Inspired by Harsh Shrivastava

Algebra Level 2

{ x + y + z = 4 x 2 + y 2 + z 2 = 6 \large{\begin{cases} x + y + z &= 4 \\ x^2 + y^2 + z^2 &= 6 \end{cases} }

If x , y , z x, y, z are three real numbers satisfying the above system of equations, then find the maximum value of x y z xyz .


The answer is 2.

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.

9 solutions

Department 8
Sep 21, 2015

x + y + z = 4 x 2 + y 2 + z 2 = 6 x y + y z + z x = 5 x+y+z=4 \\ x^{2}+y^{2}+z^{2}=6 \\ xy+yz+zx=5

Let x , y , z x,y,z be the roots of some polynomial P ( a ) P(a) . Applying discrimination of cubic equation we will get an equation as.

0 27 d 2 104 d + 100 0 \ge 27d^{2}-104d+100

Where d = x y z d=xyz . We see the above value attains it maximum at 0 0 . Therefore the roots of equation will give the maximum value x y z xyz . The roots of above equation are 2 , 100 54 2, \frac{100}{54} . But only 2 is acceptable. How? That depends upon you.

How do you get xy+yz+zx=5? I don't understand.

Benry Burfer - 5 years, 8 months ago

Log in to reply

We have x+y+z=4 Take square two sides of this equation: (x+y+z)^2 = 16 <=> x^2+y^2+z^2 + 2(xy+yz+xz) =16 <=> 6+2(xy+yz+xz) =16 (because x^2+y^2+z^2=6) <=> xy+yz+zx=5

Ngoc Chu - 5 years, 8 months ago

Log in to reply

Ok that makes sense. Thank you.

Benry Burfer - 5 years, 8 months ago

How to do it by 3D geometry? One forms a sphere and the other is a plane and the intersection of these two equations is a circle. So what can I do after that?

Aravind V - 3 years, 8 months ago

Log in to reply

You can't. \quad

Pi Han Goh - 3 years, 8 months ago

Nice! I use a similar approach; see my separate post.

Otto Bretscher - 5 years, 8 months ago

Log in to reply

Nice observation and thanks for a new trick to solve.

Department 8 - 5 years, 8 months ago

sir try this

Department 8 - 5 years, 8 months ago

Log in to reply

Thanks! I wish I could play around with these puzzles all day, but I'm off to work, teaching a little calculus. Maybe tonight...

Otto Bretscher - 5 years, 8 months ago

I must confess that I don't fully understand the wording of the problem. Take a look at my attempt, and please do explain.

Otto Bretscher - 5 years, 8 months ago

there is a more elegant solution

Dev Sharma - 5 years, 8 months ago

Log in to reply

Define "elegant", Dev ;)

Otto Bretscher - 5 years, 8 months ago

Log in to reply

@Otto Bretscher sir, it means that there is more easier solution.

Dev Sharma - 5 years, 8 months ago

Log in to reply

@Dev Sharma There's no point in claiming that you have a better solution if you're not going to show it.

Pi Han Goh - 5 years, 7 months ago

Log in to reply

@Pi Han Goh Yes you are right

Department 8 - 5 years, 7 months ago

@Pi Han Goh I solved it using cauchy schwarz, on x and y then find out min of z, then its easy

Dev Sharma - 5 years, 7 months ago

Log in to reply

@Dev Sharma Can you post your solution?

Pi Han Goh - 5 years, 7 months ago

Log in to reply

@Pi Han Goh I am not good at latex

Dev Sharma - 5 years, 7 months ago

Log in to reply

@Dev Sharma Just type it out without latex.

Pi Han Goh - 5 years, 7 months ago

@Dev Sharma I'm sure I won't be expecting a solution from you any time soon.

Pi Han Goh - 5 years, 7 months ago

Log in to reply

@Pi Han Goh Posted ...

Dev Sharma - 5 years, 7 months ago

Seriously!!!!!!!

Department 8 - 5 years, 8 months ago

What is the discriminant of of an cubic equation or how can I derive it??

naitik sanghavi - 5 years, 8 months ago

Log in to reply

Use google

Department 8 - 5 years, 8 months ago

don't understand

Istiak Ahmed - 5 years, 8 months ago

Log in to reply

First i made an equation whose roots are x , y , z x,y,z . Then we know roots are real numbers so i used the discrimination of the equation you can find it at google.

Department 8 - 5 years, 8 months ago
Otto Bretscher
Sep 23, 2015

My approach is similar to Lakshya's, but I avoid the somewhat tedious work with the discriminant.

By Viete, we have the polynomial f ( t ) = t 3 4 t 2 + 5 t p f(t) = t^3-4t^2+5t-p whose roots are x , y , z x,y,z , where p = x y z p=xyz . This polynomial has the local maximum f ( 1 ) = 2 p f(1)=2-p . If p > 2 p>2 , we have only one real solution, but for p = 2 p=2 we get the solutions 1 , 1 , 2 1,1,2 for x , y , z x,y,z , in any order. Thus the maximum product is p = 2 p=\boxed{2} .

there is one more solution:

use 2 ( y 2 + z 2 ) > = ( y + z ) 2 ) 2(y^2 + z^2) >= (y + z)^2)

Dev Sharma - 5 years, 8 months ago

Log in to reply

Spell it out for us, please... I don't get it.

Otto Bretscher - 5 years, 8 months ago

Can you write full solution .??

Harmanjot Singh - 5 years, 8 months ago

When I get to xy+xz+yz=5, I replace y+z by 4-x, and yz by d/x giving the equation (after rearranging): d=x^3-4x^2+5x Differentiating to get max d gives x=1 or 5/3 giving d = 2 or 1.85... Hence max d is 2.

Just to play the devil's advocate: What about x = 2.01 x=2.01 ? ;)

Otto Bretscher - 5 years, 8 months ago

Log in to reply

tl;dr: x, y and z have to be between 2/3 and 2.

Please, be the devil's advocate! x = 2.01 gives a complex solution for y or z. Let's have a look at the boundaries (using geometry and graphs): let z be the one we want to bound (and the rest are similar by symmetry) (1) x^2+y^2=6-z^2 (2) x+y = 4 - z

Hence, (1) is a circle in the centre with radius sqrt(6-z^2). (2) a line with gradient -1 who's distance to (0,0) is the distance at x = y = 2-z/2 So said distance is (2-z/2) sqrt(2). To bound z we need: sqrt(6-z^2) > (2-z/2) sqrt(2) Or solving this gives: 2/3 < z < 2. Hence 2.01 is not in this and cannot be a solution.

Just for completeness: if x = 2/3, d = 1.85... and if x = 2: d = 2. And d(d)/dx = 3x^2-8x+5 which is positive from 2/3 to 1 where d = 2, and therefore cannot go higher.

Quite a long solution if you really want to be precise about it.

Benjamin Katz-Crowther - 5 years, 8 months ago

Log in to reply

I'm sorry for making you do all this work... thank you! There are easier ways to see that x, y, and z must be between 2/3 and 2: by symmetry, we must have y = z at the points with highest and lowest x values, and then the equations are easy to solve.

I just wanted to point out that one has to think about these issues in your solution. (I'm taking a slightly different approach to avoid the issue in my solution.)

Otto Bretscher - 5 years, 8 months ago

Log in to reply

@Otto Bretscher No worries, my pleasure to look into it as I was puzzled why x=2.01 gives d>2.

Benjamin Katz-Crowther - 5 years, 8 months ago

Log in to reply

@Benjamin Katz-Crowther i tried A.M>=G.M Got xyz<=2.15(approx) Can u proceed from that?

Ankit Agarwal - 5 years, 8 months ago
Ritesh Singh
Oct 2, 2015

Umm....I used quite an unorthodox way.I assumed that x^2,y^2,z^2 would be positive(since they aren't complex).And the only positive numbers whose squares add up to 6 were 2,1,1 . On top of that 2+1+1 was equal to 4.

Rishab Dev Singh
Sep 24, 2015

Ha ha.., i thought it in a simpler way...... I thouht what values of xyz we can put to satisfy eq. Putting two of them as 1 and one as 2....

Thats what i did, simply substituting the valuea for variables x,y and z.... i.e. 1,1,2

Dhwani Shah - 5 years, 8 months ago
Hiru Lakhotia
Sep 24, 2015

Hit and trail rocks

Aeshit Singh
Sep 22, 2015

Apply AM>=GM on both the equations.Then select the intersection of the inequalities and then GIF.

Dev Sharma
Nov 15, 2015

x+ y = 4 - z

x 2 + y 2 = 6 z 2 x^2 + y^2 = 6 - z^2

using cauchy schwarz, then its easy

What's the next step?

Pi Han Goh - 5 years, 7 months ago

Log in to reply

Then find min of z then find x and y by forming quadratic eq

Dev Sharma - 5 years, 7 months ago

Log in to reply

Please show your work. I doubt Cauchy Schwarz is a valid approach.

Pi Han Goh - 5 years, 7 months ago
Yugesh Kothari
Sep 24, 2015

The question doesn't make sense.x=y=z yields 2.37 as the value of the product.

There is no solution with x = y = z x=y=z .

Otto Bretscher - 5 years, 8 months ago

Log in to reply

Shoot, I missed that. I was applying AM-GM directly. Thank you!

Yugesh Kothari - 5 years, 8 months ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...