Another Good Number?

True or False?

If a n b n a^n - b^n is an integer for all natural n n and distinct real numbers a , b a,b , then a a and b b must be integers.

Clarification: a n b n a^n - b^n must be an integer for all natural values of n n , not just a few values of n n .

True False

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.

3 solutions

Lemma 1:- If l p n k l^{p_n}k is an integer for a monotonically increasing sequence < p n > <p_n> with p 1 = 1 p_1 = 1 and a non-zero integer k k , then l l is an integer.

Proof :- l l can't be irrational, otherwise it would imply l k lk is irrational, contradiction. Then, for the sake of contradiction, suppose l l is a non integral rational number p q \frac{p}{q} where p , q p,q are integers, q 1 q \not = 1 and gcd ( p , q ) = 1 \gcd(p,q) =1 . Then l p n k = p p n k q p n l^{p_n}k = \frac{p^{p_n}k}{q^{p_n}} . Since gcd ( p n , q n ) = gcd ( p , q ) = 1 \gcd(p^n,q^n) =\gcd(p,q) =1 , we have q p n k q p n k q^{p_n} | k \implies q^{p_n} \leq k . But this is not possible as q p n q^{p_n} is increasing. Therefore, our suppostion is wrong and l l is an integer.

Now, ( a n b n ) 2 + a 2 n b 2 n = 2 a n ( a n b n ) (a^n - b^n)^2 + a^{2n} - b^{2n} = 2a^n(a^n-b^n) implies 2 a n ( a n b n ) 2a^n(a^n-b^n) is an integer.

Also, [ ( 2 a n ) k ( a n b n ) ] 2 + 2 k [ ( 2 a 2 n ) k ( a 2 n b 2 n ) ] = ( 2 a n ) 2 k + 1 ( a n b n ) [(2a^n)^k(a^n-b^n)]^2 + 2^k[(2a^{2n})^k(a^{2n}-b^{2n})] = (2a^n)^{2k +1}(a^n - b^n) implies that if ( 2 a n ) k ( a n b n ) (2a^n)^k(a^n - b^n) is an integer, then ( 2 a n ) 2 k + 1 ( a n b n ) (2a^n)^{2k +1}(a^n - b^n) is an integer for natural n n . Since the case k = 1 k =1 is true, we can conclude that ( 2 a n ) t n ( a n b n ) (2a^n)^{t_n}(a^n - b^n) is an integer for a monotonically increasing sequence < t n > <t_n> with t 1 = 1 t_1 =1

Then, by Lemma 1, 2 a n 2a^n must be an integer for all natural numbers n n . Applying Lemma 1 again,we see a a must be an integer.

Now since a b a -b and a a both are integers, b b must also be an integer. Hence, if a n b n a^n - b^n is an integer for all natural n n , then a a and b b must be integers.

Very nice, but in Lemma 1 you should mention that k 0 k \neq 0 .

Ariel Gershon - 6 years, 3 months ago

Log in to reply

Thanks. I've fixed it.

Siddhartha Srivastava - 6 years, 3 months ago

Log in to reply

What if a and b are not integers and a=b.?

Archit Boobna - 6 years, 3 months ago

Log in to reply

@Archit Boobna

distinct real numbers a , b , a,b, .

That will result in k = 0 k = 0 . To avoid this, I put "distinct" in the question.

Siddhartha Srivastava - 6 years, 3 months ago
Linkin Duck
Mar 3, 2015

a-b is an integer, then a=x+m, b=y+m, with x, y are integer and 0<=m<=1

a^{2} - b^{2} = (a-b) \times (a+b) is an integer => a+b is an integer also => m=0.5 or 0

a^{3} - b^{3} = (a-b) \times ( a^{2} + b^{2} + a \times b ) is integer also => m must be 0 => a and b must be integer

x y = z xy = z where x x and z z are integers does not imply that y y is also an integer. It may also be a rational number. For example, 4 1 2 = 2 4*\frac{1}{2} = 2 does not mean that 1 / 2 1/2 is an integer.

Siddhartha Srivastava - 6 years, 3 months ago

I do not quite agree for instance taking the Binet formula for Fibonacci numbers and we split into two separate terms

a^n = [(1+Sqrt(5)/2) ^n ]/ Sqrt(5)

b^n = [(1-Sqrt(5)/2) ^n ]/ Sqrt(5)

If we substact a^n -b^n we get the Fn which is integer but a an b are not.

Mariano PerezdelaCruz - 6 years, 3 months ago

Log in to reply

What would a a and b b be in this instance?

Note that the denominator 5 \sqrt{5} doesn't have a exponent of n n in it.

Siddhartha Srivastava - 6 years, 3 months ago

Log in to reply

I believe you can fix it by writing in the Binet formula [ Sqrt(5) ]^(1/n) and just plug it as the denominator inside the bracket, by making n=1 and we would get an and bn

Mariano PerezdelaCruz - 6 years, 3 months ago

Log in to reply

@Mariano PerezdelaCruz Then it wouldn't work for all n n . We need it to work for all n n

Siddhartha Srivastava - 6 years, 3 months ago

(2.5)^2 -(1.5)^2 =6.25-2.25 =4 n=2 which is a natural number. a (2.5) is not equal to b (1.5). but 4 is an integer!!!!!!!!!!!

ARUNEEK BISWAS - 6 years, 3 months ago

Log in to reply

The question states that a n b n a^n - b^n must be an integer for all natural n n , not just one value of n n .

Siddhartha Srivastava - 6 years, 3 months ago

For example if we take a and b as 3.5 and 4.5 and n=2 then answer will be an integer but a and b are not integers , so it is a ' may' condition and not 'must'.

Ritvik Arya - 6 years, 3 months ago

Log in to reply

The question states that a n b n a^n - b^n must be an integer for all natural n n , not just one value of n n .

Siddhartha Srivastava - 6 years, 3 months ago

Log in to reply

In response to @Siddhartha Srivastava this would agree for n=2,4,16 Etc

Ritvik Arya - 6 years, 3 months ago

Log in to reply

@Ritvik Arya But it won't work for n = 3 n = 3 . It has to work for all natural n n . Not just a few specified values.

Siddhartha Srivastava - 6 years, 3 months ago

Log in to reply

@Siddhartha Srivastava If one value satisfies then it is not must it is may

Ritvik Arya - 6 years, 3 months ago

Log in to reply

@Ritvik Arya The question is that if a n b n a^n - b^n is and integer for all n n , then a a and b b must be integers.

If the initial statement (if a n b n a^n - b^n is and integer for all n n ) is not true, as in the example you gave, then you can't say anything about the then statement.

You reasoning will only be correct if there exists some a , b a,b such that a n b n a^n - b^n is an integer for all n n

Siddhartha Srivastava - 6 years, 3 months ago
Frank Rodriguez
Mar 10, 2015

Just a comment. Are we saying that sqrt (2) is not a distinct real number? Because I was thinking (sqrt (3))^2-(sqrt(2))^2=1. I think we should use the term "rational" instead of "distinct".

In context, the word "distinct" meant that a b a \not = b , i.e. a a and b b are "distinct".

Your counter example only works for n = 2 n =2 . It obviously does not work for n = 1 n = 1 and all other odd numbers.

Siddhartha Srivastava - 6 years, 3 months ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...