⎩ ⎪ ⎪ ⎨ ⎪ ⎪ ⎧ x 2 + y 2 = 1 6 z 2 + t 2 = 2 5 x t − y z = 2 0
If real numbers t , x , y , z satisfy the system of equations above, find the maximum value of P = x z .
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.
Nice One..!!
Simple and best!
Liked it man congrats
not dat I understood shit but compared to the others this was so short and so good :) gratz
Best solution
WOW! I really wonder about this!
I love this solution! It is so simple and elegant.
Solution 1 (Geometric)
2 1 ( x t − y z ) is the area of the triangle with vertices ( 0 , 0 ) , ( x , y ) , ( z , t ) . But since ( x , y ) lies on a circle with radius 4 and ( z , t ) lies on a circle with radius 5, this area is less than or equal to 2 1 × 5 × 4 = 1 0 with equality if ( x , y ) and ( z , t ) are perpendicular. We have equality so we can write x = 4 cos θ and z = − 5 sin θ and x z = − 1 0 sin 2 θ which has a maximum of 10.
Solution 2 (Algebraic)
Cauchy-Schwarz gives x t − y z ≤ x 2 + y 2 z 2 + y 2 = 2 0 , since we have equality we have t = λ x , z = − λ y and x z = λ z t . Substituting into z 2 + t 2 = 2 5 gives λ 2 = 1 6 2 5 and AM-GM gives z t ≤ 2 2 5 . It is easy to show from this that x z ≤ 1 0
In both cases we can obtain equality using the points ( x , y ) = ( 2 4 , − 2 4 ) ( z , t ) = ( 2 5 , 2 5 )
Ah! I thought Cauchy-Schwarz would work but I found the geometric solution first which is more intuitive I think. Nice explanation!
I thought according to basics that it is difficult to solve as basics say that number of variable should be equal to number of equation given . So can ayone solve my query if they understood what I meant to say ?
This is just an excelent solution.
Is there a direct formula for the area of a triangle or did you derive it and didn't show it here ? (how did you get that area of triangle formula) I mean, I calculated the area of the triangle and got the same formula but how did you get it directly ?
Did by your geometry solution sir
excellent solution sir.
You knew this.
you are all wrong and are solutions are nonsense.because: let xz=10 now we have three Variables and three equations . so we can solve for x.writing the long solvation is very time consuming but if you solve for x this polynomial will appear: 0.25x^4+0.02x^3+31.8x^2+1.28x+1024=0 or maybe you achieve another polynomial with different Coefficients but the same roots.and the roots of that polynomial are : x 1 = 0.42766+7.98856*i x 2 = 0.42766-7.98856 i x_3 = -0.46766+7.98632 i x_4 = -0.46766-7.98632*i yes you can see NONE OF THEM IS REAL THEY ARE ALL COMPLEX NUMBERS but the question says x is a real number (check it) by the way maximum is 0 what do you say to that ?
Log in to reply
You've made a mistake somewhere. Let x = 2 2 , y = − 2 2 , z = t = 2 5 , and then x z = 1 0 and the three equations are satisfied.
Let u = ( x , y ) and v = ( t , − z ) . Then the three given equations are ∥ u ∥ 2 = 1 6 ∥ v ∥ 2 = 2 5 u ⋅ v = 2 0 Since 2 0 = u ⋅ v = ∥ u ∥ ∥ v ∥ cos θ = 4 ⋅ 5 cos θ where θ is the angle between the vectors, we have cos θ = 1 and the vectors are co-directional. That is, v = 4 5 u .
We are to find the maximum of P = x z = − 4 5 x y given that x and y are on the circle x 2 + y 2 = 1 6 . This will occur when x = − y = ± 2 2 and P = 1 0 .
Great thinking. I did an algebraic solution.
It's easier and understandable from the above solutions
This is my process, which may be a bit long due to unnecessary steps.
1 6 × 2 5 = 2 0 2 so ( x 2 + y 2 ) ( z 2 + t 2 ) = ( x t − y z ) 2
Expanding gives ( x z ) 2 + ( x t ) 2 + ( y z ) 2 + ( y t ) 2 = ( x t ) 2 + ( y z ) 2 − 2 x y z t ( x z ) 2 + ( y t ) 2 = − 2 x y z t
LHS is positive so RHS must be positive as well, so one or three of the variables are negative. Looking at x t − y z = 2 0 , one of y , z is negative and either both of x , t are positive or negative, and flipping their signs won't change the equations so WLOG let both of them be positive. Flip the sign of the other negative variable to get x t + y z = 2 0 as the new third equation. The other two equations aren't affected. Now all variables are positive.
Using AM-GM on 2 ( x z ) 2 + ( y t ) 2 gives ( x z ) 2 + ( y t ) 2 ≥ 2 x y z t but it was given that ( x z ) 2 + ( y t ) 2 = 2 x y z t (negative sign was removed because of new equations) so ( x z ) 2 = ( y t ) 2 due to equality and x z = y t because both are positive.
Let x z = y t = a . The first two equations gives z x + t y = a 1 6 and x z + y t = a 1 6 . Dividing the equations gives z t x y = 2 5 1 6 (I found this by letting a = z x and b = t y ). Dividing a from both sides of 2 5 x y = 1 6 z t gives 2 5 t x = 1 6 x t and 2 5 z y = 1 6 y z or 5 x = 4 t and 5 y = 4 z .
Using AM-GM on the first equation gives x y ≤ 8 . Substituting y = 5 4 z gives x z ≤ 1 0 .
like it method
I just did it the same way. :DI
From Lagrange's identity; ( x 2 + y 2 ) ( z 2 + t 2 ) = ( x t − y z ) 2 + ( x z + y t ) 2 .
Substituting we get x z + y t = 0 → x z = − y t
From equation 1 and 2; using AM-GM we get
4 0 0 = ( x 2 + y 2 ) ( z 2 + t 2 ) ≥ 4 x 2 y 2 z 2 t 2 = 4 ∣ x y z t ∣ = 4 ∣ − ( x z ) 2 ∣ = 4 ( x z ) 2
Therefore, x z ≤ 1 0 ~~~
Very clean, thanks!
Yes AM-GM is the best way to get this!!
I solved this using complex numbers.
From the first two equations, we have ∣ x + y i ∣ = 4 and ∣ t + z i ∣ = 5 . Therefore, 4 ∗ 5 = ∣ ( x + y i ) ( t + z i ) ∣ = ∣ ( x t − y z ) + ( x z + y t ) i ∣ = ∣ 2 0 + ( x z + y t ) i ∣
Thus, ∣ 2 0 + ( x z + y t ) i ∣ = 2 0 . Therefore, we must have x z + y t = 0 . Therefore, ( x + y i ) ( t + z i ) = 2 0 .
Since ∣ x + y i ∣ = 4 and ∣ t + z i ∣ = 5 , there must exist real numbers a , b such that x + y i = 4 e i a and t + z i = 5 e i b . Their product is 2 0 e i ( a + b ) which must equal 20. Hence b = − a . Using Euler's formula, we write x = 4 c o s ( a ) and z = 5 s i n ( b ) = − 5 s i n ( a ) . Therefore x z = − 2 0 c o s ( a ) s i n ( a ) = − 1 0 s i n ( 2 a ) . This has a maximum value of 1 0 .
I solved this using complex numbers.
Me too. Not identical to your method, but close enough that I don't see a need to post my solution.
Not by guessing,
x t = 20 + y z
Squaring both sides. With x^2 = 16 - y^2 and t^2 = 25 - z^2,
x^2 t^2 = 400 + 40 y z + y^2 z^2
=> (16 - y^2)(25 - z^2) = 400 + 40 y z + y^2 z^2
=> 400 - 25 y^2 - 16 z^2 + y^2 z^2 = 400 + 40 y z + y^2 z^2
=> 25 y^2 + 40 y z + 16 z^2 = 0
=> (5 y + 4 z)^2 = 0
=> z = (- 5/ 4) y
Since x = +/- sqrt(16 - y^2),
Maximum x z = (5/ 4) sqrt(16 y^2 - y^4);
Let f = 16 y^2 - y^4,
d f/ d y = 32 y - 4 y^3 and d^2 f/ d y^2 = 32 - 12 y^2
d f / d y = 0 => y^2 = 8 where d^2 f/ d y^2 = -64 < 0 {Indicates a maximum}
Maximum x z = (5/ 4) sqrt(16 x 8 - 64) = 10 where
[t, x, y, z] = [5/ sqrt(2), 2 sqrt(2), - 2 sqrt(2), 5/ sqrt(2)] OR [- 5/ sqrt(2), - 2 sqrt(2), 2 sqrt(2), - 5/ sqrt(2)] which satisfies 3 initial equations.
My solution got me the correct answer but I don't know why it works... make x^2=y^2, z^2=t^2, then we have (25/2)^(1/2) and 8^(1/2), when multiplied, gives 10.
it only works because they happen to be equal. The first two equations are circle equations. It just so happens that the maximum is where the angle is 45 degrees aka x & y are equal. But you shouldn't rely on that as a given.
I have a simplistic solution, which gave me the correct answer, but it might be incorrect, since I did not even need the 3rd equation. So how I solved this was:
X^{2} + Y^{2} =16 ---- (1) Perform partial differentiation to this, to get: 2X+2Y=0. Equate the left hand side to 0, to arrive at Y = -X. Substitute this in equation (1), and solve for X. We get X = 2 multiplied by Root of 2. This is the maximum value of X.
Perform similarly for z^{2} + t^{2} =25, solve for z, to arrive at the maximum value of z as 5 / Root of 2.
Multiply these two maximum values of x and z:
{2 . (Root of 2 )} multiplied by { 5 / Root of 2 } We get 10.
This isn't quite right, as the maxima of x 2 + y 2 = 1 6 is actually ( 4 , 0 ) and ( 0 , 4 ) respectively, and the maxima of z 2 + t 2 = 2 5 is ( 5 , 0 ) and ( 0 , 5 ) respectively, as these are both circle equations.
Your result might just be coincidence.
Without the third equation, the maximum of x z is achieved when ( x , y , z , t ) = ( 4 , 0 , 5 , 0 ) and the max is 20.
Log in to reply
Dawn ur these solutions are not satisfying the equation xt-yz=20
Their is only way to solve this question,any easy way to solve this question.
Very clean too!
What you have started with is not partial differentiation. If you take a total derivative or perform implicit differentiation you find 2 x + 2 y d x d y = 0 so above you are implicitly assuming d y / d x = 1 . The circle x 2 + y 2 = 1 6 has only the two points ( 2 2 , − 2 2 ) , ( − 2 2 , 2 2 ) which obey this, and similarly for the z and t variables. It was just lucky for you that the solution to the problem is achieved for x and z at these special points!
Try to solve the problem again with the final equation replaced by x t − y z = 1 0 and you'll see the necessity of that third constraint.
I got that x z ≤ 4 1 / 4 . Here's my work... Notice how 16,25,and 20 have a relationship: 1 6 × 2 5 = 2 0 Therefore, we can combine these three equations into x 2 + y 2 z 2 + t 2 = x t − y z Simplifying gives x 2 z 2 + 2 x z y t + y 2 t 2 = 0 -> x z + y t = 0 Now we add the first two equations together. Using our previous finding and factoring the difference of squares we get ( x + z − 4 1 ) ( x + z + 4 1 ) = − ( y − z ) 2 Since the right side is negative, the only way for this equation to be true is if − 4 1 ≤ x + z ≤ 4 1 For xz to be maximized, we look at the cases x + z = 4 1 and x + z = − 4 1 . It is pretty easy to see that that they will both yield the same maximum value of xz since the xs and zs yielding the maximum value of xz in each case will just be negatives of each other. Therefore, we can just look at x + z = 4 1 and assume that x and z are non-negative real numbers to use AM-GM. x z ≤ 2 x + z -> x z ≤ 2 4 1 -> x z ≤ 4 1 / 4
Trigonometry will be easier here.. x = 4 cos α ; y = 4 sin α
z = 5 cos β ; t = 5 sin β
x t − y z = 2 0 ⇒ 2 0 ( sin β cos α − cos β sin α ) = 2 0 ⇒ sin ( β − α ) = 1
⇒ β − α = 2 π
P = x z ⇒ P = 2 0 cos α cos β = 2 0 sin β cos β = 1 0 sin 2 β whose max value is 10!
Log in to reply
Well done pranjal.Great method
May I ask why cos a = sin B ? thnks
Log in to reply
since, b=a+90 >cos(b)=cos(a+90)=-sina >P=-20sin(a)cos(a)=-10sin(2a) >Pmax=-10(-1)=10
Yes Pranjal. That's how it is supposed to be solved :D
I solved it the same way.
the two equation represents a circle . applying their parametric form you will get relation between their eccentric angles . using this relation maximize the third equation by differentiating it. youll get 10
Same method Pranjal.
I suggest you double-enter your text to separate sentences for more clarity. Otherwise it looks like a giant block of text (sort of like an essay without paragraphs).
I think I found simple algebra in it.
eq. 1
×
eq. 2
⇒
x
2
z
2
+
y
2
t
2
+
x
2
t
2
+
y
2
t
2
=
4
0
0
⋯
(
4
)
squaring eq. 3
⇒
x
2
t
2
+
y
2
z
2
−
2
x
y
z
t
=
4
0
0
⋯
(
5
)
(
4
)
−
(
5
)
⇒
x
2
z
2
+
y
2
t
2
+
2
x
y
z
t
=
0
⇒
x
z
+
y
t
=
0
⇒
x
z
=
−
y
t
Also from (1) and (2) we've got
y
=
1
6
−
x
2
and
t
=
2
5
−
t
2
∴
x
2
z
2
=
y
2
t
2
=
4
0
0
−
2
5
x
2
−
1
6
z
2
+
x
2
z
2
⇒
(
5
x
)
2
+
(
4
z
)
2
=
2
0
2
∴
z
=
4
5
1
6
−
x
2
∴
x
z
=
4
5
×
x
2
(
1
6
−
x
2
)
=
6
4
−
(
x
+
8
)
2
⇒
max
(
x
z
)
=
1
0
Well....we could try to put all variables depended upon a particular parameter....and get an equation whose maximum value is required.....well to put it in simpler words....consider x=4cos(k) y=-4sin(k), Z=5sin(k) and and t=5cos(k)...these paramteric substitution satisfy all the given three equations...and we need to maximize P=20 sin(k)cos(k)....which should be a piece of cake....sorry for the dull solution though.... I hope it helps
take complex no. method let H = x+iy and G=z-it Use these modulus{H+G}< mod{H}+mod{G}...............................................1 HG={xz+yt}+i{yz-xt}.............................2 just use above eq.
From 1st and 2nd equations we get:-
−
4
≤
x
,
y
≤
4
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
−
5
≤
t
,
z
≤
5
.
T
o
s
a
t
i
s
f
y
e
q
u
a
t
i
o
n
3
.
.
.
.
.
.
(
x
,
y
,
z
,
t
)
=
(
4
,
0
0
,
5
)
o
r
=
(
0
,
4
,
5
,
0
)
w
h
e
n
P
=
0
.
To obtain extreme values, we can create symetray by....
x
2
=
y
2
a
n
d
t
2
=
z
2
a
n
d
t
o
s
a
t
i
s
f
y
3
r
d
e
q
u
a
t
i
o
n
w
e
l
e
t
y
=
−
x
.
T
h
e
n
(
x
,
y
,
z
,
t
)
=
(
2
∗
2
,
2
∗
2
,
5
/
2
,
5
/
2
)
a
n
d
P
=
1
0
.
{ I am not satisfied with my arguments. Can any one improve ?}
Problem Loading...
Note Loading...
Set Loading...
Let x = 4 s i n A , y = 4 c o s A . z = 5 s i n B , t = 5 c o s B .
x t − y z = 2 0 ( s i n ( B − A ) ) .
B − A = 9 0 .
P = x z = 2 0 c o s A c o s B = 1 0 s i n 2 B .
M a x : 1 0 .