Maggie Johnson Handout
#27
CS103A
Introduction to Induction
Key topics:
* Introduction and Definitions
* Examples of Weak Induction
* Proper Proof Form
* Examples of Strong Induction
* A Faulty Proof and Some Interesting Proofs
*
Some Useful Formulas
One of the most important
tasks of mathematics is to discover and characterize regular patterns or
sequences. The main mathematical tool
we use to prove statements about sequences is induction. Induction is a very important tool in
computer science for several reasons, one of which is the fact that a
characteristic of most programs is repetition of a sequence of statements.
To illustrate how induction
works, imagine that you are climbing an infinitely high ladder. How do you know whether you will be able to
reach an arbitrarily high rung? Suppose
you make the following two assertions about your climbing abilities:
1) I can definitely reach the first rung.
2) Once I get to any rung, I can always climb to the next
one up.
If both statements are true,
then by statement 1 you can get to the first one, and by statement 2, you can
get to the second. By statement 2
again, you can get to the third, and fourth, etc. Therefore, you can climb as high as you wish. Notice that both of these assertions are
necessary for you to get anywhere on the ladder. If only statement 1 is true, you have no guarantee of getting
beyond the first rung. If only statement
2 is true, you may never be able to get started.
Assume that the rungs of the
ladder are numbered with the positive integers (1,2,3...). Now think of a specific property that a
number might have. Instead of
"reaching an arbitrarily high rung", we can talk about an arbitrary
positive integer having that property.
We will use the shorthand P(n) to denote the positive integer n having
property P. How can we use the
ladder-climbing technique to prove that P(n) is true for all positive n? The two assertions we need to prove are:
1) P(1) is true
2) for any positive k, if P(k) is true, then P(k+1) is
true
Assertion 1 means we must
show the property is true for 1; assertion 2 means that if any number has property P then so does the next
number. If we can prove both of these
statements, then P(n) holds for all positive integers, just as you could climb
to an arbitrary rung of the ladder.
The foundation for arguments
of this type is the Principle of
Mathematical Induction . The Principle of Mathematical Induction can be
used as a proof technique on statements that have a particular form. First of all, we must be working with
positive integers or numbers of objects, and we must be talking about some kind
of sequence or pattern. The Principle
tells us that if we can prove the
assertion is true for 1, and then prove that it is also true for k+1 (assuming
it is true for k), we have proven the assertion for all positive integers,
because if it is true for 1 and k+1, then it is true for k. When it is true for all three, it is true
for all positive integers. Thus, the following steps are required:
i) Prove that
P(1) is true.
ii) Assume that
P(k) is true, and prove that P(k+1) is true.
Note that we do not prove
that P(k) is true (except for k = 1). Instead, we show that if
P(k) is true, then P(k+1) must
also be true. That's all that is
necessary according to the Principle of Mathematical Induction. Step (i) is called the basis of induction, and step (ii) is called the induction step. The assumption that P(k) is true is called
the induction hypothesis.
As an another illustration,
consider an infinite number of dominoes positioned one behind the other in such
a way that if any given one falls, then the one behind it falls too. In order to establish that the entire chain
will fall under a certain set of circumstances, two things are necessary. First, someone has to push over the first
domino; this corresponds to the basis of induction. Second, we must also know
that whenever any domino falls, it will knock over the next in the chain. This requirement can be expressed by saying
that whenever domino N falls, so does N+1.
This corresponds to using the induction hypothesis to establish the
result for the next value of N.
Example 1: Suppose that great, great, great.... grandpa
Fred married and had two children. Call
this generation 1, so generation 1 contains the offspring of Fred (2). Each of these children have two children, so
generation 2 has four offspring. Each
of these 4 has two children so at generation 3, we have 8 offspring. This continues, without fail, for generation
to generation. It appears that
generation n will have 2n offspring. We are guessing that P(n) = 2n. We can use induction to prove that our guess is
correct.
i) base case: prove that P(1) is true
P(1) = 21 = 2 (generation
1 has 2 offspring)
ii) assume our guess is correct for an arbitrary
generation k, and prove it is also true
for generation k+1.
induction hypothesis is to assume P(k): P(k) = 2k
and show:
P(k+1) = 2(k+1)
PROOF:
By
the rules of this family, each offspring is required to have 2 children, thus
the number of offspring at generation k+1 will be twice the number of
generation k, or P(k+1) = 2P(k). By the
inductive hypothesis, P(k) = 2k, so :
P(k+1) = 2P(k) = 2(2k) = 2(k+1)
P(k+1) is true when P(k) is true, and therefore P(n) is
true for all natural numbers.
Example 2: Prove that for every positive integer n, the sum of
the first n positive integers is n(n+1)/2.
This is the classic example of
an inductive proof:
P(n)
denotes: 1+2+3+....+n = n(n+1)/2
if n = 2 then:
1+2 = 2(2+1)/2 = 3
if n = 3 then:
1+2+3 = 3(3+1)/2 = 6 etc.
To prove this for all positive integers, we follow the
steps given above for inductive proofs:
i) base case: prove that P(1) is true
1 = 1(1+1)/2 = 1
ii) induction hypothesis is to assume P(k): 1+2+3+...+k = k(k+1)/2
and show: P(k+1):
1+2+3+...+(k+1) = (k+1)[(k+1)+1]/2 (by substituting (k+1) for k in
the
induction hypothesis)
PROOF:
By the induction hypothesis, we already have a formula
for the first k integers. So,
a formula for the first (k+1) integers may be found by simply adding (k+1)
to both sides of the induction hypothesis, and
simplifying:
1+2+3+...+k+(k+1) = k(k+1)/2 + (k+1)
=
[k(k+1) + 2(k+1)] / 2
= (k2 + 3k + 2) / 2
=
[(k+1)(k+2)] / 2
=
(k+1)[(k+1)+1] / 2
P(k+1) is true when P(k) is true, and therefore P(n) is
true for all natural numbers.
Steps to doing an inductive
proof:
1) state P(n)
2) show that P(base case) is true
3) state the inductive hypothesis (substitute k for n)
4) state what must be proven (substitute k+1 for n)
5) state that you are beginning your proof, and proceed
to manipulate the inductive hypothesis
(which we assume is true) to find a link between the inductive hypothesis and
the statement to be proven. Always state explicitly where you are
invoking the inductive hypothesis.
6) Always finish your proof with something like: P(k+1)
is true when P(k) is true, and therefore
P(n) is true for all natural numbers.
NOTE: On problem sets and exams, it is important to
present your proof in the proper form.
You must always write out
steps 1-6 as given above. Form is
important in inductive proofs.
Example 3: P(n) denotes 1 + 3 + 5 + ... + (2n-1) = n2,
i.e., the sum of the first n odd integers is n2
i) base case: prove that P(1) is true 1 = 12 = 1
ii) induction
hypothesis is to assume P(k): 1 + 3 + 5
+ ... + (2k-1) = k2
and show: P(k+1): 1 + 3 + 5 + ... + (2k-1) + (2k+1) = (k+1)2
PROOF:
Simulate adding another term to both sides of the
induction hypothesis:
1 + 3 + 5 + ... + (2k-1) + (2k+1) = k2 + (2k+1)
=
k2 + 2k + 1
=
(k+1)2
P(k+1) is true when P(k) is true, and therefore P(n) is
true for all natural numbers.
The following example
illustrates that the basis of induction does not have to be 1:
Example 4: P(n) denotes 1 + 2 + 22 + ... + 2n = 2n+1 -1 for all nonnegative
integers (this includes 0).
i) base case:
prove that P(0) is true: 20 = 1
= 21 -1
ii) induction hypothesis is to assume P(k): 1 + 2 + 22 + ... + 2k = 2k+1-1
and show: P(k+1): 1 + 2 + 22 + ... + 2k + 2k+1 = 2((k+1)+1) - 1 =
2k+2 -1
PROOF:
add 2k+1 to both sides of the inductive
hypothesis:
(1 +
2 + 22 + ... + 2k )+ 2k+1 = (2k+1-1 ) + 2k+1
=
(2 * 2k+1 ) -1 factoring
=
2k+2 - 1 exponent defin
P(k+1) is true when P(k) is true, and therefore P(n) is
true for all nonnegative numbers.
An example with two
variables (but only one induction variable):
Example 5: P(n) denotes that for any real number r
except 1 and any integer n >= 0,
the summation of ri for i = 0 to n equals (r(n+1) - 1) / r - 1
i) base case: prove that P(0) is true (r(0+1) - 1) / r - 1 = 1 and r0
= 1.
ii) induction hypothesis is to assume P(k): the summation of ri for i =
0 to k equals (r(k+1) -
1) / r - 1
and show: P(k+1): the summation of ri for i = 0
to k+1 equals (r(k+2) - 1) / r - 1
PROOF:
the summation of ri for i = 0 to k+1 equals
the summation of ri for i
= 0
to k + r(k+1)
=
(r(k+1) - 1) / r - 1 + r(k+1)
=
(r(k+1) - 1) / r - 1 + (r(k+1)
* (r-1)) / r-1
=
((r(k+1) - 1) + (r(k+1) * (r-1))) / r-1
=
((r(k+1) - 1) + (r(k+1) * r) - r(k+1)) / r-1
=
((r(k+1) * r) - 1) / r - 1
=
r(k+2) - 1 / r - 1
P(k+1) is true when P(k) is true, and therefore P(n) is
true for all natural numbers.
Thus far, we have been using
weak induction in our proofs. There
is a variation called strong
induction, Rather than assume that P(k)
is true to prove that P(k+1) is true, we assume that P(i) is true for all i
where the (basis of induction) <= i <= k. From this assumption, we prove P(k+1). It's stronger in the sense that we are allowed to come to the
same conclusion while assuming more, but the assumption is a natural one based
on our understanding of weak induction. In fact, weak induction is a special
case of strong induction.
Consider the domino
analogy. We still need to prove that
the n+1 domino will fall, but instead of just assuming the previous one has
fallen, we can assume that all the previous ones have fallen, which is correct
- they did all fall.
Strong or Complete Induction:
i)
Prove P(base) is true
ii)
Assume P(base), P(base+1)...P(k)
is true, and prove that P(k+1) is true.
Example 6: P(n) denotes that all positive integers n
have a prime factorization, i.e., n can
be written as the product of primes.
i) base case:
prove that P(1) is true 1 = 1 * 1
ii) induction hypothesis is to assume P(i): for
all positive integers i where 1 < i <= k, i has a prime factorization;
and show P(k+1): k+1 has a prime
factorization.
PROOF:
If
k+1 is a prime, then it has the obvious prime factorization k+1 * 1. If k+1 is not a prime, then by the
definition of primality, it must be true that for some a & b, a * b =
k+1. Clearly, a <= k and b <=
k. But by the inductive hypothesis,
both a and b have a prime factorization too.
Thus k+1 is the product of prime factors a & b, proving that k+1 has
a prime factorization.
P(k+1)
is true when P(i) is true, where i <= k, and therefore P(n) is true for all
natural numbers.
Strong induction is often
used to prove that a sequence of numbers has a particular property.
Example 7:
P(n) denotes the following: If b1, b2, b3... is a sequence defined by the following:
b0 = 1,
b1 = 2, b2 = 3, bk = bk-3 + bk-2
+ bk-1 for all integers k >= 3;
then, bn is
<= 3n for all integers n >= 0.
i) base case: prove that P(0), P(1) and P(2) are true. (Note: in sequence problems, you must
prove all given base cases.)
b0 = 1 which is <= 30
b1 = 2 which is <= 31
b2 = 3 which is <= 32
ii) induction hypothesis is to assume P(i) for k > 2:
for all positive integers i where
1 <= i < k, bi <= 3i, and
show that bk <= 3k.
PROOF:
The definition of the sequence tells us that bk-3 +
bk-2 + bk-1 for all integers k >= 3.
Since k > 2, 0 <= k-3 <= k, and so by the
inductive hypothesis, bk-1 <= 3k-1, bk-2 <=
3k-2
and bk-3 <= 3k-3. If we add these inequalities and apply the
laws of basic algebra, we
get
bk-3 + bk-2 + bk-1 <=
3k-1 + 3k-2 + 3k-3
3k-1 + 3k-2 + 3k-3 <= 3k-1 + 3k-1 +
3k-1
3k-1
+ 3k-1 + 3k-1 =
3 * 3k-1 = 3k
And thus by substitution, bk <= 3k which is what we set out to
prove.
P(k) is true when P(i) is true, where i < k, and
therefore P(n) is true for all natural numbers.
Now we look at an additional example presented to expand your
conception of the application of induction:
Example 8: A jigsaw puzzle consists of a number of pieces. Two or more pieces with matched boundaries
can be put together to form a "big" piece. To be more precise, we use the term block to refer to either a
single piece or a number of pieces with matched boundaries that are put
together to form a "big" piece.
Thus, we can simply say that blocks with matched boundaries can be put
together to form another block.
Finally, when all pieces are put together as one single block, the
jigsaw puzzle is solved. Putting 2
blocks together with matched boundaries is called one move. We shall prove (using strong induction) that
for a jigsaw puzzle of n pieces, it will always take n-1 moves to solve the
puzzle.
i) base case:
prove that P(1) is true: For a puzzle with 1 piece, it does not take any moves to solve it.
ii)
induction hypothesis is to assume P(i) where 1<= i <= k: for a puzzle
with i pieces, it takes i-1 moves to
solve the puzzle; and show that for a puzzle with k+1 pieces, it takes k moves
to solve the puzzle.
PROOF:
Consider
the puzzle with k+1 pieces. For the
last move that produces the solution to the puzzle, we have two blocks: one with n1 pieces and the other
with n2 pieces, where n1 + n2 = k + 1. These two blocks will then be put together to solve the puzzle.
According to the induction hypothesis, it took
n1 - 1 moves to put together the one block, and n2 - 1 moves to put
together the other block. Including the
last move to unite the two blocks, the total number of moves is equal to [(n1 - 1) + (n2 - 1)] + 1 =
(k + 1) - 1
= k
P(k+1) is true when P(i) is true, where i <= k, and
therefore P(n) is true for any puzzle size.
The last topic to discuss
concerning induction is:
The
Law of Bad Proofs:
You can prove anything you want if
your proof is wrong.
In some ways, induction
seems like it can prove any property about anything. But you have to watch out for subtle flaws in your thinking. Consider the following:
P(n) denotes that a set of n horses are all the same
color.
base case: prove that P(1) is true: If X is a set of 1 horse, then all the
horses in X are the same color.
induction hypothesis is to assume P(k) is true: every set of k horses are the same color.
and show P(k+1): every set of k+1 horses are the same
color.
PROOF:
Suppose X is a set of k+1 horses. To show that all the horses in X are the
same color, it's
enough to show that if
h1 is in X and h2 is in X
then h1 is the same color as h2. If we can prove this, we are done because
either h1 or
h2 is in a set of k horses, and we know that a set of k
horses are all the same color.
Let X1 = X - h1 and X2 = X - h2
X1 and X2 have k horses and by the inductive
hypothesis, these k horses are all the
same color. So,
if we take a horse z which is a horse in the intersection of X1 and X2
(that being one of the horses that X1 and X2 have in
common), we know that h1 must
be the same color as z, and h2 must be the same color as
z; therefore, h1 is the same
color as h2. We
have proven that all horses are the same color.
What's wrong with this?
Show that any 2n
by 2n grid can be
tiled using “L” shaped pieces (such as the one below) leaving any one square
untiled (from your textbook).

Solution: Proof by induction on n.
Basis: It is easy to
verify the case for n = 1 by
exhaustively checking all possibilities:

Induction:
Prove: A 2k+1
by 2k+1 grid
can be tiled as described above.
Assuming: Any 2k by 2k grid can be
tiled.
We begin with a 2k+1
by 2k+1
grid. Divide this grid into four
sub-grids, each of size 2k
by 2k. By the inductive hypothesis, each of these
grids can be tiled leaving any one square empty.
Tile three of the four sub-grids so
that the empty squares are adjacent and form an “L” shaped hole. The fourth sub-grid can be tiled however we
please.

Now, simply fill
the “L” shaped hole with another tile. Thus a 2k+1 by 2k+1
grid can be tiled, leaving any one square blank. This completes the proof.
Bibliography
G. Pólya, Induction and Analogy in Mathematics, Princeton,
NJ: Princeton University Press: 1954.
D. Solow, How to Read and Do Proofs, New York:
Wiley, 1982.
I.S. Sominskii, Method of Mathematical Induction, New
York: Blaisdell, 1961.
Some Useful Formulas
Here are some formulas for
sequences that are great for practicing proofs by induction. You will also find these formulas useful
later in 103 when we discuss recursion and recurrence relations. The ones we have proven in this handout are
included below.
·
1 + 2 + 3+....+ n
= n(n+1) / 2
·
1 + 3 + 5 + ... +
(2n-1) = n2
·
1 + 2 + 22 +
... + 2n = 2n+1
- 1
n
·
S ri
= (r(n+1) - 1) / r - 1
i=0
·
12 + 22
+ 32 + ... + n2 = [n(n+1)(2n+1)] / 6
·
13 + 23
+ 33 + ... + n3 = [ n(n+1) / 2]2
·
1*2 + 2*3 + ... +
n(n+1) = n(n+1)(n+2) / 3
·
1/1*2 + 1/2*3 + ... +
1/n(n+1) = n / (n+1)
·
12 + 32
+ 52 + ... + (2n+1)2 = (n+1)(2n+1)(2n+3) / 3
Historical Notes
The first known use of
mathematical induction was in the work of the 16th-century mathematician
Francesco Maurolico (1494-1575). In his
book Arithmeticorum Libri Duo, he
presented a variety of properties of the integers together with proofs of these
properties. He used induction to prove
some of these properties, the first one being the proof that the sum of the
first n odd integers is n2. The first formal explanation of
mathematical induction was presented by Augustus DeMorgan (1806-1871) in
1838. This is also the first time the
term "induction" was used in this context.