Sum it up to 1729

Algebra Level 3

If V n = a n + b n , V_{n}=a^{n}+b^{n}, where a a and b b are the roots of x 2 + x + 1 , x^{2}+x+1, what is the value of

n = 0 1729 ( 1 ) n V n ? \sum_{n=0}^{1729} (-1)^{n} \cdot \ V_{n} ?


The answer is 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.

7 solutions

Ayush Verma
Nov 20, 2014

a = ω & b = ω 2 , w h e r e ω & ω 2 a r e c o m p l e x r o o t o f x 3 = 1 ( 1 ) n V n = ( ω ) n + ( ω 2 ) n n = 0 1729 ( 1 ) n V n = n = 0 1729 ( ω ) n + n = 0 1729 ( ω 2 ) n = 1 ( ω ) 1730 1 ( ω ) + 1 ( ω 2 ) 1730 1 ( ω 2 ) = 1 ω 1730 1 + ω + 1 ω 3460 1 + ω 2 = 1 ω 2 1 + ω + 1 ω 1 + ω 2 ( a s ω 3 = 1 ) = 1 ω 4 + 1 ω 2 1 + ω + ω 2 + ω 3 = 2 ω 2 ω 0 + 1 ( a s 1 + ω + ω 2 = 0 & ω 3 = 1 ) = 3 ( a s 1 + ω + ω 2 = 0 ) a=\omega \quad \& \quad b={ \omega }^{ 2 }\quad ,where\\ \\ \omega \quad \& \quad \quad { \omega }^{ 2 }\quad are\quad complex\quad root\quad of\quad { x }^{ 3 }=1\\ \\ { \left( -1 \right) }^{ n }{ V }_{ n }={ \left( -\omega \right) }^{ n }{ +\left( { -\omega }^{ 2 } \right) }^{ n }\\ \\ \Rightarrow \sum _{ n=0 }^{ 1729 }{ { \left( -1 \right) }^{ n }{ V }_{ n }= } \sum _{ n=0 }^{ 1729 }{ { \left( -\omega \right) }^{ n } } +\sum _{ n=0 }^{ 1729 }{ { \left( { -\omega }^{ 2 } \right) }^{ n } } \\ \\ \quad \quad \quad =\cfrac { { 1-\left( -\omega \right) }^{ 1730 } }{ 1-\left( -\omega \right) } +\cfrac { { 1-\left( -{ \omega }^{ 2 } \right) }^{ 1730 } }{ 1-\left( -{ \omega }^{ 2 } \right) } \\ \\ \quad \quad =\cfrac { 1-{ \omega }^{ 1730 } }{ 1+\omega } +\cfrac { 1-{ \omega }^{ 3460 } }{ 1+{ \omega }^{ 2 } } \\ \\ \quad \quad =\cfrac { 1-{ \omega }^{ 2 } }{ 1+\omega } +\cfrac { 1-\omega }{ 1+{ \omega }^{ 2 } } \quad \left( as\quad { \omega }^{ 3 }=1 \right) \\ \\ \quad =\cfrac { 1-{ \omega }^{ 4 }+1-{ \omega }^{ 2 } }{ 1+\omega +{ \omega }^{ 2 }+{ \omega }^{ 3 } } \\ \\ \quad =\cfrac { 2-{ \omega }^{ 2 }-\omega }{ 0+1` } \quad \left( as\quad 1+\omega +{ \omega }^{ 2 }=0\quad \& \quad { \omega }^{ 3 }=1 \right) \\ \\ \quad =3\quad \left( as\quad 1+\omega +{ \omega }^{ 2 }=0\quad \right) \\ \\ \quad

Exactly , good problem

U Z - 6 years, 5 months ago

Have I missed something here because surely the sum: Sn = ( 1 ) n (-1)^{n} = 1-1+1-1+1... would produce 0 if n is odd?

Curtis Clement - 6 years, 5 months ago

Log in to reply

Actually, the repeating sextet is: 2, 1,-1, -2, -1, 1. All of them end at n=1727 and give zero total. So, the resulting sum is equal to the sum of the elements 1728 and 1729, or two first elements in sextet: 2+1 =3.

M Dub - 5 years, 9 months ago

Looks like you have missed the power 0 term which will reduce to a value 2 and moving forward your odd and even terms will be changed and reduce to 1. Final answer would be 2+1=3.

Dhruva Ahuja - 1 year, 7 months ago
Chew-Seong Cheong
Nov 20, 2014

The roots of x 2 + x + 1 = 0 x^2+x+1 = 0 , ( a , b ) = 1 2 ± i 3 2 \quad (a,b) = - \frac {1}{2} \pm i \frac {\sqrt{3}}{2} , where i = 1 i = \sqrt{-1} .

Using the Euler's identity, we have: ( a , b ) = 1 2 ± i 3 2 = cos 2 π 3 ± sin 2 π 3 = e ± i 2 π 3 (a,b) = - \frac {1}{2} \pm i \frac {\sqrt{3}}{2} = \cos {\frac {2\pi}{3}} \pm \sin {\frac {2\pi}{3}} = e^{\pm i \frac{2\pi}{3}} .

V n = a n + b n = e i 2 n π 3 + e i 2 n π 3 = 2 cos 2 n π 3 \Rightarrow V_n = a^n + b^n = e^{i \frac{2n\pi}{3}} + e^{-i \frac{2n\pi}{3}} = 2\cos {\frac{2n\pi}{3}}

V n = { + 2 n 0 ( m o d 3 ) 1 n 1 ( m o d 3 ) 1 n 2 ( m o d 3 ) \Rightarrow V_n = \begin{cases} +2 \quad & n \equiv 0 \pmod {3} \\ -1 \quad & n \equiv 1 \pmod {3} \\ -1 \quad & n \equiv 2 \pmod {3} \end{cases}

Similarly,

n 0 N ( 1 ) n V n = { 2 N 0 ( m o d 6 ) 3 N 1 ( m o d 6 ) 2 N 2 ( m o d 6 ) 0 N 3 ( m o d 6 ) 1 N 4 ( m o d 6 ) 0 N 5 ( m o d 6 ) \sum _{n-0} ^N {(-1)^n V_n} = \begin{cases} 2 \quad & N \equiv 0 \pmod {6} \\ 3 \quad & N \equiv 1 \pmod {6} \\ 2 \quad & N \equiv 2 \pmod {6} \\ 0 \quad & N \equiv 3 \pmod {6} \\ -1 \quad & N \equiv 4 \pmod {6} \\ 0 \quad & N \equiv 5 \pmod {6} \end{cases}

Since 1729 1 ( m o d 6 ) 1729 \equiv 1 \pmod {6} , the answer is 3 \boxed {3} .

Nice Approach.

Shubhendra Singh - 6 years, 6 months ago

Log in to reply

Thanks. I find Euler's identity the clearest approach in terms of reasoning.

Chew-Seong Cheong - 6 years, 6 months ago

Log in to reply

SIR, how could i develop maths like you.

Dheeraj Agarwal - 6 years, 6 months ago

Log in to reply

@Dheeraj Agarwal You can practice on Brilliant.org for a start.

Chew-Seong Cheong - 6 years, 6 months ago

Very Nice.

Ayush Verma - 6 years, 6 months ago

I did both solutions given. This one is very elegant.

Jake Lai - 6 years, 6 months ago

Same process bro, but i have done with omega(w).

Jyothi Vanka - 6 years, 6 months ago

Nice solution. I did practically the same way. Used De Moivre's Formula to get Vn and then noticed that (-1)^n * Vn sextets give zero in total. So the result is the sum of elements 1728 and 1729 or first two elements, which is 2+1=3. Thank you.

M Dub - 5 years, 9 months ago
Abhishek Sinha
Sep 10, 2015

Solution without using complex numbers : First note that V n + 1 = a n + 1 + b n + 1 = ( a + b ) ( a n + b n ) a b ( a n 1 + b n 1 ) = V n V n 1 V_{n+1}=a^{n+1}+b^{n+1}=(a+b)(a^n+b^n)-ab(a^{n-1}+b^{n-1})=-V_{n}-V_{n-1} i.e., V n + 1 + V n + V n 1 = 0 V_{n+1}+V_n+V_{n-1}=0 Writing the above equality for n n + 1 n\gets n+1 , we have V n + 2 + V n + 1 + V n = 0 V_{n+2}+V_{n+1}+V_n=0 Comparing the above two equations, we have V n 1 = V n + 2 V_{n-1}=V_{n+2} , i.e. V n = V n + 3 V_n=V_{n+3} Next we form groups of six consecutive terms in the summation n = 0 1729 ( 1 ) n V n \sum_{n=0}^{1729}(-1)^nV_n and utilize the above observation to realize that the sum of each group is zero, e.g. V 0 V 1 + V 2 V 3 + V 4 V 5 = 0. V_0-V_1+V_2-V_3+V_4-V_5=0. Hence the value of the given sum is last two left-over terms, viz n = 0 1729 ( 1 ) n V n = V 1728 V 1729 = V 0 V 1 = 2 ( 1 ) = 3 \sum_{n=0}^{1729}(-1)^nV_n=V_{1728}-V_{1729}=V_0-V_1=2-(-1)=3 .

Rudresh Tomar
Nov 21, 2014

I used de moivre's theorem !

Can you provide more details about what you did?

Calvin Lin Staff - 6 years, 6 months ago

Same here bro

Rindell Mabunga - 6 years, 6 months ago
Harry Ray
May 5, 2016

Here is a solution using Newton's identities. One form of the general identities is: p k = i = k n k 1 ( 1 ) k 1 + i e k i p i p_k = \sum_{i=k-n}^{k-1} (-1)^{k-1+i} e_{k-i} p_i where p i ( a , b ) = a i + b i p_i(a,b) = a^i + b^i and e i e_i is the ith elementary symmetric polynomial. This gives the following sequence of equations: p 1 = e 1 p 2 = e 1 p 1 2 e 2 p 3 = e 1 p 2 e 2 p 1 + 3 e 3 p_1 = e_1 \\ p_2 = e_1 p_1 - 2 e_2 \\ p_3 = e_1 p_2 - e_2 p_1 + 3 e_3 \\ \vdots Since we are dealing with the roots of a quadratic, by definition e i = 0 e_i = 0 for i > 2 i > 2 . This then simplifies to: p k = e 1 p k 1 e 2 p k 2 p_k = e_1 p_{k-1} - e_2 p_{k-2} for k 3 k \ge 3 . This gives us a recurrence for calculating values of p k p_k , but we first need to determine the values of e 1 e_1 and e 2 e_2 . Since they are symmetric polynomials of the roots of a polynomial we can use Vieta's formulas. Since a , b a, b are the roots of x 2 + x + 1 x^2 + x + 1 we have e 1 = p 1 = 1 e_1 = p_1 = -1 and e 2 = 1 e_2 = 1 . This gives us a complete recurrence for p k p_k : p k = p k 1 p k 2 , k 2 p 0 = 2 , p 1 = 1 p_k = -p_{k-1} - p_{k-2}, k \ge 2 \\ p_0 = 2, p_1 = -1 This is a second order homogeneous linear recurrence relation with constant coefficients and so a general solution can be found, but for the purposes of this problem it is enough to note that p 3 = p 0 , p 4 = p 1 , p 5 = p 2 = 1 p_3 = p_0, p_4 = p_1, p_5 = p_2 = -1 and since each term relies only on the previous two, therefore the sequence must be periodic with period 3. That is, the sequence of p k p_k satisfies: p k = { 2 k 0 ( m o d 3 ) 1 o t h e r w i s e p_k = \begin{cases} 2 & k \equiv 0 \pmod 3 \\ -1 & \mathrm{otherwise} \end{cases} Clearly, from our definition p i = V i p_i = V_i , so the value we are interested in is: n = 0 1729 ( 1 ) n p n \sum_{n=0}^{1729} (-1)^n p_n Since this sequence has period 3 (and hence also has period 6), n = a a + 5 ( 1 ) n p n = 0 \sum_{n=a}^{a+5} (-1)^n p_n = 0 for any a a , and 1729 1 ( m o d 6 ) 1729 \equiv 1 \pmod 6 , this becomes: n = 0 1729 ( 1 ) n p n = 288 n = 0 5 ( 1 ) n p n + p 1728 p 1729 = 0 + 2 + 1 = 3 \sum_{n=0}^{1729} (-1)^n p_n = 288 \sum_{n=0}^5 (-1)^n p_n + p^{1728} - p^{1729} \\ = 0 + 2 + 1 = 3 Therefore, the answer is 3 \boxed{3} .

Thanks for sharing your approach, Harry!

This solution doesn't require finding the roots of the given polynomial, and this method can be extended for polynomials of higher degree.

Pranshu Gaba - 5 years, 1 month ago
Carsten Meyer
Feb 18, 2019

Let N : = 1730 N:=1730 as a short-hand. You can do most of the job using the geometric sum:

S : = n = 0 N 1 ( 1 ) n V n = n = 0 N 1 ( a ) n + ( b ) n = 1 ( a ) N 1 + a + 1 ( b ) N 1 + b = 1 ( a ) R 1 + a + 1 ( b ) R 1 + b , N R m o d 6 \begin{aligned} S:=\sum_{n=0}^{N-1}(-1)^nV_n&=\sum_{n=0}^{N-1}(-a)^n+(-b)^n=\frac{1-(-a)^N}{1+a}+\frac{1-(-b)^N}{1+b}=\frac{1-(-a)^{R}}{1+a}+\frac{1-(-b)^R}{1+b},\qquad N\equiv R\mod 6 \end{aligned}

Use a 3 = b 3 = 1 a^3=b^3=1 in the last simplification. Now observe N = 1730 2 m o d 6 N=1730\equiv 2\mod 6 to obtain

S = 1 a 2 1 + a + 1 b 2 1 + b = 3. Binomi 1 a + 1 b = Vieta 2 + 1 = 3 \begin{aligned} S&=\frac{1-a^2}{1+a}+\frac{1-b^2}{1+b}\underset{\text{3. Binomi}}{=}1-a+1-b\underset{\text{Vieta}}{=}2+1=\fbox{3} \end{aligned}

Rem.: You can use the first equation together with Vieta's formula to gain an expression for any N N N\in\mathbb{N} :

S = { 0 , N { 0 ; 4 } m o d 6 2 , N { 1 ; 3 } m o d 6 3 , N 2 m o d 6 1 , N 5 m o d 6 \begin{aligned} S=\begin{cases} 0,&N\in\{0;\:4\}\mod 6\\ 2,&N\in\{1;\:3\}\mod 6\\ 3,&N\equiv 2\mod 6\\ -1,&N\equiv 5\mod 6 \end{cases} \end{aligned}

Akash Singh
Sep 14, 2015

note that vn always equal to -1

v0=2

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...