The Student Room Group

Proof by Induction - Proving divisibility

Hi guys,
I'm struggling a little bit understand how to prove, by induction, that something is divisible by xx

The question I'm having trouble is:

Prove by induction that, for all positive integers n,
23n12^{3n}-1 is divisible exactly by 77

Spoiler



Any help is greatly appreciated
Thanks in advance
Original post by chapmouse
Hi guys,
I'm struggling a little bit understand how to prove, by induction, that something is divisible by xx

The question I'm having trouble is:

Prove by induction that, for all positive integers n,
23n12^{3n}-1 is divisible exactly by 77

Spoiler



Any help is greatly appreciated
Thanks in advance

Consider (23(k+1)1)(23k1)(2^{3(k+1)}-1) - (2^{3k} -1). If you can show this is divisible by 7, you're done. (why?)
Reply 2
Original post by chapmouse
Hi guys,
I'm struggling a little bit understand how to prove, by induction, that something is divisible by xx

The question I'm having trouble is:

Prove by induction that, for all positive integers n,
23n12^{3n}-1 is divisible exactly by 77

Spoiler



Any help is greatly appreciated
Thanks in advance


The whole reason you test values for n=1, n=2, or w/e, depending on the question, is just to check that it holds for 'any value of n'. Thus we can ASSUME its also true for n=k, and n=k+1, and so on. So you dont need to prove its true for n=k+1, we are already assuming it is. So like Farhan said;

(23(k+1)1)(2^{3(k+1)}-1)

by subbing n=k+1 in.

The next step to proving divisiblity, is showing that (n=k+1)(n=k)(n=k+1)-(n=k)

is divisible by 7, so take Farhan's equation and work with it:

(23(k+1)1)(23k1)(2^{3(k+1)}-1) - (2^{3k} -1)

Remember, we are assuming n=k is divisible by 7, and we are also assuming n=k+1 is, so the difference between them must also be divisible by 7, that is what we are showing. Eg, 49-7=42. 42/7=6.
Reply 3
Original post by Farhan.Hanif93
Consider (23(k+1)1)(23k1)(2^{3(k+1)}-1) - (2^{3k} -1). If you can show this is divisible by 7, you're done. (why?)

Why are you subtracting? I thought I had to add the n=1n=1 term to the n=kn=k term, both of which I already had?
Reply 4
Original post by chapmouse
Why are you subtracting? I thought I had to add the n=1n=1 term to the n=kn=k term, both of which I already had?


f(k+1)f(k)+f(1)f(k+1)\neq f(k)+f(1) Check my post.
(edited 10 years ago)
Reply 5
Original post by Phichi
f(k+1)f(k)+f(1)f(k+1)\neq f(k)+f(1) Check my post.

I don't understand why this is true. I have been taught to assume true for n=kn=k and prove that n=k+1n=k+1 works. It works with sequences, series, and matrices proofs. I'm doing this question from a specimen paper, where the proof questions aren't provided with a solution, which is the only thing i'm struggling with...
Original post by chapmouse
I don't understand why this is true. I have been taught to assume true for n=kn=k and prove that n=k+1n=k+1 works. It works with sequences, series, and matrices proofs. I'm doing this question from a specimen paper, where the proof questions aren't provided with a solution, which is the only thing i'm struggling with...

Say if you have a function f(x). If f(x) is divisible by 7 for some x (just because it's from your question, it could be any number, say, p) and the difference between successive values i.e. f(x+1) - f(x) is also divisible by 7 (or p) then f(x) is divisible by 7 for all x
(edited 10 years ago)
Reply 7
Original post by chapmouse
I don't understand why this is true. I have been taught to assume true for n=kn=k and prove that n=k+1n=k+1 works. It works with sequences, series, and matrices proofs. I'm doing this question from a specimen paper, where the proof questions aren't provided with a solution, which is the only thing i'm struggling with...


Because its hard for re-arrange n=k+1, to show its divisible by 7. Thus you use the method of n=k+1 - n=k, to show that the difference is divisible by 7, its just procedure.
(edited 10 years ago)
Reply 8
Original post by chapmouse
I don't understand why this is true. I have been taught to assume true for n=kn=k and prove that n=k+1n=k+1 works. It works with sequences, series, and matrices proofs. I'm doing this question from a specimen paper, where the proof questions aren't provided with a solution, which is the only thing i'm struggling with...


Consider the function f(x)=x2f(x) = x^{2}.

f(k+1)=(k+1)2f(k + 1) = (k + 1)^{2}

f(k)=k2f(k) = k^{2}

f(1)=12=1f(1) = 1^{2} = 1

I hope you can see that (k+1)2k2+1(k + 1)^{2} \not= k^{2} + 1.

Hence f(k+1)f(k)+f(1)f(k + 1) \not= f(k) + f(1)
Reply 9
Original post by Joshmeid
Consider the function f(x)=x2f(x) = x^{2}.

f(k+1)=(k+1)2f(k + 1) = (k + 1)^{2}

f(k)=k2f(k) = k^{2}

f(1)=12=1f(1) = 1^{2} = 1

I hope you can see that (k+1)2k2+1(k + 1)^{2} \not= k^{2} + 1.

Hence f(k+1)f(k)+f(1)f(k + 1) \not= f(k) + f(1)


but is f(k)=f(k+1)f(1)f(k) = f(k+1) - f(1) ?
Reply 10
Original post by chapmouse
but is f(k)=f(k+1)f(1)f(k) = f(k+1) - f(1) ?


No

f(k)=k2f(k) = k^{2}

f(k+1)f(1)f(k + 1) - f(1)
=(k+1)212= (k + 1)^{2} - 1^{2}
=k2+2kk2= k^{2} + 2k \not= k^{2}
(edited 10 years ago)
Reply 11
Original post by chapmouse
but is f(k)=f(k+1)f(1)f(k) = f(k+1) - f(1) ?


That is the same as implying f(k)+f(1)=f(k+1)f(k) +f(1) = f(k+1)
Original post by chapmouse
but is f(k)=f(k+1)f(1)f(k) = f(k+1) - f(1) ?

f(k) is the expression you want to prove the proposition for in the question taking the value for some arbitrary integer k. Same goes for f(k+1) - check my last post, if you can show f(k+1) - f(k) is divisible by 7, then combined with your basis case, you're done
Reply 13
You need to understand, the x values you are subbing in, are not the values of your function. Sure if f(x)=x then f(k)+f(1)=f(k+1), but in any other case it doesn't, unless unique.
Reply 14
Original post by Phichi
The whole reason you test values for n=1, n=2, or w/e, depending on the question, is just to check that it holds for 'any value of n'. Thus we can ASSUME its also true for n=k, and n=k+1, and so on. So you dont need to prove its true for n=k+1, we are already assuming it is. So like Farhan said;

(23(k+1)1)(2^{3(k+1)}-1)

by subbing n=k+1 in.

The next step to proving divisiblity, is showing that (n=k+1)(n=k)(n=k+1)-(n=k)

is divisible by 7, so take Farhan's equation and work with it:

(23(k+1)1)(23k1)(2^{3(k+1)}-1) - (2^{3k} -1)

Remember, we are assuming n=k is divisible by 7, and we are also assuming n=k+1 is, so the difference between them must also be divisible by 7, that is what we are showing. Eg, 49-7=42. 42/7=6.



Just follow this method, that I described before, for all divisibility questions, and you'll be fine. If you need help on the algebra, just ask.

f(k+1)-f(k), and prove its divisible by 7.
Original post by Phichi
Because its impossible for re-arrange n=k+1 to show its divisible by 7


Not true

Original post by chapmouse
...


You might've learned it a different way, so consider this

f(k+1)=23(k+1)1=23k+31=823k1=23k1+723kf(k+1) = 2^{3(k+1)} - 1 = 2^{3k+3} - 1 = 8 \cdot 2^{3k} - 1 = 2^{3k} - 1 + 7 \cdot 2^{3k}

You can take it from there :wink:
Reply 16
Original post by Felix Felicis
Not true



You might've learned it a different way, so consider this

f(k+1)=23(k+1)1=23k+31=823k1=23k1+723kf(k+1) = 2^{3(k+1)} - 1 = 2^{3k+3} - 1 = 8 \cdot 2^{3k} - 1 = 2^{3k} - 1 + 7 \cdot 2^{3k}

You can take it from there :wink:

This is exactly what I was looking for! thanks
Original post by chapmouse
This is exactly what I was looking for! thanks

Np :biggrin: But I'd make note of the f(k+1) - f(k) method as well, it could come in handy sometimes :wink:

Edit, also if you're interested, the problem doesn't require induction :tongue:

Spoiler

(edited 10 years ago)

Quick Reply

Latest