Math 465: Introduction to Combinatorics

Sergey Fomin

Quiz #2: 30 minutes, open Friday–Saturday.

Homework #2 will be due Monday evening.

These slides will be posted on Canvas.

Another binomial identity

Theorem
X n 
k = n 2n−1
k
k

Example: n = 5
0 · 1 + 1 · 5 + 2 · 10 + 3 · 10 + 4 · 5 + 5 · 1 = 80 = 5 · 24 .

Proof
Suppose we need to choose a subset of an n-element set, then pick
an element of this subset (e.g., a committee and its chairperson).
The number of such choices can be computed in two different ways:
• choose a k-element subset, then pick one its k elements;
• pick one of the n elements, then add a subset of the rest.
Binomial Theorem
Finally, an explanation of the term “binomial coefficients:”
(x + y )0 = 1
(x + y )1 = x +y
(x + y )2 = x 2 + 2xy + y 2
(x + y )3 = x 3 + 3x 2 y + 3xy 2 + y 3
(x + y )4 = x 4 + 4x 3 y + 6x 2 y 2 + 4xy 3 + y 4
(x + y )5 = x 5 + 5x 4 y + 10x 3 y 2 + 10x 2 y 3 + 5xy 4 + y 5
·········· ················································
Binomial Theorem
n  
n
X n k n−k
(x + y ) = x y
k=0
k
Proof of the Binomial Theorem

Binomial Theorem
n  
n
X n k n−k
(x + y ) = x y
k=0
k

Proof
Expand (x + y )n as the sum of n-letter words in the alphabet {x, y }:
(x + y )3 = (x + y )(x + y )(x + y )
= xxx + xxy + xyx + xyy + yxx + yxy + yyx + yyy .
(Here we view each word as a monomial in x and y .) Collect terms.
The coefficient of x k y n−k is kn [why?].

This identity has an interpretation in probability theory (for x +y = 1).

Binomial identities via the Binomial Theorem (1)
Many binomial identities can be proved using the Binomial Theorem
n  
n
X n k n−k
(x + y ) = x y .
k=0
k

Example
Set x = y = 1 to get
n  
X n
= 2n .
k=0
k

Example
Set x = −1 and y = 1 to get
n  
k n
X
(−1) = 0.
k=0
k
Binomial identities via the Binomial Theorem (2)

Example
Start with
(x + 1)n · (x + 1)m = (x + 1)n+m .
Use the Binomial Theorem to rewrite it as
n   m   n+m  
X n k X m j X n+m `
x · x = x .
k=0
k j=0
j `=0
`
Take the coefficient of x ` on both sides to get
X n  m  n + m
= .
k
k `−k `
Binomial identities via the Binomial Theorem (3)

Example
Differentiate
n  
n
X n
(x + 1) = xk
k=0
k
to get
n  
n−1
X n
n(x + 1) = kx k−1 .
k=1
k
Now set x = 1 to obtain
X n 
n−1
n2 = k .
k
k
Multinomial theorem

Recall that the multinomial

 coefficient

n def n!
==
n1 · · · nj n1 ! · · · nj !
is the number of words in the alphabet {x1 , . . . , xj } that consist of
n1 copies of letter x1 , . . . , nj copies of letter xj . Here n = n1 +· · ·+nj .

Multinomial theorem
 
X n n
n
(x1 + · · · + xj ) = x1n1 · · · xj j .
n1 +···+nj =n
n1 · · · nj

Proof
Expand (x1 + · · · + xj )n as the sum of all words of length n in the
alphabet {x1 , . . . , xj }, viewing words as monomials. Collect terms.
Multinomial theorem: Example
(a + b + c + d)3
= (a + b + c + d)(a + b + c + d)(a + b + c + d)
= aaa + aab + aac + aad + aba + abb + abc + abd
+aca + acb + acc + acd + ada + adb + adc + add
+baa + bab + bac + bad + bba + bbb + bbc + bbd
+bca + bcb + bcc + bcd + bda + bdb + bdc + bdd
+caa + cab + cac + cad + cba + cbb + cbc + cbd
+cca + ccb + ccc + ccd + cda + cdb + cdc + cdd
+daa + dab + dac + dad + dba + dbb + dbc + dbd
+dca + dcb + dcc + dcd + dda + ddb + ddc + ddd
= a3 + b 3 + c 3 + d 3 + 6abc + 6abd + 6acd + 6bcd
+3a2 b + 3ab 2 + 3a2 c + 3ac 2 + 3a2 d + 3ad 2
+3b 2 c + 3bc 2 + 3b 2 d + 3bd 2 + 3c 2 d + 3cd 2
Multinomial theorem, continued

Problem
What is the largest coefficient of a monomial in the expansion of the
polynomial (x + y + z)6 ?

Solution
For a + b + c = 6, the largest value of
 
6 6!
=
abc a! b! c!
6!
is attained at a = b = c = 2 [why?], so the answer is = 90.
23
Generalized binomial coefficients

Definition
For α a complex number
and k a nonnegative integer, the generalized
α
binomial coefficient k is defined by
 
α def α(α − 1) · · · (α − k + 1)
== .
k k!

When α = n is a nonnegative integer, we recover the usual definition:

 
n n(n − 1) · · · (n − k + 1) n!
= = .
k k! (n − k) !k!
Note that for k > n, we get kn = 0.


Generalized binomial coefficients: the case α = −n

The generalized binomial coefficients are defined by

 
α def α(α − 1) · · · (α − k + 1)
== .
k k!

Example: α = −n
 
−n (−n)(−n − 1) · · · (−n − k + 1)
=
k k!
(n + k − 1) · · · (n + 1) n
= (−1)k
 k!

k n + k − 1
= (−1) .
k
Generalized Binomial Theorem

Meta-Theorem [I. Newton]

The formula
∞  
α
X α
(x + y ) = x k y α−k
k=0
k
holds whenever it makes sense.

In particular, it holds in each of the following cases:

• when α is an integer (possibly negative);
• when x + y and y are positive real numbers;
• in the realm of formal power series, to be discussed soon.
Proof of Newton’s generalized Binomial Theorem
Viewing (x + y )α as a function of x, we apply Taylor’s formula:
f (x) = (x + y )α ,
f 0 (x) = α(x + y )α−1 ,
f 00 (x) = α(α − 1)(x + y )α−2 ,
···············
f (x) = α(α − 1) · · · (α − k + 1)(x + y )α−k ,
(k)

···············
∞
α
X f (k) (0) k
f (x) = (x + y ) = x
k=0
k!
X α(α − 1) · · · (α − k + 1) y α−k
= xk
k
k!
X α  
= x k y α−k .
k
k
Binomial Theorem for negative powers
Setting α = −n in Newton’s theorem, we get:
X∞ ∞ ∞
−n −n

k X k n+k−1

k X n+k−1
(−1)k


(1+x) = k
x = (−1) k
x = n−1
xk.
k=0 k=0 k=0

Corollary
∞  
−n
X n+k −1 k
(1 − x) = x .
k=0
n − 1

Examples
X k  X
−1
(1 − x) = xk = xk,
k
0 k
X k + 1 X
(1 − x)−2 = xk = (k + 1)x k .
k
1 k
Generating functions

Definition
Let
h0 , h1 , h2 , h3 , . . .
be a finite (resp., infinite) sequence of nonnegative integers.
The generating function of this sequence is the polynomial (resp.,
formal power series)
X
h(x) = h0 + h1 x + h2 x 2 + h3 x 3 + · · · = hk x k .
k

Example
Let hk = k + 1, i.e., we look at the sequence 1, 2, 3, . . . Then
X
h(x) = 1 + 2x + 3x 2 + 4x 3 + · · · = (k + 1)x k = (1 − x)−2 .
k
Why generating functions?

In a typical application, hk is an answer to an enumerative question

that depends on k.

Example
(Here hk = number of k-element subsets of an n-element set.)
The generating function for the n th row of Pascal’s triangle is
X n 
x k = (1 + x)n .
k
k

The generating function is often simpler than its coefficients.

One can use algebraic manipulations with generating functions to
obtain results about sequences.
Generating functions: Example 1

Problem
How many different poker hands can be dealt from a double deck of
2 × 52 cards? (Each poker hand consists of 5 cards.)
The two instances of each card are indistinguishable from each other.

Solution #1
Each of the 52 types of cards will be drawn either 0 or 1 or 2 times.
Each potential drawing can be encoded by a 52-letter word in the
alphabet {0, 1, 2}.
For a drawing to contain 5 cards, the number of occurences of 0, 1, 2
must be either (47, 5, 0) or (48, 3, 1) or (49, 1, 2). Hence the answer is
     
52 52 52
+ + = 3748160.
47 5 0 48 3 1 49 1 2
Generating functions: Example 1, continued

Problem
How many different poker hands can be dealt from a double deck of
2 × 52 cards?
We will solve a more general problem, for hands consisting of k cards.

Solution #2 (using generating functions)

Set hk = #{k-card hands which can be dealt from a double deck}.
Then
X
hk x k = (1 + x + x 2 )52
k
= (1 + x + x 2 )(1 + x + x 2 ) · · · (1 + x + x 2 ) [why?].
| {z }
52 factors
In other words, hk is simply the coefficient of x k in (1 + x + x 2 )52 .