Root, root, root your boat

Algebra Level 5

Suppose f ( x ) f(x) is a polynomial with complex coefficients of degree 2013 with no multiple roots. What are the last three digits of the smallest possible number of distinct complex roots of f ( f ( x ) ) ? f(f(x))?

Details and assumptions

Each root of f ( f ( x ) ) f(f(x)) is counted once, regardless of multiplicity.


The answer is 157.

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.

4 solutions

Patrick Corn
Nov 5, 2013

The key fact is that the number of distinct complex roots of a polynomial p p is equal to the degree of p p minus the degree of gcd ( p , p ) (p',p) . (This is because a root that appears with multiplicity k k in p p appears with multiplicity k 1 k-1 in gcd ( p , p ) (p',p) .)

So if we are to minimize the number of distinct roots of f ( f ( x ) ) f(f(x)) , we want to maximize the degree of the gcd of it and its derivative. Its derivative is f ( f ( x ) ) f ( x ) , f'(f(x))f'(x), and note that since f f has no repeated roots, f f' and f f are relatively prime. So f ( f ( x ) f'(f(x) and f ( f ( x ) ) f(f(x)) are relatively prime. So the gcd of f ( f ( x ) ) f(f(x)) and its derivative reduces to gcd ( f ( f ( x ) ) , f ( x ) ) (f(f(x)), f'(x)) .

This gcd has degree at most deg f ( x ) = 2012 f'(x) = 2012 , so the number of distinct zeroes is at least 201 3 2 2012. 2013^2-2012. To show that this lower bound is attained, we need to find an f f such that this gcd has degree 2012 2012 , that is, a polynomial f f such that f ( x ) f'(x) divides f ( f ( x ) ) f(f(x)) .

Take a a to be a complex number satisfying a 2012 = 1 a^{2012} = -1 . Then let f ( x ) = x 2013 + a f(x) = x^{2013} + a . Then an easy computation shows that f ( f ( x ) ) f(f(x)) is divisible by x 2012 x^{2012} , which is a constant multiple of f ( x ) f'(x) . Hence we are done.

Finally, 201 3 2 2012 1 3 2 12 157 2013^2 - 2012 \equiv 13^2 - 12 \equiv \fbox{157} mod 1000 1000 .

How do you get the fact that the number of distinct complex roots of a polynomial p p is equal to deg { p } deg { gcd ( p , p ) } \textrm{deg}\{p\}-\textrm{deg}\{\gcd(p',p)\} ? Thanks.

R Y - 7 years, 7 months ago
Mark Hennings
Nov 3, 2013

If the distinct roots of f ( x ) = 0 f(x) = 0 are λ 1 , , λ 2013 \lambda_1,\ldots,\lambda_{2013} , then the set of roots of f ( f ( x ) ) = 0 f(f(x))=0 is the set R = j = 1 2013 R j = j = 1 2013 { x C f ( x ) = λ j } R \; = \; \bigcup_{j=1}^{2013} R_j \; = \; \bigcup_{j=1}^{2013}\big\{x \in \mathbb{C} \big| f(x) = \lambda_j\big\} Certainly R i R j = R_i \cap R_j = \varnothing for i j i \neq j , and so R = j = 1 2013 R j |R| \; = \; \sum_{j=1}^{2013} |R_j| .

If μ R j \mu \in R_j is a repeated root of the equation f ( x ) = λ j f(x) = \lambda_j , then f ( x ) λ j = ( x μ ) a g ( x ) f(x) - \lambda_j = (x-\mu)^ag(x) for some a 2 a \ge 2 and some polynomial g ( x ) g(x) . But then f ( x ) = ( x μ ) a 1 h ( x ) f'(x) = (x-\mu)^{a-1}h(x) for some polynomial h ( x ) h(x) , and hence μ \mu is an ( a 1 ) (a-1) -fold repeated root of the polynomial f ( x ) f'(x) , which has degree 2012 2012 . The effect of μ \mu is to reduce the size of R j |R_j| by a 1 a-1 . Thus R |R| is equal to 2013 × 2013 2013\times2013 minus a reduction due to the effects of repeated roots in the sets R j R_j . Each such repeated root contributes to this reduction by an amount equal to its multiplicity as a zero of f ( x ) f'(x) . Thus the greatest possible theoretical reduction to R |R| that is due to repeated roots is 2012 2012 , the degree of f ( x ) f'(x) . That would make the least theoretically possible size of R |R| to be 2013 × 2013 2012 = 4050157 2013\times2013 - 2012 \; = \; 4050157 It remains to show that this least theoretically possible size is achievable. If we define λ = e i π 2012 \lambda = e^{\frac{i\pi}{2012}} , so that λ 2012 = 1 \lambda^{2012} = -1 , consider f ( x ) = x 2013 + λ f(x) = x^{2013} + \lambda . Certainly the zeros of f ( x ) f(x) are the 2013 2013 distinct 2013 2013 th roots of λ -\lambda . Since λ 2013 + λ = λ ( λ 2012 + 1 ) = 0 \lambda^{2013}+\lambda = \lambda(\lambda^{2012}+1) = 0 , λ \lambda is one of these. But 0 0 is a 2013 2013 -fold repeated root of f ( x ) = λ f(x) = \lambda .

In other words, it is possible to achieve this maximum possible reduction of 2012 2012 to the size of R |R| , and so the minimum possible size of R |R| is indeed 4050157 4050157 .

Peter Byers
Nov 10, 2013

If S a S_a is the solution set of f ( x ) = a f(x)=a , then the number of distinct (complex) roots of f ( f ( x ) ) f(f(x)) is a S 0 S a \sum_{a\in S_0} |S_a| . (The S a S_a sets are all distinct, since f ( x ) f(x) cannot have two different values for the same x x .)

But if x x is a double [triple, quadruple, etc.] root of f ( x ) = a f(x)=a , then it is a single [respectively double, triple, etc.] root of f ( x ) = 0 f'(x)=0 . That is to say,

a S 0 ( 2013 S a ) d e g ( f ) = 2012 \sum_{a\in S_0} (2013- |S_a|) \le deg (f') = 2012

so a S 0 S a 201 3 2 2012 \sum_{a\in S_0}|S_a| \ge 2013^2- 2012 .

On the other hand, f ( x ) = ( x + 2 ) 2013 1 f(x)=(x+2)^{2013}-1 has no multiple roots, and it yields

a S 0 S a = 2013 2012 + S 1 = 2013 2012 + 1 \sum_{a\in S_0}|S_a| = 2013*2012 + |S_{-1}| = 2013*2012 +1

which is the same as 201 3 2 2012 = 4050157 2013^2- 2012 =4050157 , and the last three digits are 157 157 .

Calvin Lin Staff
Nov 19, 2015

Suppose f ( x ) = a ( x a 1 ) . . . ( x a 2013 ) . f(x)=a(x-a_1)\cdot...\cdot(x-a_{2013}). Then the roots of f ( f ( x ) ) f(f(x)) are the roots of f ( x ) a i , f(x)-a_i, for i = 1 , 2 , . . . , 2013. i=1,2,...,2013. Since for i j i\neq j a i a j , a_i\neq a_j, so for a fixed x x f ( x ) f(x) cannot equal a i a_i and a j a_j simultaneously. Thus, none of the roots of f ( x ) a i f(x)-a_i is a root of f ( x ) a j . f(x)-a_j. The following lemma is useful for more than just this problem.

Lemma. If x 0 x_0 is a root of a polynomial g , g, then its multiplicity in g g is exactly one bigger than its multiplicity in g , g', the derivative of g . g. (if x 0 x_0 is not a root of g , g', then we say that its multiplicity in g g' is zero).

Proof. If the multiplicity of x 0 x_0 in g g is k 1 , k\geq 1, then g ( x ) = ( x x 0 ) k h ( x ) , g(x)=(x-x_0)^k\cdot h(x), where h ( x 0 ) 0. h(x_0)\neq 0. So g ( x ) = k ( x x 0 ) k 1 h ( x ) + ( x x 0 ) k h ( x ) = ( x x 0 ) k 1 ( k h ( x ) + ( x x 0 ) h ( x ) ) g'(x)=k(x-x_0)^{k-1}h(x)+(x-x_0)^{k}h'(x)=(x-x_0)^{k-1}\left( kh(x)+ (x-x_0)h'(x)\right) Note that k h ( x 0 ) + ( x 0 x 0 ) h ( x 0 ) = k h ( x 0 ) 0 , kh(x_0)+ (x_0-x_0)h'(x_0)=kh(x_0)\neq 0, so the lemma is proven.

For each i , i, the total number of roots of f ( x ) a i , f(x)-a_i, if counted with multiplicity, is 2013 , 2013, for a total of 201 3 2 2013^2 roots of f ( f ( x ) ) . f(f(x)). Since we count the roots once, we must subtract k 1 k-1 for each root x 0 x_0 of some f ( x ) a i f(x)-a_i of multiplicity k k . By the above lemma, applied to g ( x ) = f ( x ) a i , g(x)=f(x)-a_i, this number k 1 k-1 equals the multiplicity of x 0 x_0 in g ( x ) . g'(x). Note that for every i i g ( x ) = f ( x ) , g'(x)=f'(x), so the sum of the numbers ( k 1 ) (k-1) that we need to subtract, over all roots of f ( x ) a i , f(x)-a_i, is at most the degree of f ( x ) , f'(x), which equals 2012. 2012. Therefore, the total number of roots is at least 201 3 2 2012. 2013^2-2012. We will now construct an example of f ( x ) f(x) that achieves this bound.

Consider f ( x ) = x 2013 + c , f(x)=x^{2013}+c, where c c is some complex number, such that c 2012 = 1. c^{2012}=-1. Clearly, all roots of f f are simple (by the Lemma above, or directly). Then f ( f ( x ) ) = ( x 2013 + c ) 2013 + c = ( x 201 3 2 + . . . + 2013 x 2013 c 2012 + c 2013 ) + c f(f(x))=(x^{2013}+c)^{2013}+c=\left(x^{2013^2}+...+2013x^{2013}c^{2012}+c^{2013}\right) +c Note that c 2013 + c = 0 , c^{2013}+c=0, so 0 0 is a root of f ( f ( x ) ) f(f(x)) with multiplicity 2013. 2013. Therefore, the total number of roots of f ( f ( x ) ) f(f(x)) , counted without multiplicity, is at most 201 3 2 2012. 2013^2-2012.

Finally, the last three digits of 201 3 2 2012 2013^2-2012 are the same as 1 3 2 12 = 157. 13^2-12=157.

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...