' $

CH 8 Bivariate Distribution

F Reading:
ch 8.1 , 8.2 , 8.3 , 8.4

F Topics:
1. Joint Distribution of Two Random Variables
2. Indepedent Random Variables
3. Conditional Distributions
4. Transformations of Two Random Variables

& %
1
' $
⌥ ⌅
⌃Ch 8.1 Joint Distribution of Two Random Variables ⇧

F Def.: Let X and Y be two discrete r.v’s defined on the same sample space.
Let the sets of possible values of X and Y be A and B, respectively. The
function p(x, y) = P (X = x, Y = y) is called the joint probability mass
P
function of X and Y . Then the functions pX (x) = y2B P (x, y) and
P
pY (y) = x2A P (x, y) are called, respectively, the
marginal probability mass functions of X and Y .
Explain:

& %
2
' $
F Example 8.2: Roll a balanced die and let the outcome be X. Then toss a
fair coin X times and let Y denote the number of tails. What is the joint P.M.F.
of X and Y and the marginal P.M.F. of X and Y ?
Sol.:

& %
3
' $
F Theorem 8.1: Let p(x, y) be the joint P.M.F. of discrete r.v’s X and Y . Let A
and B be the set of possible values of A and B, respectively. If h is a
function of two variables from R2 to R, then h(X, Y ) is a discrete r.v. with
the expected value given by
XX
E[h(X, Y )] = h(x, y)p(x, y)
x2A y2B

provided that the sum is absolutely convergent.

Proof: Later in Ch10.

F Corollary: For discrete r.v’s X and Y

E(X + Y ) = E(X) + E(Y )

Proof:

& %
4
' $
F

F Def.: Two r.v’s X and Y , defined on the same sample space, have a
continuous joint distribution if there exists a nonnegative function of two
variables, f (x, y) on R ⇥ R, such that for any region R in the xy-plane that
can be formed from rectangles by a countable number of set operations ,
ZZ
P ((X, Y ) 2 R) = f (x, y)dxdy · · · (eg.1)
R

The ft. f (x, y) is called the joint p.d.f. of X and Y

F Note that if the region R is a plane curve, then

P ((X, Y ) 2 R) = 0

& %
5
' $
F Let R = {(x, y) : x 2 A, y 2 B} , where A and B are any subsets of real
numbers that can be constructed from intervals by a countable number of
R R
set operations using (eg.1). Then P (X 2 A, Y 2 B) = B A f (x, y)dxdy.
Rd Rb
Hence for a < b and c < d, P (a < X  b, c < Y  d) = c ( a f (x, y)dx)dy.
In addition, if ✏ and are very small positive numbers, then
P (a ✏ < X < a + ✏, b < Y < b + ) ⇡ 4✏ f (a, b).
Explain:

& %
6
' $
F Def.: Let X and Y have joint p.d.f. f (x, y); then the ft’s fX and fY , given by
R1 R1
fX (x) = 1 f (x, y)dy and fY (y) = 1 f (x, y)dx , are called,
respectively, the marginal p.d.f’s of X and Y .

F From the joint p.d.f. of X and Y , we can find the marginal p.d.f.s of X and Y
while the converse is in general not true.

F Def.: Let X and Y be two r.v’s (discrete, continuous, or mixed). The joint
probability distribution function, or joint cumulative probability distribution
function, or simply the joint distribution of X and Y , is defined by
F (t, u) = P (X  t, Y  u) for all 1 < t, u < 1. The marginal probability
distribution functions of X, FX , can be found from F as follows:
FX (t) = F (t, 1) and FY (u) = F (1, u).
Explain:

& %
7
' $
F Now suppose that the joint p.d.f. of X and Y is f (x, y). Then
Ry Rx
F (x, y) = 1 1 f (t, u)dtdu. Assuming that the partial derivatives of F
@2
exist, we get f (x, y) = @x@y F (x, y).
Explain:

& %
8
' $
F Example 8.4: The joint p.d.f. of r.v’s X and Y is given by

8
< xy 2 0xy1
f (x, y) =
: 0 otherwise.

(a) Determine the value of

(b) Find fX (x) and fY (y)
(c) Calculate E(X) and E(Y )
Sol.:

& %
9
' $
F Example 8.5: For > 0, let

8
< 1 e (x+y)
if x > 0, y > 0
F (x, y) =
: 0 otherwise.

Determine if F is the joint p.d.f. of two r.v’s X and Y .

Sol.:

& %
10
' $
F Example 8.6: A circle of radius 1 is inscribed in a square with sides of
length 2. A point is selected random from square. What is the probability
that it is inside the circle?
Sol.:

& %
11
' $

F Def.: Let S be a subset of the plane with area A(S). A point is said to be
randomly selected from S if for any subset R of S with area A(R), the
probability that R contains the point is A(R)/A(S). gometr
pobability ,

F Example 8.7: A man invites his fiancée to a fine hotel for a Sunday brunch.
They decide to meet in the lobby of the hotel between 11 : 30 A.M. and 12
noon. If they arrive at random times during this period, what is the
probability that they will meet within 10 minutes?
Sol:
& %
12
' $
F

& %
13
' $
F Theorem 8.2: Let f (x, y) be the joint p.d.f. of random variables X and Y . If
h is a function of two variables from R2 to R, then h(x, y) is a r.v. with the
R1 R1
expected value given by E[h(X, Y )] = 1 1 h(x, y)f (x, y)dxdy,
provided that the integral is absolutely convergent.

F Example 8.9: Let X and Y have joint p.d.f.

8
< 3
(x 2
+ y2 ) if 0 < x < 1, 0 < y < 1
2
f (x, y) =
: 0 otherwise.

Find E(X 2 + Y 2 )
Sol.:

& %
14
' $
⌥ ⌅
⌃ ⇧
Ch 8.2 Independent Random Variables

F Theorem 8.3 : Let X and Y be two r.v’s defined on the same space. If F is
the joint probability distribution ft. of X and Y , then X and Y are
independent iff for all real numbers t and u, F (t, u) = FX (t)FY (u).
Explain:

& %
15
' $
F Theorem 8.4: Let X and Y be two discrete r.v’s defined on the same space.
If p(x, y) is the joint probability mass function of X and Y , then X and Y
are independent iff for all real numbers x and y, P (x, y) = PX (x)PY (y).

F X and Y are independent if knowing the value of one of them does not
change the probability mass function of the other.
Explain:

& %
16
' $
F Suppose that n + m independent Bernoulli trials, each with parameter p,
are performed. If X and Y are the number of successes in the first n and
in the last m trials, then X and Y are binomial r.v’s with parameters (n, p)
and (m, p), respectively. Furthermore, they are independent. Hence
P (X = i, Y = j) = P (X = i)P (Y = j) = ni m j p
i+j
(1 p)(n+m) (i+j)

F Theorem 8.5: Let X and Y be independent r.v’s and g : R ! R and

h : R ! R be real-valued functions; then g(X) and h(Y ) are also
independent r.v 0 s.
Proof:

& %
17
' $
F Theorem 8.6: Let X and Y be independent r.v’s. Then for all real-valued
functions g : R ! R and h : R ! R, E[g(X)h(Y )] = E[g(X)]E[h(Y )],
where, as usual, we assume that E[g(X)] and E[h(Y )] are finite.
Sol.:

& %
18
' $
F If X and Y are independent, then E(XY ) = E(X)E(Y ). However , two
r.v’s X and Y might dependent while E(XY ) = E(X)E(Y ).

F Example 8.11: Let X be a r.v. with the set of possible values{ 1, 0, 1} and
P.M.F. p( 1) = p(0) = p(1) = 13 . Letting Y = X 2 . Please show that X and
Y are dependent but E(XY ) = E(X)E(Y ).

& %
19
' $
F Theorem 8.7: Let X and Y be jointly continuous r.v’s with joint p.d.f. f (x, y).
Then X and Y are independent iff f (x, y) is the product of their marginal
densities fX (x) and fY (y).
Explain:

& %
20
' $
F Example 8.13: A point is selected at random from the rectangle
R = {(x, y) 2 R2 : 0 < x < a, 0 < y < b}. Let X be the x-coordinate and Y
be the y-coordinate of the point selected. Determine if X and Y are
independent r.v’s.
Sol.:

& %
21
' $
F Example 8.14: (Buffon’s Needle Problem) A plane is ruled with parallel
lines a distance d apart. A needle of length l, l < d, is tossed at random
onto the plane. What is the probability that the needle intersects one of the
parallel lines?
Sol.:

& %
22
' $
F The result of Theorem 8.5 and 8.6 are valid for continuous r.v’s:
Proof :

& %
23
' $
⇤
⇥Ch 8.3 Conditional Distributions

F Def.: Let X(Y ) be a discrete r.v. with set of possible values A(B). The
conditional P.M.F. of X given that Y = y is used, denoted by pX|Y (X|Y ) , is
defined as follows :
P (X = x, Y = y) p(x, y)
pX|Y (x|y) ⌘ P (X = x|Y = y) = = ,
P (Y = y) pY (y)

where x 2 A, y 2 B, and pY (y) > 0.

F Note: For any fixed y 2 B, pX|Y (x|y) is itself a P.M.F. with the set of values
A. If X and Y are independent, pX|Y coincides with pX .
Explain:

& %
24
' $
F The conditional distribution function of X, given that Y = y is defined as
follows:
X X
FX|Y (x|y) = P (X  x|Y = y) = P (X = t|Y = y) = pX|Y (t|y).
tx tx

F Example 8.16: Let the joint P.M.F. of X and Y be given by

8
< 1
if x = 0, 1, 2, y = 1, 2
15 (x + y)
p(x, y) =
: 0 otherwise.

Find pX|Y (x|y) and P (X = 0|Y = 2).

FSol:

& %
25
' $

F The conditional expectation of the random variable X given that Y = y is

as follows :
X X
E(X|Y = y) = xP (X = x|Y = y) = xpX|Y (x|y),
x2A x2A

where A denotes the set of possible values of X and pY (y) > 0. In

P
addition, we have E[h(X)|Y = y] = x2A h(x)pX|Y (x|y), where h is an
ordinary function from R to R.

F Example 8.18: Calculate the expected number of aces in a randomly

selected poker hand that is formed to have exactly two jacks.
Sol.:
& %
26
' $

& %
27
' $
F Def.: Let X and Y be two continuous r.v’s with the joint p.d.f. f (x, y). The
conditional p.d.f. of X given that Y = y, denoted as fX|Y (x|y), is defined
as follows :
f (x, y)
fX|Y (x|y) = , provided that fY (y) > 0.
fY (y)

Similarly, the conditional p.d.f. of Y given that X = x, denoted as

fY |X (y|x), is defined by

f (x, y)
fY |X (y|x) = , provided that fX (x) > 0.
fX (x)

Also FX|Y (x|y), the conditional probability distribution function of X given

Rx
that Y = y, is defined by FX|Y (x|y) = P (X  x|Y = y) = 1 fX|Y (t|y)dt.
Explain:

& %
28
' $
F Example 8.22: First, a point Y is selected at random from the interval
(0, 1). Then another point X is chosen at random from the interval (0, Y ).
Find the p.d.f. of X.
Sol.:

& %
29
' $
F For continuous r.v’s X and Y with joint p.d.f.. f (x, y), the conditional
expectation of X given that Y = y is as follows:
Z 1
E(X|Y = y) = xfX|Y (x|y)dx, where fY (y) > 0.
1

In addition, if h is an ordinary function R to R , then

R1
E[h(X)|Y = y] = 1 h(x)fX|Y (x|y)dx

F Example 8.25: The lifetimes of batteries manufactured by a certain

company are identically distributed with probability distribution and
probability density function F and f , respectively. In terms of F , f , and s,
find the expected value of the lifetime of an s-hour-old battery.
Sol.:

& %
30
' $

& %
31
' $
⇤
⇥Ch 8.4 Transformations of Two Random Variables

F Theorem 8.8 : Let X and Y be continuous random variables with joint

probability density function f (x, y). Let h1 and h2 be real-valued functions
of two variables, U = h1 (X, Y ) and V = h2 (X, Y ).
Suppose that :

(a) u = h1 (x, y) and v = h2 (x, y) defines a one-to-one transformation of a set

R in the xy-plane onto a set Q in the uv-plane. That is, for (u, v) 2 Q, the
system of two equations in two unknowns,

8
< h (x, y) = u
1
: h2 (x, y) = v,

has a unique solution x = w1 (u, v) and y = w2 (u, v) for x and y, in terms of

u and v; and
& %
32
' $
(b) the functions w1 and w2 have continuous partial derivatives, and the
Jacobian of the transformation x = w1 (u, v) and y = w2 (u, v) is nonzero at
all points (u, v) 2 Q; that is, the following 2 ⇥ 2 determinant is nonzero on
Q:
@w1 @w1
@u @v @w1 @w2 @w1 @w2
J= = 6= 0. (1)
@w2 @w2 @u @v @v @u
@u @u

Then the random variables U and V are jointly continuous with the joint
probability density function g(u, v) given by

8
< f (w (u, v), w (u, v))|J| (u, v) 2 Q
1 2
g(u, v) =
: 0 elsewhere.

Proof :

& %
33
' $

& %
34
' $
F Example : 8.27 (Box-Muller’s Theorem) : Let X and Y be two independent
uniform random variables over (0, 1); show that random variables
p p
U = cos(2⇡X) 2 ln Y and V = sin(2⇡X) 2 ln Y are independent
standard normal variables.
Sol :

& %
35
' $

& %
36
' $

& %
37
' $
F Theorem 8.9 (Convolution Theorem): Let X and Y be continuous
independent random variables with probability density functions f1 and f2
and probability distribution functions F1 and F2 , respectively. Then g and
G, the probability density and distribution functions of X + Y , respectively,
are given by

R +1
g(t) = 1
f1 (x)f2 (t x)dx,
R +1
G(t) = 1
F1 (x)F2 (t x)dx,
Proof :

& %
38
' $

& %
39
' $

& %
40