Invariance Principle Question - from Arthur Engels book

• August 5th 2012, 06:28 PM
cosmonavt
Invariance Principle Question - from Arthur Engels book
Quoting from Arthur Engel's "Problem Solving Strategies":

Attachment 24433

To the part in red, how is this possible? If $x_{n}$ keeps increasing, $y_{n}$ has to decrease in order to keep the product of $x_{n}$ and $y_{n}$ constant at $ab$.

To the part in blue, I don't get this.

To the part in green, isn't harmonic and arithmetic mean equal in magnitude always?

Thanks. Also I don't get the inequality which he has generated. What is he trying to show by $x_{n+1} - y_{n+1}$?
• August 5th 2012, 07:12 PM
GJA
Re: Invariance Principle Question - from Arthur Engels book
Hi, cosmonavt.

If I'm reading this right, the red part is what he's proving with the blue and green lines, so it's not supposed to make perfect sense until those other parts are established.

The blue part is how you find the midpoint between two real numbers. If x and y are real numbers, then (x+y)/2 is the midpoint between x & y.

The green part comes from an inequality that is used in analysis frequently. The inequality states that for nonnegative numbers x and y:

$\sqrt{xy}\leq \frac{x+y}{2}$.

The left side of the inequality is known as the "Geometric mean" and the right hand side is known as the "Arithmetic mean."

Note: There is a more general statement of the above inequality. If you're curious you can find it here Inequality of arithmetic and geometric means - Wikipedia, the free encyclopedia

Does that clear things up? Let me know. Good luck!
• August 6th 2012, 02:24 PM
cosmonavt
Re: Invariance Principle Question - from Arthur Engels book
Thanks for the reply but it doesn't clear anything. FIrstly, I know that $x_{n+1}$ is halfway between $x_{n}$ and $y_{n}$ but how is it showing that $y_{n}$ is always less than $x_{n}$?

Secondly, he didn't say geometric mean, he said harmonic mean. Harmonic mean and geometric mean have the same magnitude, always.
• August 6th 2012, 02:56 PM
GJA
Re: Invariance Principle Question - from Arthur Engels book
The statement

Quote:

Harmonic mean and geometric mean have the same magnitude, always.
is incorrect. The relationship is actually that the harmonic mean is less than or equal to the geometric mean - see Theorem 2 of http://faculty.ccc.edu/mhidegkuti/sh...eans/means.pdf for a proof.

Second, you're right, he does quote the harmonic mean. But justifying the inequality while I read it, I used a method using the arithmetic/geometric mean inequality; I left the steps for you to see how this would work. If you would like a more explicit hint, I would suggest forming the ratio $\frac{y_{n+1}}{x_{n+1}}$ and seeing what you can do.