For n = 2 k , where k is a positive integer, we have the identity x n + y n ≡ ( x + y + c 1 x y ) ( x + y + c 2 x y ) ⋯ ( x + y + c m x y ) , where c 1 , c 2 , … , c m ∈ R .
If we define f ( k ) = 1 ⩽ i ⩽ n max c i , what is the value of k → ∞ lim f ( k ) ?
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.
This problem could be generalised to any positive integer n by
x n + y n = k = 1 ∏ n ( x + y + 2 cos ( 2 n ( 2 k − 1 ) π ) x y )
Currently I have a proof using very ugly collection of terms, and linking this problem to Chebychev polynomials of first kind. I'll post the details if I find a better proof.
This problem can be generalised to any positive integer n by
k = 1 ∏ n ( x + y + 2 cos ( 2 n ( 2 k − 1 ) π ) x y ) = x n + y n
To find that, notice that for a fixed y , x n + y n = 0 must have the same roots as ( x + y + c 1 x y ) ( x + y + c 2 x y ) … ( x + y + c n x y ) = k = 1 ∏ n ( x + y + c k x y ) = 0
Now, x n + y n = 0 has roots x = y e i ( 2 k − 1 ) π / n for 1 ≤ k ≤ n . Since the product above has exactly n terms, which is equal to the number of roots for x n + y n = 0 , we can conclude that for every 1 ≤ k ≤ n , there must exist a term in the product that equates to 0 when x = y e i ( 2 k − 1 ) π / n . Substituting the roots into the product yields:
k = 1 ∏ n y e i ( 2 k − 1 ) π / n + y + c k y e i ( 2 k − 1 ) π / n y = k = 1 ∏ n y ( e i ( 2 k − 1 ) π / n + 1 + c k e i ( 2 k − 1 ) π / 2 n ) = 0 , [ 1 ≤ k ≤ n ]
From here we can solve for c k :
e i ( 2 k − 1 ) π / n + 1 + c k e i ( 2 k − 1 ) π / 2 n = 0
c k = − 2 cos ( 2 n ( 2 k − 1 ) π ) , 1 ≤ k ≤ n
The final answer of 2 follows easily from here by considering k = n
Problem Loading...
Note Loading...
Set Loading...
Given a 2 + b 2 + c a b for some real number c ∈ ( − 2 , 2 ) , we will factor it to ( a + b + p a b ) ( a + b + q a b ) for some real numbers p , q . We will determine what p , q should be.
If we multiply out the factorization, we obtain
( a + b + p a b ) ( a + b + q a b ) = ( a + b ) 2 + ( p + q ) ⋅ ( a + b ) a b + p q ⋅ a b = a 2 + 2 a b + b 2 + ( p + q ) ⋅ ( a + b ) a b + p q ⋅ a b = a 2 + b 2 + ( 2 + p q ) ⋅ a b + ( p + q ) ⋅ ( a + b ) a b
Thus, if we want this to be equal to a 2 + b 2 + c a b , we need 2 + p q = c and p + q = 0 . The latter means q = − p ; substituting to the former gives 2 − p 2 = c or p = ± 2 − c . Without loss of generality, let p be the positive value 2 − c , so q gets the negative value − 2 − c .
First, notice that since we assume c < 2 , the square root is defined, so p , q are defined. Next, since we also assume c > − 2 , we have p < 2 (and thus q > − 2 ). Also, clearly 0 < p and 0 > q . Thus p , q ∈ ( − 2 , 2 ) .
The important point is that if we have a factor x + y + c x y , and we substitute x = a 2 , y = b 2 and factorize it further to ( a + b + p a b ) ( a + b + q a b , the following fact holds: if c ∈ ( − 2 , 2 ) , then p , q ∈ ( − 2 , 2 ) as well. Thus if we factorize each of its factors and so on, all coefficients will stay being real numbers.
Now we're ready for the real problem. However, before computing f ( k ) , we will first compute g ( k ) = min 1 ≤ i ≤ 2 k c i . We claim that g ( 1 ) = − 2 and g ( k + 1 ) = − 2 − g ( k ) . The first claim is easy; just plug in c = 0 to the formula q = − 2 − c above to get q = − 2 . Since p > 0 , we have g ( 1 ) = q = − 2 .
To prove the recurrence, consider x 2 k + 1 + y 2 k + 1 . Substitute t = x 2 , u = y 2 , so our expression becomes t 2 k + u 2 k . We can factorize this to have the factor t + u + g ( k ) t u = x 2 + y 2 + g ( k ) x y . Plugging this to the above formula, we obtain that this can be further factorized into ( x + y + 2 − g ( k ) ⋅ x y ) ( x + y − 2 − g ( k ) ⋅ x y ) , so we see g ( k + 1 ) ≤ − 2 − g ( k ) .
To see that it is indeed the minimum, consider any factor t + u + c t u . By definition, c ≥ g ( k ) . Thus if we factorize x 2 + y 2 + c x y = ( x + y + p x y ) ( x + y + q x y ) , we obtain q = − 2 − c ≥ − 2 − g ( k ) . (And clearly p > 0 .) So in fact g ( k + 1 ) = − 2 − g ( k ) ; all factors of x 2 k + 1 + y 2 k + 1 have those coefficients at least − 2 − g ( k ) .
Finally, we will now go to f ( k ) . A very similar argument as above gives that f ( k + 1 ) = 2 − g ( k ) : this is caused from taking the positive coefficient of factoring x 2 + y 2 + g ( k ) x y , and any other factor x 2 + y 2 + c x y will give p = 2 − c ≤ 2 − g ( k ) . This also shows that f ( k ) = − g ( k ) , so we obtain the recurrence f ( k + 1 ) = 2 + f ( k ) (with starting value f ( 1 ) = 2 ).
Finally, we're looking for the limit, so it will be in the well-known form 2 + 2 + 2 + … = 2 .