Basic Induct Q (1 Viewer)

Estel

Tutor
Joined
Nov 12, 2003
Messages
1,261
Gender
Male
HSC
2005
For a Q such as proving n^3+2n is divisible by 12 for even n, do you have to prove it for negative n as well? And if so, how?
 

CM_Tutor

Moderator
Moderator
Joined
Mar 11, 2004
Messages
2,644
Gender
Male
HSC
N/A
Usually, no, as a question would usually say 'for positive even integers n', or something similar.

Note, if this question did not say to use induction, this can be proven in about 3 - 4 lines. :)
 

Estel

Tutor
Joined
Nov 12, 2003
Messages
1,261
Gender
Male
HSC
2005
I'm referring to the cambridge questions in the Yr11 volume... they don't indicate 'for positive even integers n'...

But I'll take it that they just mean positive integers.
Thanks :)
 

CM_Tutor

Moderator
Moderator
Joined
Mar 11, 2004
Messages
2,644
Gender
Male
HSC
N/A
Originally posted by hatty
how so CM_Tutor?
Let n = 2k, where k is a positive integer:
Then n<sup>3</sup> + 2n = (2k)<sup>3</sup> + 2(2k) = 8k<sup>3</sup> + 4k = 4k(2k<sup>2</sup> + 1), which is obviously divisible by 4.
We now need only prove that k(2k<sup>2</sup> + 1) is divisible by 3: k(2k<sup>2</sup> + 1) = k(2k<sup>2</sup> - 2 + 3) = 2k(k + 1)(k - 1) + 3k
Both of these terms are divisible by 3. So, for even n, n<sup>3</sup> + 2n is divisible by both 4 & 3, and thus is divisible by 12.
 

CM_Tutor

Moderator
Moderator
Joined
Mar 11, 2004
Messages
2,644
Gender
Male
HSC
N/A
Note - the above proof does not require k to be positive, and so it proves the result for all even n, including the negative cases.

Estel: if you need to prove the result for all even n, and wish to use induction, prove the result for all positive even n, and then take n = -k, where k is a positive even integer. It then follows that:

n<sup>3</sup> + 2n = (-k)<sup>3</sup> + 2(-k) = -k<sup>3</sup> - 2k = -(k<sup>3</sup> + 2k), which is divisible by 12 from the induction proof.

This completes the proof for positve and negative even integers.
 

Estel

Tutor
Joined
Nov 12, 2003
Messages
1,261
Gender
Male
HSC
2005
Can somebody double check I've done this right?

A
When n=2
n^3+2n=12
Which is divisible by 12

B
Suppose that k is a positive even integer for which the statement is true.
That is, suppose
k^3+2k=12m, where m is some integer
We prove the statement for n=k+2
That is, we prove
(k+2)^3+2(k+2) is divisible by 12
(k+2)^3+2(k+2)
=k^3+6k^2+14k+12
=12m+6k^2+12k+12, by the induction hypothesis
=12(m+k^2/2 +k+1), which is divisible by 12 as required (k is an even integer, therefore k^2>=4 and is even, and therefore k^2/2 is an integer)

C
From parts A and B by mathematical induction the statement is true for all positive even integers n.

Suppose n is a negative even number,
then n=-k
n^3 + 2n
=(-k)^3 + 2(-k)
=-k^3 - 2k
=-(k^3 + 2k), which is divisible by 12, as proven by induction above.

Therefore, n^3+2n is divisble by 12 for all even integers n.
 

CM_Tutor

Moderator
Moderator
Joined
Mar 11, 2004
Messages
2,644
Gender
Male
HSC
N/A
Seems fine to me, but in the last bit, I'd add after "then n = -k" the statement, "where k is a positive even integer". This is needed for you to use the induction above.

Note: this is a minor, fairly nit-picky point.
 

Estel

Tutor
Joined
Nov 12, 2003
Messages
1,261
Gender
Male
HSC
2005
One last question (sorry): if the question asks for you to prove it for n>=11 for example, without referring to n being an integer, do you just take it as a given? Or would you have n=k+x or something like that? And is 0 even?... well that's really 3 questions.

Very big thankyou. :D
 
Last edited:

CM_Tutor

Moderator
Moderator
Joined
Mar 11, 2004
Messages
2,644
Gender
Male
HSC
N/A
You can usually take n to be an integer, as it is usually implied in the question. Eg.

Show, using mathematical induction, that 1 + 2 + 3 + ... + n = n(n + 1) / 2, for n > 0.

This doesn't specifically say that n must be an integer, but it's implied by the fact that 'n' is given as the general term of a series with specific terms '1', '2', and '3'. Under these circumstances, n must be an integer.

Similarly, in the above question, 'even n' requires n to be an integer - after all, how can a non-integer be even?
 

Estel

Tutor
Joined
Nov 12, 2003
Messages
1,261
Gender
Male
HSC
2005
But when you're given an inequality to prove by mathematic induction, where it's just something like prove 2^n>3n^2 for n>2, I don't see how n being an integer is implied. I think I'm just confused.
 

CM_Tutor

Moderator
Moderator
Joined
Mar 11, 2004
Messages
2,644
Gender
Male
HSC
N/A
You're right, being an integer is not implied by 2<sup>n</sup> > 3n<sup>2</sup> for n > 2 (actually, this statement isn't true for integers until n > 7). However, induction works on the integers. To prove such a result, for all values n => 8, for example, is extremely complicated by induction. (It could be done by an Extn 2 student, using strong induction.) This would be much easier done by defining a function f(n) = 2<sup>n</sup> - 3n<sup>2</sup>, and showing that, for n => 8, f'(n) > 0, and that f(8) > 0. It would then follow that f(n) > 0 for n => 8.
 

Users Who Are Viewing This Thread (Users: 0, Guests: 1)

Top