Equation with x,y as exponential functions


by cng99
Tags: exponential equation, olympiad, polynomial exponent, two variable, x y
cng99
cng99 is offline
#1
Mar8-12, 11:52 PM
P: 44
1. The problem statement, all variables and given/known data

This question was asked in the Indian National Maths Olympiad.

The question is:

Find all the possible real ordered pairs of (x,y) for equation
16^[(x^2) + y] + 16^[x + (y^2)]=1


2. Relevant equations

That was the only equation.

3. The attempt at a solution

Using 0.5 + 0.5 =1, (x^2) + y = -0.25 and (y^2) + x= -0.25
(Because [16^(-0.25)]=0.5 , I'm using decimals here instead of fractions)

Solving them gives one real ordered pair of (-0.5,-0.5). But how do I know if that's the only answer?
Phys.Org News Partner Science news on Phys.org
NASA's space station Robonaut finally getting legs
Free the seed: OSSI nurtures growing plants without patent barriers
Going nuts? Turkey looks to pistachios to heat new eco-city
cng99
cng99 is offline
#2
Mar8-12, 11:54 PM
P: 44
This is my first thread ever, by the way. Tell me if I did something wrong.
epenguin
epenguin is offline
#3
Mar9-12, 06:03 PM
HW Helper
epenguin's Avatar
P: 1,932
Quote Quote by cng99 View Post
Solving them gives one real ordered pair of (-0.5,-0.5). But how do I know if that's the only answer?
Have you checked that that is a solution?

epenguin
epenguin is offline
#4
Mar10-12, 01:20 AM
HW Helper
epenguin's Avatar
P: 1,932

Equation with x,y as exponential functions


Was up till 3 a.m. here last night for a family arrival, and wrote last post just before going to bed. Had thought to write 'is it plausible that the answer is one or a few pairs of numbers like that?'.

Woke about 7, and in a bit saw what you are supposed to do. It starts looking nice but at the end the algebra got looking ugly.

Then realised you are supposed to describe answer geometrically, not algebraically.

Have not needed to write anything down.

Wondering whether to go back to bed or have breakfast.
cng99
cng99 is offline
#5
Mar11-12, 12:01 AM
P: 44
Quote Quote by epenguin View Post
Have you checked that that is a solution?
(-0.5,-0.5) fits in perfectly. But the question is to find all possible real ordered pairs. How do I do that?
cng99
cng99 is offline
#6
Mar11-12, 12:03 AM
P: 44
Quote Quote by epenguin View Post
Then realised you are supposed to describe answer geometrically, not algebraically.
What do you mean???
NascentOxygen
NascentOxygen is offline
#7
Mar11-12, 12:19 AM
HW Helper
P: 4,715
I tried over two days to see what wolframapha would make of this, but it's not functioning. Anyone know whether the free version has been decommissioned?
cng99
cng99 is offline
#8
Mar11-12, 12:57 AM
P: 44
Quote Quote by NascentOxygen View Post
I tried over two days to see what wolframapha would make of this, but it's not functioning. Anyone know whether the free version has been decommissioned?
I've tried that already. It would just do a differentiation. The problem can only be solved by a human.

(It might be possible to plot a graph of 16^((x^2) + y) + 16^((y^2) + x) = z, and then intersecting the surface with z=1. But I couldn't find anything that would plot something like that. They all accept y as a function of x. This doesn't really get us anywhere. )

Is there anyway to find the minimum or maximum value of 16^((x^2) + y) + 16^((y^2) + x)?
That might help. Possible ways to do so might be using the Arithmetic Mean >= Geometric mean >= Harmonic mean property. Or maybe by just differentiating it and putting dy/dx=0. But I couldn't find anything there. Do tell me if anyone finds anything.
Mentallic
Mentallic is offline
#9
Mar11-12, 01:15 AM
HW Helper
P: 3,436
I'm having quite a lot of trouble with this problem, it's really got me stumped. But then again, all the Olympiad questions seem to give me a hard time

Since we have an exponential of the form [itex]a^b+a^c=1[/itex] and we know that [itex]a^b>0, a^c>0[/itex] then what we must have is that [itex]b<0, c<0[/itex]

So

[tex]x^2+y<0[/tex]
[tex]x+y^2<0[/tex]
And this gives us the inequality
[tex]-1<x<0[/tex]
[tex]-\sqrt{-x}<y<-x^2[/tex]
But from here I can't think of a way to further narrow down our possible solution set.
cng99
cng99 is offline
#10
Mar11-12, 03:20 AM
P: 44
Hey, Mentallic. Nice try...

This did help a bit...

http://www.wolframalpha.com/input/?i...B+-1%3C+y+%3C0
cng99
cng99 is offline
#11
Mar11-12, 03:27 AM
P: 44
Actually, However, that doesn't narrow down to it.
epenguin
epenguin is offline
#12
Mar11-12, 04:27 AM
HW Helper
epenguin's Avatar
P: 1,932
Quote Quote by epenguin View Post
Then realised you are supposed to describe answer geometrically, not algebraically.
Quote Quote by cng99 View Post
What do you mean???
I mean - that I had misread the question (missed the ^)
NascentOxygen
NascentOxygen is offline
#13
Mar11-12, 07:17 AM
HW Helper
P: 4,715
Quote Quote by cng99 View Post
You get some sort of result when you click on this link, do you? As I said previously, wolframalpha seems non-functioning to me. I clicked on your link and for 10 mins watched it spinning endlessly on "Computing ..."

Given this, and in light of my inability to get it to draw even a straight-line graph, I think the owners must have made it regionally selective, so it's ignoring certain IP ranges. When I go to http://www.wolframalpha.com I am presented with a totally blank page. Anyone else?

Maybe this facility comes under the umbrella of "National security ..........?"
cng99
cng99 is offline
#14
Mar11-12, 10:47 AM
P: 44
Quote Quote by NascentOxygen View Post
Maybe this facility comes under the umbrella of "National security ..........?"

Haha! Probably.
cng99
cng99 is offline
#15
Mar14-12, 04:26 AM
P: 44
Hmmmm, no solutions then?

(-0.5,-0.5) only?
epenguin
epenguin is offline
#16
Mar15-12, 05:47 AM
HW Helper
epenguin's Avatar
P: 1,932
Quote Quote by cng99 View Post
Hmmmm, no solutions then?

(-0.5,-0.5) only?
Can hardly be. One equation relating two variables is a continuous curve in general. The present curve must be symmetrical about the 'diagonal' line x = y and go through that point. The curve might have more than one branch but if so no other branch cuts that line. Don't know what else they want you to say about it.

You might therefore try rotating the thing counter clockwise by 45 , then it is symmetrical around the new X axis f(X) = f(-X) and see if that inspires anything.
cng99
cng99 is offline
#17
Mar15-12, 07:42 AM
P: 44
Well, if it's a curve, why wouldn't anyone plot it?
Wolfram Alpha plots a curve for 16^[(x^2) + y] + 16^[x + (y^2)]= anything more than 1.

Maybe whatever the curve is, it gets reduced to a point when 16^[(x^2) + y] + 16^[x + (y^2)]= 1
epenguin
epenguin is offline
#18
Mar15-12, 09:49 AM
HW Helper
epenguin's Avatar
P: 1,932
Quote Quote by cng99 View Post
Well, if it's a curve, why wouldn't anyone plot it?
Wolfram Alpha plots a curve for 16^[(x^2) + y] + 16^[x + (y^2)]= anything more than 1.

Maybe whatever the curve is, it gets reduced to a point when 16^[(x^2) + y] + 16^[x + (y^2)]= 1
Can you show us what the family looks like, so we see what happens as you reduce this something to 1?


Register to reply

Related Discussions
Exponential functions - converting a linear equation Calculus & Beyond Homework 5
Advanced Functions,Exponential Functions Precalculus Mathematics Homework 9
Advanced Functions,Exponential Functions Precalculus Mathematics Homework 4
Exponential Functions Precalculus Mathematics Homework 2
exponential functions Calculus & Beyond Homework 5