Am I right or wrong?

Algebra Level 5

One day I was trying to prove the Cauchy Schwarz inequality using the rearrangement inequality. However, at last I succeeded in proving it. Assume that a 1 , a 2 , a n a_1,a_2,\dots a_n and b 1 , b 2 , b n b_1,b_2, \dots b_n are sequences of real numbers. Here are my steps:

Step 1: Let A = i = 1 n a i 2 0 , B = i = 1 n b i 2 0 A=\displaystyle\sum_{i=1}^n a_i^2 \neq 0 \ , \ B=\displaystyle\sum_{i=1}^n b_i^2 \neq 0

Step 2: 2 = 1 + 1 \ 2 = 1+1

Step 3: 2 = i = 1 n a i 2 A + i = 1 n b i 2 B \ 2=\dfrac{\displaystyle\sum_{i=1}^n a_i^2}{A} + \dfrac{\displaystyle\sum_{i=1}^n b_i^2}{B}

Step 4: 2 = a 1 2 A + a 2 2 A + a n 2 A + b 1 2 B + b 2 2 B + b n 2 B \ 2 = \dfrac{a_1^2}{A} +\dfrac{a_2^2}{A} + \dots \dfrac{a_n^2}{A} +\dfrac{b_1^2}{B} +\dfrac{b_2^2}{B} + \dots \dfrac{b_n^2}{B}

Step 5: So by rearrangement inequality, we have :

2 a 1 b 1 A B + a 2 b 2 A B + + a n b n A B + b 1 a 1 A B + b 2 a 2 A B + + b n a n A B 2 \geq \dfrac{a_1b_1}{\sqrt{AB}}+\dfrac{a_2b_2}{\sqrt{AB}} + \dots + \dfrac{a_nb_n}{\sqrt{AB}}+\dfrac{b_1a_1}{\sqrt{AB}}+\dfrac{b_2a_2}{\sqrt{AB}}+ \dots +\dfrac{b_na_n}{\sqrt{AB}}

Step 6: 2 A B 2 ( a 1 b 1 + a 2 b 2 + + a n b n ) \ 2\sqrt{AB} \geq 2(a_1b_1+a_2b_2 + \dots + a_nb_n)

Step 7: A B ( a 1 b 1 + a 2 b 2 + + a n b n ) \ \sqrt{AB} \geq (a_1b_1+a_2b_2 + \dots + a_nb_n)

Step 8: Squaring both sides, we have:

A B ( a 1 b 1 + a 2 b 2 a n b n ) 2 AB \geq (a_1b_1+a_2b_2 \dots a_nb_n)^2

Step 9: By replacing A,B we restore the form and we get the famous Cauchy Schwarz Inequality:

( i = 1 n a i 2 ) ( i = 1 n b i 2 ) ( i = 1 n a i b i ) 2 \left(\displaystyle\sum_{i=1}^n a_i^2\right)\left(\displaystyle\sum_{i=1}^n b_i^2\right) \geq \left(\displaystyle\sum_{i=1}^n a_ib_i \right)^2

If you think I made a mistake in the proof, click the step number where I made a mistake. If you think that my proof is valid, click 0 0 .

7 0 4 8 1 6 5 3

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.

1 solution

Nihar Mahajan
May 30, 2015

In step 8 , we cannot square until we guarantee that a 1 b 1 + a 2 b 2 + a n b n a_1b_1+a_2b_2 + \dots a_nb_n is always non negative.Hence the step 8 is wrong.

So how did I complete the proof? I used a suitable trick.Firstly we have that

A B ( a 1 b 1 + a 2 b 2 + a n b n ) ( 1 ) \sqrt{AB} \geq (a_1b_1 + a_2b_2 + \dots a_nb_n) \dots (1)

Then I replaced all the a i a_i with a i -a_i to get the following :

A B ( a 1 b 1 + a 2 b 2 + a n b n ) ( 2 ) \sqrt{AB} \geq -(a_1b_1 + a_2b_2 + \dots a_nb_n) \dots (2)

Using ( 1 ) , ( 2 ) (1),(2) , we have

A B a 1 b 1 + a 2 b 2 + a n b n \sqrt{AB} \geq |a_1b_1 + a_2b_2 + \dots a_nb_n|

And only then we can square both the sides to get :

A B ( a 1 b 1 + a 2 b 2 + a n b n ) 2 AB \geq (a_1b_1 + a_2b_2 + \dots a_nb_n)^2

And the proof becomes valid.

Cheers!

@Calvin Lin Can I write a note on this problem having a title "Proof of Cauchy Schwarz Inequality using Rearrangement inequality" ?

Nihar Mahajan - 6 years ago

Log in to reply

@Calvin Lin Sir , I am waiting for your reply.

Nihar Mahajan - 6 years ago

Log in to reply

Sure!

Why not? I was wondering why you were asking.

Calvin Lin Staff - 6 years ago

Log in to reply

@Calvin Lin Sir , I thought If I wrote a note , then some people will just read it and get this question correct. Hence I preferred asking you. Anyways I will soon post a note. Thanks!

Nihar Mahajan - 6 years ago

Log in to reply

@Nihar Mahajan Ah I see.

It will likely be best to add this as a (corrected) proof to the CS wiki directly, as one of the various ways of proving CS.

Calvin Lin Staff - 6 years ago

I don't really understand. Why should we guarantee if it is non-negative? Even if it is negative, we get a positive answer, right?

I would really appreciate it if someone could explain it to me.

Thank you.

Muathasim Mohamed - 6 years ago

Log in to reply

You see, 4>-9 but (4)^2 < (-9)^2

Abhijeet Verma - 6 years ago

Log in to reply

Thank you very much.

Muathasim Mohamed - 6 years ago

See Abhijeet's example. It explains the whole thing.

Edited

Nihar Mahajan - 6 years ago

Log in to reply

Thank you very much.

Muathasim Mohamed - 6 years ago

really, I too didn't get the logic. Had it been negative, then also we could have got the proof.

Raushan Sharma - 6 years ago

Log in to reply

See Abhijeet's example. It explains the whole thing.

Nihar Mahajan - 6 years ago

Log in to reply

@Nihar Mahajan Ok, got it

Raushan Sharma - 6 years ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...