvin Lin – Thanks. I will think about it again.
Putting
x
=
1
and
f
(
y
)
=
1
.
We get
f
(
0
)
=
2
×
f
(
1
)
⟶
(
1
)
Now, let's put
x
=
2
and
f
(
y
)
=
1
.
We get
f
(
2
)
=
−
1
Now we put
x
=
2
and
f
(
y
)
=
0
which gives
f
(
0
)
=
1
.
If we put this value into the 1st equation we will get
f
(
1
)
=
2
1
Till now, We have
f
(
0
)
=
1
f
(
1
)
=
2
1
f
(
2
)
=
−
1
Now, let's try and find a pattern in the equation. If we put f ( y ) = 1 in the given equation,
f ( x − 1 ) = f ( 1 ) + x + f ( x ) − 1
i.e. f ( x ) = f ( x − 1 ) − x + 2 1 ⟶ ( 2 )
If we put x = 2 0 1 6 in 2nd equation,
f ( 2 0 1 6 ) = f ( 2 0 1 5 ) − 2 0 1 6 + 2 1 ⟶ ( 3 )
Similarly, if x = 2 0 1 5 , f ( 2 0 1 5 ) = f ( 2 0 1 4 ) − 2 0 1 5 + 2 1
We can put back the value of
f
(
2
0
1
5
)
in the 3rd equation
We will drill down to
f
(
2
)
similarly.
Therefore, f ( 2 0 1 6 ) = f ( 2 ) + 2 0 1 4 × 2 1 − ( 2 0 1 6 + 2 0 1 5 . . . . . . + 4 + 3 )
Putting the value of f ( 2 ) = − 1 .
We will get f ( 2 0 1 6 ) as -2032127.
So the correct answer is − 2 0 3 2 1 2 7
This is a good start. However, your explanation has a slight gap (though it's not exactly what the question asked for)
How do we know that such a function must exist? It is defined over the reals, and not just the integers.
Keep writing more solutions and you will get the hang of this!
Log in to reply
Sorry sir, i forgot to add negative sign to my question.
Sorry for others too.
At first this answer was not accepted. I was wondering where I went wrong, when it seemed obvious.
Log in to reply
The creator posted the wrong answer. It was reported by several members of the community and then corrected.
If you spot an error with a problem, you can select "Report Problem" from the menu. This allows the creator to respond to you accordingly.
Log in to reply
I usually do, but thought I may have erred somewhere so waited.
Log in to reply
@Siva Bathula – That works too :) Though, once if you also post your approach/reasoning, it will allow the creator (and staff) to see where you are wrong / determine if he is wrong.
Best solution. Its an IMO 1999 problem. All solutions are great.
Log in to reply
Note: None of the solutions are currently completely correct (as answers to IMO 1999/6). See my individual comments.
Log in to reply
Okay.
I will write my solution also and then compare it with these. I have not written it yet because it is in too detail and lengthy also.
I suggest a slightly faster solution where we don't need to solve for f ( 0 ) :
Rearranging,
f ( x ) − f ( x − f ( y ) ) = 1 − f ( f ( y ) ) − x f ( y )
From this we can see all real c can be expressed as f ( m ) − f ( n ) for some m , n .
(To see this, we choose y such that f ( y ) is not 0 (and this exists as f ( x ) = 0 ∀ x does not satisfy the functional equation). Then we can find a value of x such that 1 − f ( f ( y ) ) − x f ( y ) = c . So in this case m = x , n = x − f ( y ) .)
For all real m , n , letting x = f ( m ) , y = n we see that f ( f ( m ) − f ( n ) ) = f ( f ( n ) ) + f ( f ( m ) ) + f ( m ) f ( n ) − 1 which is symmetric about m , n . Thus f ( f ( m ) − f ( n ) ) = f ( f ( n ) − f ( m ) ) . But by the above fact, for all c , we have f ( c ) = f ( − c ) , so the function is even.
Now substitute x = 0 . 5 f ( y ) . We can do this as 0 . 5 f ( y ) is real. Then f ( x ) = f ( x − f ( y ) ) since the function is even, hence 1 − f ( f ( y ) ) − x f ( y ) = 0 .
Hence we have f ( f ( y ) ) = − 0 . 5 ( f ( y ) ) 2 + 1 for all y .
Let c = f ( m ) − f ( n ) , then
f ( c ) = f ( f ( m ) − f ( n ) ) = f ( f ( m ) ) + f ( f ( n ) ) + f ( m ) f ( n ) − 1
= ( − 0 . 5 [ f ( m ) ] 2 + 1 ) + ( − 0 . 5 [ f ( n ) ] 2 + 1 ) + f ( m ) f ( n ) − 1 = − 0 . 5 ( f ( m ) − f ( n ) ) 2 + 1 = − 0 . 5 c 2 + 1
Thus that's our solution! (Checking can be done easily.)
Problem Loading...
Note Loading...
Set Loading...
The only solution is f ( x ) = 1 − 2 x 2 . This problem is IMO 1999/6, and it's actually pretty hard to proof rigorously.
Note: Most solutions make the error of substituting in f ( y ) = C for some value C , but provide no justification for why this is possible. For example, to calculate f ( 2 0 1 6 ) , some solutions use "Let f ( y ) = 2 0 1 6 , then ... ". However, if observe the function f , you will see that 2016 is not in the range, and thus there is no corresponding y value that we can use.
On to the solution. I've broken it up into 6 parts, so that it's clearer to see why I took those steps.
Let M be the image f ( R ) . Let c = f ( 0 ) .
1) We will show that for m ∈ M , f ( m ) = 2 1 + c − m 2
Since there is a real value p such that f ( p ) = m , we can set x = m , y = p :
f ( m − m ) = f ( m ) + m 2 + m − 1 ⇒ f ( m ) = 2 1 + c − m 2 ( 1 )
2) We will show that c = 0 .
Proof by contradiction. Suppose not, then with x = x , y = 0 , f ( x ) = 0 + x × 0 + f ( x ) − 1 = f ( x ) − 1 which is a contradiction.
3) We will show that M − M = R . What this means it that given any real value r , there exists a , b ∈ M such that a − b = r .
Set x = x , y = 0 : f ( x − c ) = f ( c ) + x c + f ( x ) − 1 .
Hence f ( x − c ) − f ( x ) = x c + f ( c ) − 1
Observe that the RHS is a linear (non-constant) function in x , and thus the range is all real numbers.
4) We will show that f ( x ) = c − 2 r 2 for all real x .
Now, given any r ∈ R , we have a , b ∈ M such that r = a − b , with a = f ( p ) , b = f ( q ) .
Set x = a , y = q , we get f ( a − b ) = f ( b ) + a b + f ( a ) − 1 .
Using equation 1, we get f ( r ) = 2 1 + c − b 2 + a b + 2 1 + c − a 2 − 1 = c − 2 r 2 ( 4 )
5) We will find the value of c .
In equation (1), set m = c to get f ( c ) = 2 1 + c − c 2 .
In equation (4), set r = c to get f ( c ) = c − 2 c 2 .
Equating these 2 expressions, c = 1 .
6) In conclusion, f ( x ) = 1 − 2 x 2
Now, it remains to show that this satisfies the original functional equation, which can be manually done.