Order axioms and x/y > 1 (For all x, y, z ∈ ℝ, if x<y and 0<z, then xz<yz)

Four Muffins

New member
Joined
Jun 29, 2022
Messages
25
Hello.

I was working on a problem and made a mistake, and I'm trying to understand exactly what it was. I didn't realise when I was writing the solution that
x/y > 1 ---> x > y
doesn't work when x and y are negative.

Have I not followed the axiom
(3d) For all x, y, z ∈ ℝ, if x<y and 0<z, then xz<yz
from here, and that by using x and y with a -1 factor, I haven't flipped the inequality sign? Or is it something else, like a general math thing that axioms aren't needed for, and I just have to notice it doesn't work and qualify the statement with 'for x, y ∈ ℝ+'? Or is it just not valid to multiply both sides of the inequality by y?


20221228_130207.jpg
 
Hello.

I was working on a problem and made a mistake, and I'm trying to understand exactly what it was. I didn't realise when I was writing the solution that
x/y > 1 ---> x > y
doesn't work when x and y are negative.

Have I not followed the axiom
(3d) For all x, y, z ∈ ℝ, if x<y and 0<z, then xz<yz
from here, and that by using x and y with a -1 factor, I haven't flipped the inequality sign? Or is it something else, like a general math thing that axioms aren't needed for, and I just have to notice it doesn't work and qualify the statement with 'for x, y ∈ ℝ+'? Or is it just not valid to multiply both sides of the inequality by y?


20221228_130207.jpg
Please show us the problem you are working on, which is not clearly stated here; and tell us why you think you are wrong.

When you say "Are x and y positive", is that a question, or are you trying to say "Suppose x and y are positive"?

Is it given that 2x - 2y = 1 and also that x/y > 1?

What is your goal? What was your conclusion that you think is wrong?

If you know that x and y are positive, and that x/y > 1, then you can multiply by y to conclude that x > y, because the axiom says you can. But if y can be negative, then you can't.
 
I beg your pardon, I should have been clearer. That is the question, here's the source.

1672204655604.png

And the solution is here: https://gmatclub.com/forum/inequali...ions-from-my-collection-86939-20.html#p653709
1672204695853.png

I was wondering if there was a particular rule or something that I violated, since using simple algebraic manipulation on x/y > 1 led me astray, with me not noticing that multiplying both sides of x/y > 1 by y doesn't work if x and y are both negative.

(Presumably 'Answer: C.' in the solution refers to an answer sheet I haven't found.)

Edit: missed words
 
if x and y are both negative
Hi Four Muffins. If x and y are both negative, then they each contain a factor of -1, which cancel. In other words, you end up with |x|/|y| which is the same as x and y both positive. Yet, the case where x and y are both negative doesn't work in this exercise because y<x yields |y|>|x| after the negations cancel in the ratio x/y.

Your example: x = -1 and y = -3/2

\(\displaystyle \frac{x}{y} = \frac{\cancel{(\text{-}1)}(1)}{\cancel{(\text{-}1)}(\frac{3}{2})} = \frac{2}{3}\)

That's the same case as x = 1 and y = 3/2, which is not a valid (x,y) pair because we know that y must be less than x to make x/y greater than 1.



We can confirm graphically that x and y must both be positive, whenever both Statement1 and Statement2 are true.

Statement1 tells us that y is the same as x – ½

Substitute that expression for y in Statement2:

x/(x – 1/2) > 1

Here's a graph of each side of that inequality. NOTE that the vertical axis is not called the y-axis in this plot; the vertical axis measures the value of ratio x/y, not the value of y.

The left-hand side is plotted in green and the right-hand side in yellow. We see that the green curve shows x/y greater than 1 (i.e., the green curve appears above the yellow line) only when x is greater than ½ (line x=½ shown in red). That means y must be positive, also, because y is the result of subtracting ½ from a number that's bigger than ½. :)

xy-ratio.PNG

Alternatively, by taking cases, we can solve the inequality x/(x–½)>1 algebraically and find that it's true for all x>½.
[imath]\;[/imath]
 
I beg your pardon, I should have been clearer. That is the question, here's the source.

View attachment 34695

And the solution is here: https://gmatclub.com/forum/inequali...ions-from-my-collection-86939-20.html#p653709
View attachment 34696

I was wondering if there was a particular rule or something that I violated, since using simple algebraic manipulation on x/y > 1 led me astray, with me not noticing that multiplying both sides of x/y > 1 by y doesn't work if x and y are both negative.

(Presumably 'Answer: C.' in the solution refers to an answer sheet I haven't found.)

Edit: missed words
Apparently, the choices given are something like:

A: (1) alone is sufficient to answer​
B: (2) alone is sufficient to answer​
C: (1) and (2) together are sufficient to answer​
D: There is not enough information to know​

Omitting this was a major error in presenting the problem. I would not trust this person.

As for your work, you are correct that [imath]\frac{x}{y}>0[/imath] does not imply that [imath]x>y[/imath], so that is a mistake.

One alternative method to theirs would be to take two cases, according to the sign of y, allowing you to do your multiplication and know the result.
 
If x-y>0 (as in x-y=1/2), then surely x>y.
The proof is trivial. In x-y>0, just add y to both sides to get x>y.

Now if x/y>1 we have two cases to consider.
Case 1: x and y are both positive.
If x/y >1, multiplying both sides by y, we get x>y.

Case 2: x and y are both negative.
In x/y>1, multiply both sides by y and get x<y
OR multiply x/y by -1/-1 to get -x/-y>1. Now multiply both sides by -y and get -x>-y. Multiplying both sides by -1 yields that x<y
 
Top