Given a function f : N → N which is strictly increasing and satisfies f ( f ( x ) ) = 3 x for all x , find f ( 1 6 ) .
Notation : N denotes the set of positive integers .
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.
Why can't f ( x ) = 3 x ?
Log in to reply
Natural numbers...
As you have said it is strictly increasing function so it is a bijection between natural numbers Then what is pre-image of 1 ???
Log in to reply
A strictly increasing function does not necessarily imply a bijection.
Log in to reply
Can you give a proof for that ???
Log in to reply
@Kushal Bose – Certainly. A counterexample will do just fine.
Let f ( x ) : N → N which satisfies f ( x ) = 2 x . Then, no natural number maps to 3 , which is also a natural number.
@Kushal Bose – There can be discontinuities, in which case it can still be increasing, but not bijection. Since the domain and range are N, it's discontinuous
Go Pinoy! :)
There is another solution. f(odd) = 2 times odd and f(even)= 1.5 times even Then f(1) = 2 and f(2) = 3, so f(f(1))= 3 times 1. Next f(2) = 3 and f(3) = 6 so f(f(2)) = 3 times 2. And so on. Therefore f(16) = 24.
Log in to reply
Good idea, but doesn't work with multiple of 4
Relevant wiki: Induction - Functional Equations
We will prove via induction on all non-negative integers n a claim P n that:
First, we prove that f ( 1 ) = 2 .
If f ( 1 ) = 1 , we have f ( f ( 1 ) ) = 1 = 3 , a contradiction, hence f ( 1 ) > 1 and with the condition f ( n + 1 ) > f ( n ) , we have f ( n ) > n for all positive integers n .
If f ( 1 ) ≥ 3 , we'd have f ( f ( 1 ) ) ≥ f ( 3 ) > 3 , also a contradiction. Hence we must have f ( 1 ) = 2 . Also, we have f ( 2 ) = f ( f ( 1 ) ) = 3 and f ( 3 ) = f ( f ( 2 ) ) = 6 .
We note that we have our claim P n be true for n = 0 , given our values of f ( 1 ) , f ( 2 ) , f ( 3 ) . Now assume for some non-negative integer k , we have P k be true, i.e. that:
We then consider the claim P k + 1 :
For 3 k + 1 ≤ x ≤ 2 ⋅ 3 k + 1 , consider the x where x = 3 y for some integer 3 k ≤ y ≤ 2 ⋅ 3 k . We then have x = f ( f ( y ) ) = f ( y + 3 k ) , and consequently f ( x ) = f ( f ( y + 3 k ) ) ) = 3 ( y + 3 k ) = 3 y + 3 k + 1 = x + 3 k + 1 .
For 3 k + 1 ≤ x ≤ 2 ⋅ 3 k + 1 , there also exist x , x + 1 where 3 y < x < x + 1 < 3 ( y + 1 ) for some ( 3 k ≤ y ≤ 2 ⋅ 3 k − 1 . We can infer that x = 3 y + 1 , x + 1 = 3 y + 2 . Then using our results that f ( 3 y ) = 3 y + 3 k + 1 , f ( 3 y + 3 ) = 3 y + 3 + 3 k + 1 from just now, we have:
3 y + 3 k + 1 = f ( 3 y ) < f ( x ) < f ( x + 1 ) < f ( 3 y + 3 ) = 3 y + 3 + 3 k + 1 We thus have f ( x ) = 3 y + 1 + 3 k + 1 = x + 3 k + 1 , f ( x + 1 ) = 3 y + 2 + 3 k + 1 = x + 1 + 3 k + 1 , hence we have it that for all 3 k + 1 ≤ x ≤ 2 ⋅ 3 k + 1 , we have f ( x ) = x + 3 k + 1
For 2 ⋅ 3 k + 1 ≤ x ≤ 3 k + 2 , we note that 3 k + 1 ≤ x − 3 k + 1 ≤ 2 ⋅ 3 k + 1 and from our previous result we have f ( x − 3 k + 1 ) = x − 3 k + 1 + 3 k + 1 = x . Consequently, we have f ( x ) = f ( f ( x − 3 k + 1 ) ) = 3 x − 3 k + 2 .
Thus we have proven that with P 0 being true and P k being true ⇒ P k + 1 being true, we have proven our claim that P n is true for all non-negative integers n.
Given that 3 2 ≤ 1 6 ≤ 2 ⋅ 3 2 , we apply our formula to get f ( 1 6 ) = 2 5 .
It seems very helpful here to consider stuff in base 3!
Log in to reply
Yes, it does seem to help. (Assuming you mean base 3, not base 3! = base 6) :P
AMAZING, bro. You took it waaay further than I ever imagined doing so.
Log in to reply
I simply considered the general case. I was curious, so I explored it.
Log in to reply
I generalised it slightly differently (see the badly formatted proof below).
Great job 😃 For some reason I really cant see the contradiction f(1)>= 3. Could you enlighten me? Sorry for not LaTeX'ing, on my phone and it's late 😅
Log in to reply
If f ( 1 ) = 1 , then f ( f ( 1 ) ) = f ( 1 ) = 1 , which contradicts f ( f ( 1 ) ) = 3 . Also, since f is strictly increasing, f ( n ) > n . If f ( 1 ) ≥ 3 , then f ( f ( 1 ) ) ≥ f ( 3 ) . But, since f ( n ) > n , f ( 3 ) > 3 , so f ( f ( 1 ) ) > 3 , which contradicts f ( f ( 1 ) ) = 3 .
2 questions good sir. Why does taking f(f(x) not return 3x with the solution you suggested, the motivation for which is invisible to me. Also, why can't the function simply be f(x)= (3^1/2) x
Log in to reply
@Mohamed Elsayed - we are told f(x) is a positive integer, so it can not be (3^1/2)x. As for the motivation, I imagine the solution was found in a similar way as was done in the first solution, by Manuel Kahayon, and then proved by induction. It's a rigorous proof, but the procedure for finding the solution has not been shown.
Log in to reply
@Anthony Cutler So sorry to have wasted your time I didn't notice that last problem statement at all. I'm still confused as to why f(f(x)) does not return 3x.
Log in to reply
@Mohamed Elsayed – At the beginning of the proof there is a link explaining induction proofs. As you can see, they are indirect proofs. A direct proof that f(f(x))=3x would be complicated as f(x) has different definitions depending on the size of x. (Note that the solution does not claim that the given definition of f(x) is the only possible one; instead it only shows that the given definition satisfies the question.)
The given equation yields :
∀ n ∈ N , f ( f ( f ( n ) ) ) = { f 2 ( f ( n ) ) = 3 f ( n ) f ( f 2 ( n ) ) = f ( 3 n )
Thus :
∀ n ∈ N , f ( 3 n ) = 3 f ( n )
Hence, by induction :
∀ m , n ∈ N , f ( 3 n m ) = 3 n f ( m )
But it also known for sure that
∀ n ∈ N , 0 < f ( n ) < f ( f ( n ) ) = 3 n
Therefore :
0 < f ( 1 ) < 3
And as f ( 1 ) = 1 (otherwise 3 × 1 = f ( f ( 1 ) ) = f ( 1 ) = 1 ) :
f ( 1 ) = 2
Therefore :
{ f ( 2 ) = f ( f ( 1 ) ) = 3 × 1 f ( 3 ) = 3 × f ( 1 ) = 6 ⟹ f ( 5 ) > f ( 4 ) > f ( 3 ) = 6 ⟹ f ( 5 ) ≥ 8
2 4 = 3 × 8 ≤ 3 f ( 5 ) = f ( 1 5 ) < f ( 1 6 ) < f ( 1 7 ) < f ( 1 8 ) = 9 f ( 2 ) = 9 × 3 = 2 7
Hence :
f ( 1 6 ) = 2 5
f ( 0 ) doesn't exist.
Log in to reply
My bad : in France, N , which denotes the natural numbers, comprises 0 . I will correct that, thank you for your attention !
Problem Loading...
Note Loading...
Set Loading...
Relevant wiki: Functional Equations - Problem Solving
We cannot have f ( 1 ) = 1 as this implies f ( f ( 1 ) ) = 1 . If f ( 1 ) = 3 then f ( f ( 1 ) ) = 3 = f ( 3 ) , which is absurd since f is strictly increasing therefore f ( 1 ) = f ( 3 ) . Similar arguments force f ( 1 ) = 2 .
Therefore, f ( f ( 1 ) ) = f ( 2 ) = 3 ⟹ f ( f ( 2 ) ) = f ( 3 ) = 6 ⟹ f ( f ( 3 ) ) = f ( 6 ) = 9 .
Since f ( 3 ) = 6 and f ( 6 ) = 9 , this forces f ( 4 ) = 7 , f ( 5 ) = 8 .
f ( 6 ) = 9 ⟹ f ( f ( 6 ) ) = f ( 9 ) = 1 8 ⟹ f ( f ( 9 ) ) = f ( 1 8 ) = 2 7 .
Similarly, f ( 5 ) = 8 ⟹ f ( f ( 5 ) ) = f ( 8 ) = 1 5 ⟹ f ( f ( 8 ) ) = f ( 1 5 ) = 2 4 .
Since f ( 1 5 ) = 2 4 and f ( 1 8 ) = 2 7 , this forces f ( 1 6 ) = 2 5 , f ( 1 7 ) = 2 6 (Since f is increasing).
Therefore, our answer is 2 5