Mathematical Induction

CSE 191, Class Note 08

Computer Sci & Eng Dept
SUNY Buffalo

c Xin He (University at Buffalo) CSE 191 Descrete Structures 1 / 14

Climbing ladder

Suppose we have a ladder of n rungs. Let’s say we can guarantee two

things:
We can reach the first rung of the ladder.
If we can reach the ith rung of the ladder, then we can reach the
next (i.e., the (i + 1)st) rung.
What can we conclude, then?
We can conclude that we can reach the nth rung for any n.

c Xin He (University at Buffalo) CSE 191 Descrete Structures 3 / 14

Mathematical induction

Similar to the above argument, we have a proof method called mathematical

induction:
Goal: to prove P(n) is true (where n is a positive integer).
First step (called the basis step): show P(1) is true.
Second step (called the inductive step): show P(k) → P(k + 1) is true for
every positive integer k. Here P(k) is called the inductive assumption (or
inductive hypothesis).
Clearly, the above method makes sense because from

P(1), P(1) → P(2), P(2) → P(3), . . . , P(n − 1) → P(n)

we can easily get P(n).

c Xin He (University at Buffalo) CSE 191 Descrete Structures 4 / 14

First example

Example: Show that, for any positive integer n, 2n > n.

Proof: Basis step: When n = 1, we have 2n = 2 > 1 = n. So the proposition

is true for n = 1.
Inductive step: Assume that the proposition is true for n = k (where k is a
positive integer), i.e., 2k > k.
Now we prove that it is also true for n = k + 1, i.e., 2k+1 > k + 1.
From 2k > k we get that 2k+1 = 2 × 2k > 2 · k ≥ k + 1.
This completes the induction proof.

c Xin He (University at Buffalo) CSE 191 Descrete Structures 5 / 14

Understanding first example

In the first example, we have shown two things:

(a) 21 > 1;
(b) If 2k > k for positive integer k, then 2k+1 > k + 1.
Hence, we have the following statements being true:

(1) 21 > 1; (This is (a))

(2) If 21 > 1, then 22 > 2; (This is (b) when k = 1)
(3) If 22 > 2, then 23 > 3; (This is (b) when k = 2)
..
.
(n) If 2n−1 > n − 1, then 2n > n; (This is (b) when k = n − 1)

Putting all of them together, we see that 2n > n.

c Xin He (University at Buffalo) CSE 191 Descrete Structures 6 / 14

Second example

Example 2: Show that 3|n3 − n for positive integer n.

Proof: Basis step: When n = 1, we have n3 − n = 0. Clearly, 3|n3 − n.

Inductive step: Assume that 3|k3 − k for positive integer k. We’ll show that
3|(k + 1)3 − (k + 1).
It is easy to see (k + 1)3 − (k + 1) = k3 + 3k2 + 2k = (k3 − k) + 3(k2 + k).
Since 3|k3 − k, we can write k3 − k = 3j where j is an integer. So,

(k + 1)3 − (k + 1) = 3j + 3(k2 + k) = 3(j + k2 + k)

Hence, 3|(k + 1)3 − (k + 1).

c Xin He (University at Buffalo) CSE 191 Descrete Structures 7 / 14

Understanding second example

In the second example, we have shown two things:

(a) 3|13 − 1;
(b) If 3|k3 − k for positive integer k, then 3|(k + 1)3 − (k + 1).

Hence, we have the following statements being true:

(1) 3|13 − 1; (This is (a))
(2) If 3|13 − 1, then 3|23 − 2; (This is (b) when k = 1)
(3) If 3|23 − 2, then 3|33 − 3; (This is (b) when k = 2)
...
(n) If 3|(n − 1)3 − (n − 1), then 3|n3 − n; (This is (b) when k = n − 1)
Putting all of them together, we see that 3|n3 − n.

c Xin He (University at Buffalo) CSE 191 Descrete Structures 8 / 14

Variants of mathematical induction

In the mathematical induction we just studied, the constraint is that n is a

positive integer. In fact, we can have variants:
n is a non-negative integer;
or, n is a positive integer ≥ m.
To deal with the above situations, all we need is:
adjust the basis step, so that it considers n = 0 or n = m instead of
n = 1.
adjust the inductive step, so that P(k) → P(k + 1) is proved for all
non-negative integer k or all integer k ≥ m.

c Xin He (University at Buffalo) CSE 191 Descrete Structures 9 / 14

Example for variant

Example:
Suppose that, for a finite set S, |S| = n. Show that |P(S)| = 2n .

Note that we cannot consider n = 1 in the basis step! Because S could

be the empty set and thus n could be 0.
That means, we have to make sure the above statement is true for all
non-negative integer n (not just all positive integer n).
If we consider n = 1 in the basis step, then the entire proof ignores the
possibility of n = 0.
Similarly, when we do the inductive step, we cannot just prove it for all
positive integer k. We should prove it for all non-negative integer k.

c Xin He (University at Buffalo) CSE 191 Descrete Structures 10 / 14

Example for variant

Proof: Basis step: When n = 0, S is the empty set. Hence, P(S) = {∅},
which means |P(S)| = 1 = 20 .
Inductive step: Assume that, for all S such that |S| = k (where k s a
non-negative integer), |P(S)| = 2k .
Now we show that, for all S0 such that |S0 | = k + 1, |P(S0 )| = 2k+1 .
Clearly, all S0 such that |S0 | = k + 1 can be written as S0 = S ∪ {a}, where
|S| = k and a is not in S.
To count |P(S0 )|, i.e., the number of subsets of S0 , we only need to count:
(a) |P(S)|, i.e., the number of subsets of S;
By the inductive assumption, we know that |P(S)| = 2k .
(b) The number of subsets of S0 that contains a.
We note that each subset containing a uniquely corresponds to a subset not
containing a (by eliminating a from the subset).
Hence, this number is also |P(S)| = 2k .
We sum up these two numbers and get that |P(S0 )| = 2k + 2k = 2k+1 .
c Xin He (University at Buffalo) CSE 191 Descrete Structures 11 / 14
Strong induction
We have another important variant called strong induction:
Goal: to prove P(n) is true (where n is a positive integer).
Basis step: show P(1) is true.
Inductive step: show P(1) ∧ P(2) ∧ · · · ∧ P(k) → P(k + 1) is true for
every positive integer k.
Clearly, the above method makes sense because from P(1),
P(1) → P(2), P(1) ∧ P(2) → P(3), . . . , P(1) ∧ P(2) ∧ · · · ∧ P(n − 1) → P(n)
we can easily get P(n).

Example:
Show that any positive integer n > 1 can be written as the product of primes.

Note this is actually part of the fundamental theorem of arithmetic. Here we

prove it using strong induction.
Proof: Basis step: Here we consider n = 2 in stead of n = 1, because there
is a restriction n > 1.
When n = 2, since 2 is by itself a prime, the proposition is clearly true.
c Xin He (University at Buffalo) CSE 191 Descrete Structures 12 / 14

Example for strong induction

Inductive step: Assume every n such that 1 < n ≤ k (where k is an integer

> 1) can be written as the product of primes.
Now we show that k + 1 can also be written as the product of primes. We
consider two cases:
Case A: k + 1 is a prime. Then we are done.
Case B: k + 1 is a composite.

Then there exist positive integers a > 1 and b > 1 such that k + 1 = a · b.
Since a > 1, we know a ≥ 2, and thus b ≤ (k + 1)/2 < k.
By the inductive assumption, b can be written as the product of primes.
Similarly, a can also be written as the product of primes.
Combining these two results, we see that k + 1 = a · b can be written as
the product of primes.

c Xin He (University at Buffalo) CSE 191 Descrete Structures 13 / 14

Understanding example for strong induction
In this example, we have shown two things:
(a) 2 can be written as the product of primes;
(b) If all n such that 1 < n ≤ k can be written as the product of primes, then
k + 1 can be written as the product of primes.
Hence, we have the following statements being true:
(1) 2 can be written as the product of primes; (This is (a))
(2) If 2 can be written as the product of primes, then 3 can be
written as the product of primes; (This is (b) when k = 2)
(3) If 2 and 3 can be written as the product of primes, then 4 can
be written as the product of primes; (This is (b) when k = 3) . . .
(n-1) If 2, 3, . . ., and n − 1 can be written as the product of primes,
then n can be written as the product of primes; (This is (b)
when k = n − 1)
Putting all of them together, we see that n can be written as the product of
primes.

c Xin He (University at Buffalo) CSE 191 Descrete Structures 14 / 14