Integer roots

Algebra Level 3

Let x 3 + a x 2 + b x + c x^3+ax^2+bx+c be a polynomial with positive integer coefficients which are in arithmetic progression in that order.

Determine the maximum possible number of integer roots of the polynomial.

0 1 2 3

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.

2 solutions

Pi Han Goh
Jun 24, 2017

Since the positive integers 1 , a , b , c 1,a,b,c follows an arithmetic progression , then a 1 = b a = c b a-1 = b-a = c-b . Solving this equation in terms of a a gives b = 2 a 1 , c = 3 a 2 b = 2a-1, c = 3a-2 .

We want to find the maximum number of integer roots (in x x ) of 0 = x 3 + a x 2 + ( 2 a 1 ) x + ( 3 a 2 ) = a ( x 2 + 2 x + 3 ) + ( x 3 x 2 ) . 0 =x^3 + ax^2 + (2a-1)x + (3a-2) = a(x^2 + 2x + 3) + (x^3 - x- 2) .

Since a a is an integer, then x 3 x 2 x 2 + 2 x + 3 = 4 x 2 + 2 x + 3 + ( x 2 ) \dfrac{ x^3 - x-2}{x^2 + 2x+3} = \dfrac4{x^2 + 2x+3} + (x-2) is also integer.

Thus, x 2 + 2 x + 3 = ( x + 1 ) 2 + 2 x^2 + 2x + 3 = (x+1)^2 + 2 must divides 4. So, the possible values of ( x + 1 ) 2 + 2 (x+1)^2 + 2 are ± 1 , ± 2 , ± 4 \pm 1, \pm 2, \pm 4 . Trial and error shows that x = 1 x = -1 is the only solution (with a = b = c = 1 a=b=c=1 ). Our answer is 1 \boxed1 .

Umm, I think it should be a 1 = b a = c b a-1=b-a=c-b since 1 , a , b , c 1,a,b,c are in AP.

Skanda Prasad - 3 years, 11 months ago

Log in to reply

Whoops. Fixed.Thanks.

Pi Han Goh - 3 years, 11 months ago

Wow, that conversion makes the answer immediate :)

Calvin Lin Staff - 3 years, 11 months ago
Sándor Daróczi
Jun 22, 2017

The common difference of the coefficients is a 1 a-1 , so we have b = 2 a 1 b=2a-1 and c = 3 a 2 c=3a-2 .

First we are going to show that not all of the roots are integers. Suppose on the contrary the opposite. Clearly the polynomial does not have nonnegative roots (since we would get x 3 + a x 2 + b x + c c > 0 x^3+ax^2+bx+c \geq c > 0 ), so the roots are x 1 -x_1 , x 2 -x_2 and x 3 -x_3 where x 1 x_1 , x 2 x_2 and x 3 x_3 are positive integers.

By Vieta's formulas we obtain the following equations:

x 1 + x 2 + x 3 = a x_1 + x_2 + x_3 = a

x 1 x 2 + x 1 x 3 + x 2 x 3 = 2 a 1 x_1x_2 + x_1x_3 + x_2x_3 = 2a-1

x 1 x 2 x 3 = 3 a 2 x_1x_2x_3 = 3a-2

From the last equation it can be seen that the product of the roots gives a remainder of 1 divided by 3, which is only possible if all of the roots are 1 mod 3, or two of them are 2 mod 3 and the third is 1 mod 3. This yields us two cases.

  1. x 1 , x 2 , x 3 1 ( m o d 3 ) x_1, x_2, x_3 \equiv 1 \pmod{3} In this case we would have x 1 + x 2 + x 3 = a x_1 + x_2 + x_3 = a and x 1 x 2 + x 1 x 3 + x 2 x 3 = 2 a 1 x_1x_2 + x_1x_3 + x_2x_3 = 2a-1 to be divisible by 3 which is a contradiction.

  2. x 1 1 ( m o d 3 ) x_1 \equiv 1 \pmod{3} and x 2 , x 3 2 ( m o d 3 ) x_2, x_3 \equiv 2 \pmod{3} From this we would obtain x 1 x 2 + x 1 x 3 + x 2 x 3 = 2 a 1 x_1x_2 + x_1x_3 + x_2x_3 = 2a-1 to give a remainder of 2 + 2 + 4 2 ( m o d 3 ) 2 + 2 + 4 \equiv 2 \pmod{3} which implies 3 a 3|a , but at the same time a = x 1 + x 2 + x 3 2 ( m o d 3 ) a = x_1 + x_2 + x_3 \equiv 2 \pmod{3} , so this is also a contradiction. Thus the answer to the question can not be 3.

The assumption that exactly 2 of the roots are integers would lead us to the fact that a = x 1 + x 2 + x 3 a = x_1 + x_2 + x_3 is not an integer, yielding a contradiction. This means that the answer is at most 1.

One can easily see that 1 is possible, for example considering the equation below with root -1:

x 3 + x 2 + x + 1 = 0 x^3+x^2+x+1=0

Thus the answer is 1.

If anybody will have come up with a simpler solution, let me know.

Sándor Daróczi - 3 years, 11 months ago

Why common difference is a-1 ? Also in the last equation the every coefficient is 1 So, they are not in A.P.

Kushal Bose - 3 years, 11 months ago

Log in to reply

The first coefficient is 1, so the common difference is expressible as the difference of the second and first coefficient which is a-1.

A.P. is defined if the common difference is 0, so I don't see a problem with the example.

Sándor Daróczi - 3 years, 11 months ago

Log in to reply

Common difference zero of an A.P. does not make any sense.So, this violates statement of ur problem

Kushal Bose - 3 years, 11 months ago

Log in to reply

@Kushal Bose Why do you think it does not make sense? It satisfies the property of being an A.P. since the difference of 2 consecutive terms always remains constant.

Sándor Daróczi - 3 years, 11 months ago

Log in to reply

@Sándor Daróczi Nope, AP is created only when there is non-zero common difference.

Kushal Bose - 3 years, 11 months ago

Log in to reply

@Kushal Bose I've always known AP to be defined if the c.d. is 0, and this is more logical to me than your point of view.

To be sure about the validity of my conception I've browsed on the internet in the topic and found quite controversial opinions. Sadly none of the most popular sites (such as Wikipedia) have pointed out the d=0 case (thus not accepting or denying the possibility), however, in other sources my point seems to be still more endorsed than yours.

https://www.quora.com/If-the-common-difference-is-0-will-the-series-be-an-AP

Sándor Daróczi - 3 years, 11 months ago

Log in to reply

@Sándor Daróczi Ok let's see what @Calvin Lin sir says about this

Kushal Bose - 3 years, 11 months ago

Log in to reply

@Kushal Bose For this problem, since the only integer solution is x = 1 , d = 0 x = -1, d = 0 , it would be preferable to explicitly allow for d = 0 d = 0 in the problem statement. (Had there been other integer solutions, this consideration would not be relevant.)


In general, the answer would depend on how you (or your teacher) define an AP. Personally, I consider d = 0 d= 0 valid, especially since there isn't any advantage in understanding of AP if we avoid the case of d = 0 d = 0 .

Note: For geometric progressions, I would disallow r = 0 r = 0 , since given the second term you cannot determine the first term.

Calvin Lin Staff - 3 years, 11 months ago

Log in to reply

@Calvin Lin Ok then its fine

Kushal Bose - 3 years, 11 months ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...