Newton's sums in reverse

Algebra Level 4

a + b + c N a 4 + b 4 + c 4 = 32 a 5 + b 5 + c 5 = 186 a 6 + b 6 + c 6 = 803 \begin{array} { l l} a + b + c & \in \mathbb{N} \\ a^4+b^4+c^4 & =32\\ a^5+b^5+c^5& =186\\ a^6+b^6+c^6& =803\\ \end{array}

What is the value of a + b + c a+b+c ?


The answer is 6.

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

Well, this was my attempt to solve this:

Let the polynomial P ( x ) = x 3 + m x 2 + n x + p P(x)=x^3+mx^2+nx+p have roots a a , b b and c c . Now, if we apply Newton's sums recursively we obtain:

a + b + c = m a+b+c=-m

a 2 + b 2 + c 2 = m 2 2 n a^2+b^2+c^2=m^2-2n

a 3 + b 3 + c 3 = m 3 + 3 m n 3 p a^3+b^3+c^3=-m^3+3mn-3p

a 4 + b 4 + c 4 = m 4 4 m 2 n + 4 m p + 2 n 2 a^4+b^4+c^4=m^4-4m^2n+4mp+2n^2

a 5 + b 5 + c 5 = m 5 + 5 m 3 n 5 m 2 p 5 m n 2 + 5 n p a^5+b^5+c^5=-m^5+5m^3n-5m^2p-5mn^2+5np

a 6 + b 6 + c 6 = m 6 6 m 4 n + 6 m 3 p + 9 m 2 n 2 12 m n p 2 n 3 + 3 p 2 a^6+b^6+c^6=m^6-6m^4n+6m^3p+9m^2n^2-12mnp-2n^3+3p^2

With the known values, the following system is formed:

m 4 4 m 2 n + 4 m p + 2 n 2 = 32 m^4-4m^2n+4mp+2n^2=32

m 5 + 5 m 3 n 5 m 2 p 5 m n 2 + 5 n p = 186 -m^5+5m^3n-5m^2p-5mn^2+5np=186

m 6 6 m 4 n + 6 m 3 p + 9 m 2 n 2 12 m n p 2 n 3 + 3 p 2 = 803 m^6-6m^4n+6m^3p+9m^2n^2-12mnp-2n^3+3p^2=803

It's easy to solve for p p from the first equation:

p = 32 + 4 m 2 n 2 n 2 m 4 4 m p=\dfrac{32+4m^2n-2n^2-m^4}{4m}

But the things get worse when we substitute this value, we get the following two equations after simplifying:

m 6 + 10 m 2 n 2 10 n 3 160 m 2 5 m 4 n 744 m + 160 n = 0 m^6+10m^2n^2-10n^3-160m^2-5m^4n-744m+160n=0

14 m 4 n 2 + 32 m 2 n 34 m 2 n 3 m 6 n + 12 n 4 384 n 2 224 m 4 3720 m 3 12848 m 2 + 3072 = 0 14m^4n^2+32m^2n-34m^2n^3-m^6n+12n^4-384n^2-224m^4-3720m^3-12848m^2+3072=0

I got stuck there, so the only thing I could do was solving for n n from the first equation of these two and substitute on the second equation. But the expression for n n is horrible because it's from a 3rd degree polynomial. So, after hours and hours, I got an equation in m m with the help of Excel and Wolfram Alpha to do the sums with very large numbers. This is the equation in m m of degree 22:

m 22 4320 m 18 84816 m 17 401500 m 16 + 1228800 m 14 + 29283840 m 13 + 329484096 m 12 + 1792296000 m 11 + 3273475750 m 10 9256550400 m 9 83841807360 m 8 434166425856 m 7 1696694220000 m 6 5555724768000 m 5 9731157199200 m 4 + 296518164480 m 3 + 1071516905856 m 2 = 0 m^{22}-4320m^{18}-84816m^{17}-401500m^{16}+1228800m^{14}+29283840m^{13}+329484096m^{12}\\+1792296000m^{11}+3273475750m^{10}-9256550400m^9-83841807360m^8-434166425856m^7\\-1696694220000m^6-5555724768000m^5-9731157199200m^4+296518164480m^3+\\1071516905856m^2=0

Since the coefficients of m m and m 0 m^0 are 0 0 and m 0 m \neq 0 , the degree of the equation is now 20, it's a lot easier now, don't you think? :D

Finally, using Newton-Raphson method, we see that the equation has 4 real roots and 16 complex roots. Since m m , n n and p p are real, these are the solutions for m m :

m = 6 m=-6 , m 0.3482217748938019 m\approx -0.3482217748938019 , m 0.3159770443715433 m\approx 0.3159770443715433 or m 11.02483070176856 m\approx 11.02483070176856

Hence, a + b + c = 6 a+b+c=6 , a + b + c 0.3482217748938019 a+b+c \approx 0.3482217748938019 , a + b + c 0.3159770443715433 a+b+c \approx -0.3159770443715433 or a + b + c 11.02483070176856 a+b+c \approx -11.02483070176856 .

OH MY GOSH, I'M SO SORRY. I DIDN'T MEAN TO DO THIS TO YOU.

Trevor Arashiro - 6 years, 6 months ago

Log in to reply

What was you method?

Alan Enrique Ontiveros Salazar - 6 years, 6 months ago

What was your method?

Razvan Dita - 2 years, 3 months ago

It is given a b c are natural numbers that never fits in to the question.pl correct the question

nagarjuna reddy - 5 years, 4 months ago

Log in to reply

I agree with Nagarjuna about the question posted. It does fit.

Dante Hollanda - 5 years ago

Log in to reply

a+b+c is a natural number. the question says nothing about the individual values

Emmalee Kitty James - 2 years, 9 months ago

Hello, on your other question , I had a similar problem. I got a + b + c = 16.506078736905175 a+b+c=-16.506078736905175

a + b + c = 2.4165089132142988 a+b+c=2.4165089132142988

a + b + c = 9.5895698236908771 a+b+c=9.5895698236908771

But no integer. I am sure I am missing something but this shows that there are more than 1 1 solution for your question. Are you sure that there is less than 1 1 integer solution for your question (both for a + b + c a+b+c and a 10 + b 10 + c 10 {a}^{10}+{b}^{10}+{c}^{10}

Amazing solution by the way

Julian Poon - 6 years, 6 months ago

Log in to reply

At the end you should get a quartic equation with one integer solución, discard the others. I checked it and it's only one integer solución both for a + b + c a+b+c and a 10 + b 10 + c 10 a^{10}+b^{10}+c^{10} . Also, thanks :D

Alan Enrique Ontiveros Salazar - 6 years, 6 months ago

Log in to reply

Oh okay thanks!

Julian Poon - 6 years, 6 months ago

I would like to see Trevor's method.

Jeremy Weissmann - 4 years, 9 months ago

A supercomputer is needed to solve a set of three equations with three unknowns with reduced powers to a, a^1.25, a^1.5 ... etc. Wolfram alpha times out

Harout G. Vartanian - 4 years, 5 months ago

let a=z; b=x+y i; c=x-y i (1 real and 2 complex, cause we have the second equation, and real constants), where x,y,z are real, and substituting this into the above equations, then we can deal with a system of 3x3 equations, but with only reals, solvable with less computer effort (maybe), cause i cant think yet of a better way to solve this :S

Eliud Alejandro Maldonado Sanchez - 3 years, 5 months ago
Ron van den Burg
Feb 10, 2018

Assume a third grade monic polynomial with roots a, b and c. With notations from the wiki on Newton's Identities, in excel create a table with the columns: p 1 , p 2 , p 3 , e 2 , e 3 , p 4 , p 5 , p 6 p_1, p_2, p_3, e_2, e_3, p_4, p_5, p_6 Use the following values/formula's:

p 1 = 1 p_1=1 a positive integer we start with

Skip p 2 p_2 for a moment

p 3 = ( p 1 4 / 24 p 1 2 p 2 / 4 + p 2 2 / 8 32 / 4 ) / ( p 1 / 3 ) p_3=(p_1^4/24-p_1^2*p_2/4+p_2^2/8-32/4)/(-p_1/3)

e 2 = ( p 1 2 p 2 ) / 2 e_2=(p_1^2-p_2)/2

e 3 = ( p 1 3 3 p 1 p 2 + 2 p 3 ) / 6 e_3=(p_1^3-3*p_1*p_2+2*p_3)/6

p 4 = p 1 p 3 e 2 p 2 + e 3 p 1 p_4=p_1*p_3-e_2*p_2+e_3*p_1

p 5 = p 1 p 4 e 2 p 3 + e 3 p 2 p_5=p_1*p_4-e_2*p_3+e_3*p_2

p 6 = p 1 p 5 e 2 p 4 + e 3 p 3 p_6=p_1*p_5-e_2*p_4+e_3*p_3

Now use the 'What-If Analysis' and then 'Goal seek...' tool in Excel (see tab 'Data') to

'Set cell' = p 6 p_6

'To value' = 803

'By changing' = p 2 p_2

You will then get p 5 = 718.72 p_5=-718.72 , clearly wrong. Changing p 1 p_1 to 2 and 'Goal seek...' again, gives p 5 = 529.95 p_5=529.95 . Repeat this until the values of p 4 , p 5 , p 6 p_4, p_5, p_6 are 32, 186 and 803.

This happens at p 1 = a + b + c = 6 p_1=a+b+c=6 .

=============

Now it appears that the monic polynomial is

x 3 6 x 2 + 14 x 15 x^3-6x^2+14x-15

and thus a , b , c a, b, c equal 3 , 3 + 11 i 2 , 3 11 i 2 3, \frac{3+\sqrt{11}i}2, \frac{3-\sqrt{11}i}2 or any other permutation.

Nice apprach.

While you have found the correct monic cubic polynomial satisfying these equations, how would you know that the polynomial that you've found is unique?

Pi Han Goh - 3 years, 4 months ago

Log in to reply

Good question.... I will think about it.

Ron van den Burg - 3 years, 3 months ago

Not helpful

Ankita Bailwad - 1 year, 6 months ago
Lu Chee Ket
Jan 31, 2015

{6, 14, 15}

S = 6

P = 15

I = 14/ 15

This is simple with computer. The answer is 6.

Care to elaborate?

Pi Han Goh - 6 years, 4 months ago

Log in to reply

I think he programmed the solution to brute-force it.

Josiah Kiok - 5 years, 1 month ago

Lu listed the values of the elementary polynomials { e 1 , e 2 , e 3 } \{e_1, e_2, e_3\} wrt to the monic polynomial p ( x ) = ( x a ) ( x b ) ( x c ) = x 3 e 1 x 2 + e 2 x e 3 p(x)=(x-a)(x-b)(x-c)=x^3-e_1x^2+e_2x-e_3 .

Then he lists:

S = e 1 = a + b + c = 6 S = e_1 = a + b + c = 6 , the Sum.

P = e 3 = a b c = 15 P = e_3 = abc = 15 , the Product.

I don't know what computer program was used. I did however write a solution using the 'Goal Seek' function in Excel.

Ron van den Burg - 3 years, 4 months ago

I think that it hasn't right because 1.{a,b,c}∈2n (32) 2.{a,b,c}∈2n-1 (803) So it can't exist

Mahrad Poustindoz - 3 years, 11 months ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...