Roots Revisited

Algebra Level 5

A polynomial f ( x ) f(x) of degree n n with real coefficients has roots x 1 , x 2 , , x n 1 x_1, x_2, \ldots , x_n\ne1 is such that f ( 1 ) = 10 f(1)=-10 , f ( 1 ) = 1 f'(1)=1 and f ( 1 ) = 6 f''(1)=6 .

Find i = 1 n 1 ( 1 x i ) 2 \displaystyle \sum_{i=1}^{n}\dfrac{1}{(1-x_i)^2} .

Clarification : f ( x ) f'(x) and f ( x ) f''(x) represents the first and second derivative of f ( x ) f(x) , respectively.


Inspiration .


The answer is 0.61.

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.

3 solutions

If f ( x ) = a 0 x n + a 1 x n 1 + + a n \displaystyle f(x)=a_0 x^n+a_1 x^{n-1}+\cdots +a_n has roots x 1 , x 2 , , x n x_1,x_2,\cdots,x_n then the substitution y = 1 1 x y=\frac{1}{1-x} transforms f ( x ) f(x) into a polynomial with roots 1 1 x 1 , , 1 1 x n \displaystyle \frac{1}{1-x_1},\cdots,\frac{1}{1-x_n} and thus,

f ( x ) = x n k = 0 n a k x n 1 k = 1 n ( k 1 ) a k + + 1 = 0 \displaystyle f(x)=x^n\sum_{k=0}^{n}a_k - x^{n-1}\sum_{k=1}^{n}\binom{k}{1}a_k+\cdots+1=0 and by Vieta's we have

i = 1 n 1 1 x i = k = 1 n a k ( k 1 ) k = 0 n a k = f ( 1 ) f ( 1 ) \displaystyle \sum_{i=1}^{n} \frac{1}{1-x_i}=\frac{\sum_{k=1}^{n}a_k\binom{k}{1}}{\sum_{k=0}^{n}a_k} = \frac{f'(1)}{f(1)}

Again by Newton's identities we have p 0 S 2 + p 1 S 1 + 2 p 2 = 0 \displaystyle p_0 S_2+p_1 S_1 +2p_2=0 where p i p_i are the coefficients of the transformed equation and S m S_m are the sum of m-th power of roots, Simplifying we get i = 1 n 1 ( 1 x i ) 2 = ( f ( 1 ) ) 2 f ( 1 ) f ( 1 ) ( f ( 1 ) ) 2 = 61 100 = 0.61 \displaystyle \sum_{i=1}^{n}\frac{1}{(1-x_i)^2} = \frac{(f'(1))^2-f(1)f''(1)}{(f(1))^2} = \frac{61}{100}=\boxed{0.61}

Note : \text{Note : } Every coefficient p n p_n of the transformed equation can be related to f ( x ) f(x) by p n = ( 1 ) n f ( n ) ( 1 ) n \displaystyle p_n = \frac{(-1)^n f^{(n)}(1)}{n} which we can observe easily.

Nice solution and thank you for the inspiration... However, you don't need Vieta, Newton, substitution...Do you know how I solved it? If you suppose that there is one only polynomial fulfilling these constraints, then the polynomial must be a second degree polynomial. I mean f ( x ) = a x 2 + b x + c f ( 1 ) = 10 = a + b + c , f ( 1 ) = 1 = 2 a + b , f ( 1 ) = 6 = 2 a f(x) = ax^2 + bx + c \Rightarrow f(1) = -10 = a + b + c, \space f'(1) = 1 = 2a + b, \space f''(1) = 6 = 2a \Rightarrow f ( x ) = 3 x 2 5 x 8... f(x) = 3x^2 - 5x - 8... Now, you only have to solve the equation 3 x 2 5 x 8 = 0 3x^2 - 5x - 8 = 0 Thus, x 1 = 8 3 , x 2 = 1 x_1 = \frac{8}{3}, \space x_2 = -1 The rest is simple... With his, I have solved the question and furthemore I have got the roots, even the polynomial... What do you think about my process?is it difficult? am I wrong?...

Guillermo Templado - 4 years, 9 months ago

Log in to reply

this is excellent ! you found the scratch and the info I proived was for this polynomial only. I mad the problem from this polynomial and gave it as a nth degree polynomial so to confuse. But to make sure its a second-degree polynomial we dont have sufficient logic for that.

Aditya Narayan Sharma - 4 years, 9 months ago

Log in to reply

thank you very much!... as I said I assumed that the polynomial should be unique (this doesn't mean it's right), and because of the restrictions(constraints) that you gave the only possibility is that it was a second-degree polynomial.. By the way, I'm going to continue working about this wiki . I tell you because I couldn't mention you when I was going to write Bernouilli equation. Now , I'm going to continue with variable changes for the independient variable x x including Euler's equation, and later I'll make variable changes involving the two variables x y x \wedge y . You can and you are invited to contribute, if you want... Greetings...

Guillermo Templado - 4 years, 9 months ago

Log in to reply

@Guillermo Templado I must say the wiki is going great till and I would contribute to the best of my knowledge :)

Aditya Narayan Sharma - 4 years, 9 months ago
Kushal Bose
Sep 17, 2016

Let f ( x ) = ( x x 1 ) ( x x 2 ) ( x x n ) f(x)=(x-x_1)(x-x_2) \cdots (x-x_n)

Differentiating w.r.t. x

f ( x ) = ( x x 2 ) ( x x 3 ) ( x x n ) + ( x x 1 ) ( x x 3 ) ( x x n ) + ( x x 1 ) ( x x 2 ) ( x x n 1 ) f'(x)=(x-x_2)(x-x_3) \cdots (x-x_n) + (x-x_1)(x-x_3) \cdots (x-x_n) + (x-x_1)(x-x_2) \cdots (x-x_{n-1})

f ( x ) = f ( x ) i = 0 n 1 x x i f'(x)=f(x) \sum_{i=0 }^{ n } \frac{1}{x-x_i}

Putting x = 1 x=1 we get: i = 0 n 1 x x i = f ( 1 ) f ( 1 ) = 1 / 10 \sum_{i=0 }^{ n } \frac{1}{x-x_i}=\frac{f'(1)}{f(1)}=-1/10

Now again differentiate w.r.t. x :

f ( x ) = i = 0 n f ( x ) ( x x i ) f ( x ) ( x x i ) 2 f''(x)=\sum_{i=0}^{n} \frac{f'(x) (x-x_i)- f(x)}{(x-x_i)^2}

f ( x ) = f ( x ) i = 0 n 1 x x i f ( x ) i = 0 n 1 ( x x i ) 2 ) f''(x)=f'(x) \sum_{i=0}^{n} \frac{1}{x-x_i} -f(x) \sum_{i=0}^{n} \frac{1}{(x-x_i)^2)}

i = 0 n 1 ( x x i ) 2 \sum_{i=0}^{n} \frac{1}{(x-x_i)^2} = f ( x ) i = 0 n 1 x x i f ( x ) f ( x ) \dfrac{f'(x) \sum_{i=0}^{n} \frac{1}{x-x_i}-f''(x)}{f(x)}

Putting x=1 in this equation:

i = 0 n 1 ( 1 x i ) 2 \sum_{i=0}^{n} \frac{1}{(1-x_i)^2} = f ( 1 ) ( 1 / 10 ) f ( 1 ) ) / f ( 1 ) = 0.61 f'(1)(-1/10)-f''(1))/f(1)=\boxed{0.61}

When you went to find f ( x ) f'' (x) , the second line is actually just the product rule. So, you could have skipped listing out the quotient rule.

That's a nice thing about the quotient rule, which is that it is the product rule in disguise.

Calvin Lin Staff - 4 years, 8 months ago

Given x 1 , x 2 , x n x_1, x_2, \cdots x_n are roots.

So, f ( x ) = k 1 n ( x x i ) f(x) = k\prod_1^n (x-x_i)

Differentiating once, f ( x ) = 1 n f ( x ) x x i = f ( x ) 1 n 1 x x i f'(x) = \sum_1^n \dfrac{f(x)}{x-x_i} =f(x)\sum_1^n \dfrac1{x-x_i}

So, 1 n 1 x x i = f ( x ) f ( x ) \sum_1^n \dfrac1{x-x_i} = \dfrac{f'(x)}{f(x)}

Differentiating again, f ( x ) = 1 n [ f ( x ) x x i f ( x ) ( x x i ) 2 ] = f ( x ) f ( x ) f ( x ) f ( x ) 1 n 1 ( x x i ) 2 1 n 1 ( x x i ) 2 = ( f ( x ) ) 2 f ( x ) f ( x ) ( f ( x ) ) 2 = 0.61 \begin{aligned}f''(x) &= \sum_1^n \left[\dfrac{f'(x)}{x-x_i} - \dfrac{f(x)}{(x-x_i)^2}\right]\\ &= f'(x)\cdot \dfrac{f'(x)}{f(x)} - f(x)\sum_1^n \dfrac1{(x-x_i)^2}\end{aligned}\\ \Rightarrow \sum_1^n \dfrac1{(x-x_i)^2} = \dfrac{(f'(x))^2 - f''(x)f(x)}{(f(x))^2}=\boxed{0.61}

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...