Can you guess it?

Algebra Level 5

If the three real roots of the equation 7 x 3 + 7 x 2 7 x + 1 = 0 7x^3+7x^2-7x+1=0 are:

x 1 = a a sec ( m π b ) x 2 = a a + sec ( n π b ) x 3 = a a + sec ( p π b ) \begin{aligned} x_1 &=&\dfrac{a}{a-\sec\left(\dfrac{m \pi}{b}\right)} \\ x_2 &=&\dfrac{a}{a+\sec\left(\dfrac{n \pi}{b}\right)} \\ x_3&=&\dfrac{a}{a+\sec\left(\dfrac{p \pi}{b}\right)} \\ \end{aligned}

Where x 1 < x 2 < x 3 x_1 < x_2 < x_3 ; a , b , m , n , p a, b, m, n, p are positive integers with gcd ( m , b ) = gcd ( n , b ) = gcd ( p , b ) = 1 \text{gcd}(m,b)=\text{gcd}(n,b)=\text{gcd}(p,b)=1 and 2 m , 2 n , 2 p b 2m,2n,2p \leq b

Find 2 b ( a + m + n + p ) 2b-(a+m+n+p) .

Inspired by this and this .


The answer is 7.

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

Make the substitution x = t t 2 x=\dfrac{t}{t-2} :

7 ( t t 2 ) 3 + 7 ( t t 2 ) 2 7 ( t t 2 ) + 1 = 0 7\left(\dfrac{t}{t-2}\right)^3+7\left(\dfrac{t}{t-2}\right)^2-7\left(\dfrac{t}{t-2}\right)+1=0

Expand everything and multiply both sides by ( t 2 ) 3 8 \dfrac{(t-2)^3}{8} :

t 3 + t 2 2 t 1 = 0 t^3+t^2-2t-1=0

That is the minimal polynomial of 2 cos ( 2 π k 7 ) 2\cos \left(\dfrac{2 \pi k}{7}\right) , for k [ 1 , 3 ] k\in[1,3] . To prove that just let z 7 = 1 z^7=1 , that implies that z = cos ( 2 π k 7 ) + i sin ( 2 π k 7 ) z=\cos\left(\dfrac{2 \pi k}{7}\right)+i\sin\left(\dfrac{2 \pi k}{7}\right) and ( z 1 ) ( z 6 + z 5 + z 4 + z 3 + z 2 + z + 1 ) = 0 (z-1)(z^6+z^5+z^4+z^3+z^2+z+1)=0 . Now let t = z + 1 z = 2 cos ( 2 π k 7 ) t=z+\dfrac{1}{z}=2 \cos\left(\dfrac{2 \pi k}{7}\right) . Dividing both sides of the equation in z z by z 3 z^3 we obtain; z 3 + z 2 + z + 1 + 1 z + 1 z 2 + 1 z 3 = 0 z^3+z^2+z+1+\dfrac{1}{z}+\dfrac{1}{z^2}+\dfrac{1}{z^3}=0 . After a little algebraic manipulation we obtain ( z + 1 z ) 3 + ( z + 1 z ) 2 2 ( z + 1 z ) 1 = 0 \left(z+\dfrac{1}{z}\right)^3+\left(z+\dfrac{1}{z}\right)^2-2\left(z+\dfrac{1}{z}\right)-1=0 , so t 3 + t 2 2 t 1 = 0 t^3+t^2-2t-1=0 .

Then, our roots are:

x = t t 2 = 1 1 2 t = 1 1 sec ( 2 π k 7 ) x=\dfrac{t}{t-2}=\dfrac{1}{1-\frac{2}{t}}=\dfrac{1}{1-\sec\left(\dfrac{2 \pi k}{7}\right)} .

By the given conditions, after adjusting the angles in the range ( 0 , π 2 ) (0, \frac{\pi}{2}) , we conclude that:

x 1 = 1 1 sec ( 2 π 7 ) x 2 = 1 1 + sec ( 3 π 7 ) x 3 = 1 1 + sec ( π 7 ) x_1=\dfrac{1}{1-\sec\left(\dfrac{2 \pi}{7}\right)} \\ x_2=\dfrac{1}{1+\sec\left(\dfrac{3 \pi}{7}\right)} \\ x_3=\dfrac{1}{1+\sec\left(\dfrac{\pi}{7}\right)}

Hence, a = 1 a=1 , b = 7 b=7 , m = 2 m=2 , n = 3 n=3 , p = 1 p=1 and 2 b ( a + m + n + p ) = 7 2b-(a+m+n+p)=\boxed{7} .

Since you asked for us to guess it , I did guess it !

A Former Brilliant Member - 6 years, 2 months ago

Log in to reply

Yes, you can do that, and the hint is that there are a lot of sevens :D But it's still a 400 points problem!

Alan Enrique Ontiveros Salazar - 6 years, 2 months ago

Log in to reply

Exactly ! And the rest was easy , assign arbitrary values of 1,2,3 and start with the guessing.

Yeah , it's a 400 point problem and I didn't do it justice by not solving it:P

A Former Brilliant Member - 6 years, 2 months ago
Chew-Seong Cheong
Mar 29, 2015

As the problem's title suggests, I actually guess the solution. The below is how I did it with a spreadsheet.

Let f ( x ) = 7 x 3 + 7 x 2 7 x + 1 f(x) = 7x^3+7x^2-7x+1 , Then x 1 x_1 , x 2 x_2 and x 3 x_3 can be found by Newton's method x n + 1 = x n f ( x n ) f ( x n ) = x n 7 x n 3 + 7 x n 2 7 x n + 1 21 x n 2 + 14 x n 7 x_{n+1} = x_n - \dfrac {f(x_n)}{f'(x_n)} = x_n - \dfrac {7x_n^3+7x_n^2-7x_n+1}{21x_n^2+14x_n-7}

The three roots are found to be x 1 = 1.655970555 x_1 = -1.655970555 , x 2 = 0.182018097 x_2=0.182018097 and x 3 = 0.473952458 x_3 = 0.473952458 .

Let θ 1 = m π b \theta_1 = \dfrac {m\pi}{b} , θ 2 = n π b \theta_2 = \dfrac {n\pi}{b} and θ 3 = p π b \theta_3 = \dfrac {p\pi}{b} ; then, we have:

x 1 = a a sec θ 1 = a a 1 cos θ 1 = a cos θ 1 a cos θ 1 1 a cos θ 1 = x 1 x 1 1 x_1 = \dfrac {a}{a - \sec {\theta_1}} = \dfrac {a}{a - \dfrac {1}{\cos {\theta_1}}} = \dfrac {a\cos{\theta_1}}{a\cos{\theta_1}-1} \quad \Rightarrow a\cos{\theta_1} = \dfrac {x_1}{x_1-1}

Similarly a cos θ 2 = x 2 1 x 2 a\cos{\theta_2} = \dfrac {x_2}{1-x_2} and a cos θ 3 = x 3 1 x 3 a\cos{\theta_3} = \dfrac {x_3}{1-x_3}

Assuming a = 1 a = 1 , we find that:

\(\begin{array} {} \cos{\theta_1} = 0.623489802 & \theta_1 = 0.897597901 & \dfrac {\theta_1}{\pi} = \dfrac {m}{b} = 0.285714286 \\ \cos{\theta_2} = 0.222520934 & \theta_2 = 1.346396852 & \dfrac {\theta_2}{\pi} = \dfrac {n}{b} = 0.428571429\\ \cos{\theta_3} = 0.900968868 & \theta_3 = 0.448798951 & \dfrac {\theta_3}{\pi} = \dfrac {p}{b} = 0.142857143 \end{array} \)

By try and error, we find that:

\(\begin{array} {} \dfrac {\theta_1}{\pi} \times 7 = 0.285714286 \times 7 = 2 & \Rightarrow b = 7 \quad m = 2 \\ \dfrac {\theta_2}{\pi} \times 7 = 0.428571429 \times 7 = 3 &\Rightarrow b = 7 \quad n = 3 \\ \dfrac {\theta_3} {\pi} \times 7 = 0.142857143 \times 7 = 1 & \Rightarrow b = 7 \quad p = 1 \end{array} \)

Therefore, 2 b ( a + m + n + p ) = 2 × 7 ( 1 + 2 + 3 + 1 ) = 7 2b - (a+m+n+p) = 2\times 7 - (1+2+3+1) = \boxed{7}

As far as the solution is concerned, it seems ok! to me. But i found the following using Wolfram Alpha, which says the solutions of the equations are not real. https://www.wolframalpha.com/input/?i=solve%287 x 3+%2B+7 x 2+-+7*x+%2B+1%29 . Now i doubt that, since i used some other method and found the roots are in fact not real. Correct me, if you find something wrong.

Gaurav Dhingra - 6 years, 2 months ago

Log in to reply

Just substitute the three values in the equation and you will find that they are roots. I also plotted the graph before I get the more accurate answers.

Chew-Seong Cheong - 6 years, 2 months ago

Log in to reply

Actually the roots turn out to be 1.65597055521136 7.0 1 0 24 i -1.65597055521136 - 7.0 \cdot 10^{-24} i , 0.182018097012207 + 7.0 1 0 21 i 0.182018097012207 + 7.0 \cdot 10^{-21} i and 0.473952458199156 + 7.0 1 0 21 i 0.473952458199156 + 7.0 \cdot 10^{-21} i now which have almost the same values as the values you found. And substituting the these values it turns out that, these are the roots.

Gaurav Dhingra - 6 years, 2 months ago

Log in to reply

@Gaurav Dhingra PLZ TRY OUT MY MECHANICS PROBLEM

Karan Gujar - 6 years, 2 months ago

@Gaurav Dhingra The imaginary parts are actually floating point errors which are inherent to all computing languages because of the inability to represent a decimal number to infinite accuracy. They are actually 0 instead of negligible imaginary numbers.

Chew-Seong Cheong - 6 years, 2 months ago

Log in to reply

@Chew-Seong Cheong I did not knew that even the imaginary numbers could be inherent as errors in computing, even if the actual numbers are pure real numbers. Thanks. :)

Gaurav Dhingra - 6 years, 2 months ago

Log in to reply

@Gaurav Dhingra I think it's because it involves 1/7 in the calculations and 1/7 is a non-terminating decimal. The calculations also involve square roots. It just happens that all three have negative errors instead of zeros in the square roots.

Chew-Seong Cheong - 6 years, 2 months ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...