Suppose and are coprime monic polynomials with complex coefficients, such that divides and divides . For all such polynomials of degree , what is the largest possible value of ?
Details and assumptions
A polynomial is monic if its leading coefficient is 1. For example, the polynomial is monic but the polynomial is not.
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.
f ( x ) ∣ g ( x ) 2 − x ⇒ f ( x ) ∣ f ( x ) 2 + g ( x ) 2 − x g ( x ) ∣ f ( x ) 2 − x ⇒ g ( x ) ∣ f ( x ) 2 + g ( x ) 2 − x
Since f,g coprime, we have f g ∣ f 2 + g 2 − x . So f ( x ) 2 + g ( x ) 2 − x = a ( x ) f ( x ) g ( x ) for some function a(x).
If degree g is at least 1 and degree f = degree g, then degree if LHS = 2 deg(f) and degree of RHS = deg(a) + deg(f) + deg(g) = deg(a) + 2 deg(f). So degree of a(x)=0 and f ( x ) 2 + g ( x ) 2 − x = a f ( x ) g ( x ) for some constant a. Since f(x) and g(x) are both monic, then the coefficient of the highest power is 2 for LHS, and a for RHS. Therefore a=2. Rearranging the equation, we get [ f ( x ) − g ( x ) ] 2 = x which is not possible.
Therefore, both degrees of f and g are different.
Now when we let degree of f > degree of g ≥ 1, we have the following:
f ( x ) ∣ g ( x ) 2 − x ⇒ g ( x ) 2 − x = h ( x ) f ( x ) ⇒ h ( x ) ∣ g ( x ) 2 − x ( ∗ ) for some function h(x).
g ( x ) ∣ f ( x ) 2 − x ⇒ g ( x ) ∣ h ( x ) 2 f ( x ) 2 − x h ( x ) 2
⇒ g ( x ) ∣ ( g ( x ) 2 − x ) 2 − x h ( x ) 2
⇒ g ( x ) ∣ g ( x ) 4 − 2 g ( x ) 2 x + x 2 − x h ( x ) 2
⇒ g ( x ) ∣ x 2 − x h ( x ) 2 (since g ( x ) ∣ g ( x ) 4 − 2 g ( x ) 2 x )
⇒ g ( x ) ∣ x ( x − h ( x ) 2 )
⇒ g ( x ) ∣ ( h ( x ) 2 − x ) x
We note that since g ( x ) ∣ f ( x ) 2 − x , f ( x ) 2 = g ( x ) b ( x ) + x for some b(x). This means that if x is a root of g(x), then x is also a root of f(x). But we know this is not possible since f(x) and g(x) are coprime. Thus, g ( x ) ∣ h ( x ) 2 − x ( ∗ ∗ ) . From ( ∗ ) , ( ∗ ∗ ) , we can conclude that f ( x ) ∣ g ( x ) 2 − x , g ( x ) ∣ f ( x ) 2 − x ⇒ h ( x ) ∣ g ( x ) 2 − x , g ( x ) ∣ h ( x ) 2 − x . From g ( x ) 2 − x = h ( x ) f ( x ) , we note that the LHS and thus RHS are monic polynomials, and thus h(x) is monic. Also, if (x-t) is a common root of g(x) and h(x), g ( t ) 2 − t = h ( t ) f ( t ) ⇒ 0 − t = 0 ⇒ t = 0 . But this means x is a root of g(x), and we proved earlier otherwise. Hence, we know that g(x) and h(x) in this case are also monic and coprime. Also from g ( x ) 2 − x = h ( x ) f ( x ) , we know that 2 deg (g) = deg (f) + deg (h). Since degree of f > degree of g, then degree of g > degree of h.
This means that for every pair of functions (f,g) such that deg(f) > deg(g), we can "jump down" and find exactly one pair (g,h) such that deg(g) > deg(h). Conversely, for every pair (g,h), we can "jump up" and find exactly one pair (f,g), since g ( x ) 2 − x = h ( x ) f ( x ) ⇒ f ( x ) = h ( x ) g ( x ) 2 − x . Now we consider all functions (g,h), such that g is of degree 1 and thus h is of degree 0. The only possible h(x) = 1 since h(x) is monic. We know g ( x ) ∣ h ( x ) 2 − x = 1 − x , so g ( x ) = x − 1 . Therefore, there is only 1 such pair of (g,h) with degree 1 and 0 respectively. Also since 2 deg (g) = deg (f) + deg (h), we have deg (f) - deg (g) = deg (g) - deg (h). Since deg (g) - deg (h) = 1, we can conclude that any pair of such functions will have difference of degree 1. So we conclude that there is exactly a pair of functions (f,g) with degree n and n-1 that fits the criteria. We call the specific function of degree n f n ( x ) .
Now, we consider f n ( 4 ) . We want to prove by induction that f n ( 4 ) = 2 n + 1 for all non-negative integers n. It is true that f 0 ( 4 ) = 1 . Also we know that g ( x ) 2 − x = h ( x ) f ( x ) so we can rewrite them as f k ( x ) 2 − x = f k − 1 ( x ) f k + 1 ( x ) . So we get f k ( 4 ) 2 − x = f k − 1 ( 4 ) f k + 1 ( 4 ) , and if f n ( 4 ) = 2 n + 1 for all integers from 0 to k, we get ( 2 k + 1 ) 2 − 4 = ( 2 k − 1 ) f k + 1 ( 4 ) . Therefore, f k + 1 ( 4 ) = 2 k + 3 = 2 ( k + 1 ) + 1 , and thus f n ( 4 ) = 2 n + 1 holds true for k+1. Therefore by mathematical induction, f n ( 4 ) = 2 n + 1 for all non-negative integers n.
Now we can say that f 1 0 0 ( 4 ) = 2 0 0 + 1 = 2 0 1 .