Chapter 4: Joint Probability Distributions

Dr. Suresh Kumar 1 / 50

Discrete Bivariate Random Variable
Let X and Y be two discrete random variables. Then
the ordered pair (X, Y) is called a two dimensional or
bivariate discrete random variable.
Joint Density Function
A function f such that (i) f(x, y) ≥ 0
(ii) f(x,
Xy) X= P[X = x, Y = y]
(iii) f(x, y) = 1
X=x Y=y
is called joint density function of (X, Y).
Dr. Suresh Kumar 2 / 50
Cumulative Distribution Function X X
The cdf of (X, Y) is given by F(x, y) = f(x, y).
X≤x Y≤y
Expectation
The expectation
XX of X and H(X, Y) are
E[X] = xf(x, y) = µX.
X=x Y=y
XX
E[H(X, Y)] = H(x, y)f(x, y).
X=x Y=y

Dr. Suresh Kumar 3 / 50

Covariance
If µX and µY are the means of X and Y respectively,
then covariance of X and Y, denoted by Cov(X, Y) is
defined as

Cov(X, Y) = E[(X − µX)(Y − µY )]

= E[XY] − E[X]E[Y].

Dr. Suresh Kumar 4 / 50

Marginal Density Functions
The marginal densities of X and Y are defined as
X
fX(x) = f(x, y).
Y=y
X
fY (y) = f(x, y).
X=x
Independent random variables
The discrete random variables X and Y are said to
be independent if and only if f(x, y) = fX(x)fY (y) for all
(x, y).
Dr. Suresh Kumar 5 / 50
Toss of two fair coins
Let X denote the number of heads and Y denote the number
of tails.
Then X = 0, 1, 2, Y = 0, 1, 2, and f (x, y ) is
X /Y 0 1 2 fX (x)
0 0 0 1/4 1/4
1 0 1/2 0 1/2
2 1/4 0 0 1/4
fY (y ) 1/4 1/2 1/4 1
Dr. Suresh Kumar 6 / 50
Ex. Two ballpoint pens are selected at random from a box that
contains 3 blue pens, 2 red pens, and 3 green pens. If X is the
number of blue pens selected and Y is the number of red pens
selected, find
(a) the joint probability mass function f (x, y ),
(b) P[X + Y ≤ 1].

Dr. Suresh Kumar 7 / 50

 (x3)(y2)(2−x−y
3
)
Sol. (a) We have f (x, y ) = 8 ,
(2)
x = 0, 1, 2; y = 0, 1, 2; 0 ≤ x + y ≤ 2.
X /Y 0 1 2 fX (x)
3 6 1 10
0 28 28 28 28
9 6 15
1 28 28 0 28
3 3
2 28 0 0 28
15 12 1
fY (y ) 28 28 28 1

(b) P[X + Y ≤ 1] = f (0, 0) + f (0, 1) + f (1, 0) = 9/14.

Dr. Suresh Kumar 8 / 50
Ex. In an automobile plant, two tasks are performed by robots,
the welding of two joints and tightening of three bolts. Let X
denote the number of defective joints and Y denote the number
of improperly tightened bolts produced per car. The probabili-
ties of (X, Y) are given in the table.
X/Y 0 1 2 3
0 0.84 0.03 0.02 0.01
1 0.06 0.01 0.008 0.002
2 0.01 0.005 0.004 0.001
Dr. Suresh Kumar 9 / 50
 X/Y 0 1 2 3
0 0.84 0.03 0.02 0.01
1 0.06 0.01 0.008 0.002
2 0.01 0.005 0.004 0.001
(i) Is it a density function?
X2 X 3
Yes. For, f(x, y) = 1.
X=0 Y=0

Dr. Suresh Kumar 10 / 50

 X/Y 0 1 2 3 fX(x)
0 0.84 0.03 0.02 0.01 0.9
1 0.06 0.01 0.008 0.002 0.08
2 0.01 0.005 0.004 0.001 0.02
fY (y) 0.91 0.045 0.032 0.013 1
(ii) Find the probability that there would be exactly one error
made by the robots.
P[X = 1, Y = 0] + P[X = 0, Y = 1]
= f(1, 0) + f(0, 1) = 0.06 + 0.03 = 0.09.
Dr. Suresh Kumar 11 / 50
 X/Y 0 1 2 3 fX(x)
0 0.84 0.03 0.02 0.01 0.9
1 0.06 0.01 0.008 0.002 0.08
2 0.01 0.005 0.004 0.001 0.02
fY (y) 0.91 0.045 0.032 0.013 1
(iii) Find the probability that there would be no improperly
tightened bolts.
X2
P[Y = 0] = f(x, 0)
X=0
= f(0, 0) + f(1, 0) + f(2, 0) = 0.91.
Dr. Suresh Kumar 12 / 50
 X/Y 0 1 2 3 fX(x)
0 0.84 0.03 0.02 0.01 0.9
1 0.06 0.01 0.008 0.002 0.08
2 0.01 0.005 0.004 0.001 0.02
fY (y) 0.91 0.045 0.032 0.013 1
(iv) Are the variables X and Y independent?
No. From the Table, we notice that f(0, 0) = 0.84,
fX(0) = 0.9 and fY (0) = 0.91. So we have
fX(0)fY (0) = 0.819 6= f(0, 0).
Dr. Suresh Kumar 13 / 50
(v) Find Cov(X, Y).
X2 X3
E[X] = xf(x, y) = 0.12,
X=0 Y=0
X2 X 3
E[Y] = yf(x, y) = 0.148,
X=0 Y=0
X2 X 3
E[XY] = xyf(x, y) = 0.064.
X=0 Y=0
Hence, Cov(X, Y) = E[XY] − E[X]E[Y] = 0.046.
Dr. Suresh Kumar 14 / 50
Ex. Suppose X and Y are two discrete random variables
taking only integer values. The joint density function of
(X , Y ) is
f (x, y ) = c/[n(n + 1)], 1 ≤ y ≤ x ≤ n, where n is some
positive integer.
(i) Find the value of c.
(ii) Find the marginal densities.
(iii) Given that n = 5, find P[X ≤ 3, Y ≤ 2].
Ans. c = 2, P[X ≤ 3, Y ≤ 2] = 1/3.
Dr. Suresh Kumar 15 / 50
Continuous Bivariate Random Variable
It is the ordered pair (X, Y) of two continuous random
variables X and Y.
Joint Density Function
A f is joint density function of (X, Y) if
(i) f(x, y) ≥ 0 ZZ
(ii) P[(X, Y) ∈ S]= f(x, y)dxdy
Z ∞Z ∞ S

(iii) f(x, y)dxdy = 1.

−∞ −∞
Dr. Suresh Kumar 16 / 50
Cumulative Distribution Function Z x Z y
The cdf of (X, Y) is given by F(x, y) = f(x, y)dxdy.
−∞ −∞
Marginal Density Functions
The marginal densities of X and Y are defined as
Z ∞
fX(x) = f(x, y)dy.
−∞
Z ∞
fY (y) = f(x, y)dx.
−∞
Dr. Suresh Kumar 17 / 50
Independent random variables
The continuous random variables X and Y are said to
be independent if and only if f(x, y) = fX(x)fY (y) for all
(x, y).
Expectation
The expectation of X and H(X, Y) are
Z Z ∞ ∞
E[X] = xf(x, y)dxdy = µX.
−∞ Z −∞ Z
∞ ∞
E[H(X, Y)] = H(x, y)f(x, y)dxdy.
−∞ −∞
Dr. Suresh Kumar 18 / 50
Covariance
If µX and µY are the means of X and Y respectively,
then covariance of X and Y, denoted by Cov(X, Y) is
defined as

Cov(X, Y) = E[(X − µX)(Y − µY )]

= E[XY] − E[X]E[Y].

Dr. Suresh Kumar 19 / 50

Ex. Let X denote a person’s blood calcium level and Y , the
blood cholesterol level. The joint density function of (X , Y ) is
f (x, y ) = k, 8.5 ≤ x ≤ 10.5, 120 ≤ y ≤ 240.
(i) Find the value of k.
f (x, y ) being joint density function, we have
Z 240 Z 10.5
1= kdxdy = 240k.
120 8.5

So k = 1/240 and f (x, y ) = 1/240.

Dr. Suresh Kumar 20 / 50
(ii) Find the marginal densities of X and Y .
The marginal density of X is
Z 240
1 1
fX (x) = dy = .
120 240 2
Similarly, the marginal density of Y is
Z 10.5
1 1
fY (y ) = dy = .
8.5 240 120

Dr. Suresh Kumar 21 / 50

(iii) Find the probability that a healthy person has a
cholesterol level between 150 to 200.
Z 200
5
P[150 ≤ Y ≤ 200] = fY (y )dy = .
150 12
(iv) Are the variables X and Y independent?
1 1 1
fX (x)fY (y ) = × = f (x, y ).
2 120 240
This shows that X and Y are independent.
Dr. Suresh Kumar 22 / 50
(v) Find ZCov(X , Y ).
240 Z 10.5
x
E [X ] = dxdy = 9.5,
120 8.5 240
Z 240 Z 10.5
y
E [Y ] = dxdy = 180,
120 8.5 240
Z 240 Z 10.5
xy
E [XY ] = dxdy = 1710.
120 8.5 240
Hence, Cov(X , Y ) = 1710 − 9.5 × 180 = 0.
Dr. Suresh Kumar 23 / 50
Ex. The joint density function of (X, Y) is

f(x, y) = c/x, 27 ≤ y ≤ x ≤ 33.

(i) Find the value of c.

To find c, we use
Z 33 Z x
f(x, y)dydx = 1
27 27
1
and we get c = 6−27 ln(33/27) .
Dr. Suresh Kumar 24 / 50
 33

y=x
x = 33

y
27

y = 27
23
23 27 33
x

Figure: The shaded golden region is the triangular region common to

the three regions given by the inequalities y ≥ 27, y ≤ x and x ≤ 33.

Dr. Suresh Kumar 25 / 50

(ii) Find the marginal densities and hence check the
independence of X and Y.
Z y=x
c
fX(x) = dy = c(1 − 27/x).
y=27 x
Z x=33
c
fY (y) = dx = c(ln 33 − ln y).
x=y x
We observe that
f(x, y) = c/x 6= fX(x)fY (y).
So XDr.and Y are not independent.
Suresh Kumar 26 / 50
(iii) Evaluate P[X ≤ 30, Y ≤ 28].
Z 28 Z 30
c
P[X ≤ 30, Y ≤ 28] = dxdy = 0.15.
27 y x

Dr. Suresh Kumar 27 / 50

Theorem: If X and Y are two independent random variables
with joint density f , then
E [XY ] = E [X ]E [Y ], that is, Cov(X , Y ) = 0.
Proof: We have Z Z ∞ ∞
E [XY ] = xyf (x, y )dxdy
Z−∞ −∞
∞ Z ∞
= xyfX (x)fY (y )dxdy
Z−∞
∞
−∞ 
Z ∞ 
= yfY (y ) xfX (x)dx dy
−∞ −∞
Dr. Suresh Kumar 28 / 50
 Z ∞ Z ∞ 
E [XY ] = yfY (y ) xfX (x)dx dy
Z−∞
∞
−∞

= yfY (y )E [X ]dy
−∞ Z
∞
= E [X ] yfY (y )dy
−∞
= E [X ]E [Y ].

Dr. Suresh Kumar 29 / 50

If E [XY ] = E [X ]E [Y ], then X and Y need not be inde-
pendent. For, consider
X /Y −2 −1 1 2 fX (x)
1 0 1/4 1/4 0 1/2
4 1/4 0 0 1/4 1/2
fY (y ) 1/4 1/4 1/4 1/4 1

We find that E [X ] = 5/2, E [Y ] = 0 and E [XY ] = 0. So E [XY ] =

Next, fX (1)fY (−2) = (1/2)(1/4) 6= 0 = f (1, −2).
Dr. Suresh Kumar 30 / 50
Pearson coefficient of correlation
If X and Y are two random variables with means µX,
µY , and variances σX2 and σY2 , then correlation between
X and Y is given by
Cov(X, Y)
ρXY = .
σX σY
It can be proved that ρXY lies in the range [−1, 1].
Further, |ρXY | = 1 if and only if Y = a + bX for some
real numbers a and b 6= 0.
Dr. Suresh Kumar 31 / 50
Figure: In case of large negative covariance, we have ρXY ≈ −1. In case
of nearly zero covariance, ρXY ≈ 0 while in case of very large positive
covriance, ρXY ≈ 1.

Dr. Suresh Kumar 32 / 50

Note that if ρXY = 0, we say that X and Y are uncorre-
lated. It does not imply that X and Y are unrelated. Of
course, the relationship, if exists, would not be linear.

In Robot’s example, σX2 = 0.146, σY2 = 0.268, Cov(X, Y) = 0.0

and therefore ρXY = 0.23.

Dr. Suresh Kumar 33 / 50

Covariance matrix
Covariance matrix of X and Y is written as
 2 
σX σXY
σYX σY2

Notice that the covariance matrix is always symmetric since

σXY = σYX .

Dr. Suresh Kumar 34 / 50

Correlation matrix
The correlation matrix is given by
 
1 ρXY
ρYX 1

It is also symmetric since ρXY = ρYX . In addition, its diagonal

elements are unity.

Dr. Suresh Kumar 35 / 50

Q.1 Let X denote the number of times a photocopy machine
will malfunction: 0, 1, 2 or 3 times, on any given month. Let
Y denote the number of times (0, 1 or 2) a technician is called
on an emergency service. The joint pmf is given as: f (0, 0) =
0.15, f (0, 1) = 0.05, f (0, 2) = 0, f (1, 0) = 0.30, f (1, 1) =
0.15, f (1, 2) = 0.05, f (2, 0) = 0.05, f (2, 1) = 0.05, f (2, 2) =
0.10, f (3, 0) = 0, f (3, 1) = 0.05, and f (3, 2) = 0.05. Find
(i) P(X < Y ), (ii) the marginal pmfs of X and Y , and (iii)
Cov(X , Y ).
Dr. Suresh Kumar 36 / 50
Sol. The given joint pmf in tabular form is
X /Y 0 1 2 fX (x)
0 0.15 0.05 0 0.20
1 0.30 0.15 0.05 0.50
2 0.05 0.05 0.10 0.20
3 0 0.05 0.05 0.10
fY (y ) 0.50 0.30 0.20 1

Dr. Suresh Kumar 37 / 50

(i) To find P(X < Y ), do the sum of probabilities of the pairs
(X , Y ) where X < Y . Such pairs are (0, 1), (0, 2), (1, 2). So
we have
P(X < Y ) = f (0, 1)+(0, 2)+f (1, 2) = 0.05+0+0.05 = 0.1.
(iii) E (X ) = 1.2, E (Y ) = 0.7, E (XY ) = 1.2 and
Cov(X , Y ) = 0.36.

Dr. Suresh Kumar 38 / 50

Q.2 Consider two continuous random variables X and Y with
pdf (
2 2
81 x y , 0 < x < k, 0 < y < k
f (x, y ) =
0, elsewhere
Find (i) k, (ii) P(X > 3Y ), (iii) P(X + Y > 3), and (iv) the
marginal pdfs of X and Y . Are X and Y independent?

Dr. Suresh Kumar 39 / 50

 Z k Z k
2 2
x y = 1 =⇒ k = 3.
0 0 81
Z 3 Z x/3
2 2 1
P(X > 3Y ) = x ydydx = .
0 0 81 15
Z 3Z 3
2 2
P(X + Y > 3) = x ydydx = 0.99.
0 3−x 81
1 2
fX (x) = x 2, fY (y ) = y .
9 9
Dr. Suresh Kumar 40 / 50
Q.3 Consider two continuous random variables X and Y with
pdf
(
k(x + y ), x > 0, y > 0, 3x + y < 3
f (x, y ) =
0, elsewhere

Find (i) k, (ii) P(X < Y ), (iii) the marginal pdfs of X and Y ,
(iii) Cov(X + 2, Y − 3), (iv) Corr(−2X + 3, 2Y + 7), and (iii)
Cov(−2X + 3Y − 4, 4X + 7Y + 5).

Dr. Suresh Kumar 41 / 50

Sol. (i) k = 2,
(ii) P(X < Y ) = 27/32,
(iii) the marginal pdfs of X and Y are
9 3
− 3x + x 2, 0 < x < 1
fX (x) =
4 4
1 1 5
fY (y ) = + y − y 2, 0 < y < 3
4 3 36
Verify E (X ) = 5/16, E (Y ) = 21/16, E (XY ) = 3/10

Dr. Suresh Kumar 42 / 50

Q.4 Consider two continuous random variables X and Y with
pdf (
ke −y , −y < x < y , y > 0
f (x, y ) =
0, elsewhere
Find (i) k, (ii) the marginal pdfs of X and Y , and (iii) the
conditional pdfs of X and Y .
Sol. (i) k = 1/2

Dr. Suresh Kumar 43 / 50

Q.5 Two persons A and B have agreed to meet for lunch
between noon (0:00 pm) to 1:00 pm. Denote A’s arrival time
by X , B’s by Y , and suppose X and Y are independent with
density functions:
(
3x 2, 0 < x < 1,
fX (x) =
0, elsewhere
(
2y , 0 < y < 1,
fY (y ) = .
0, elsewhere
Dr. Suresh Kumar 44 / 50
(i) Find the probability that A arrives before B, and hence
compute the expected amount of time A would have to wait
for B to arrive.
Sol. Since X and Y are independent, their joint pdf is given
by
(
3x 2y , 0 < x < 1, 0 < y < 1
f (x, y ) = fX (x)fY (y ) =
0, elsewhere

(i) P(A arrives before B) = P(X < Y ) = 2/5.

Dr. Suresh Kumar 45 / 50
Next, expected amount of time A would have to wait for B to
arrive is given by E (Y − X ) provided Y > X . Solving the
double integral of (y − x)f (x, y ) over the region
0 < x < 1, 0 < y < 1, y > x, we get E (Y − X ) = 1/12
hours. (Verify!).
(ii) If they have pre-decided on a condition that whoever
comes first will only wait for 15 minutes for the other, what is
the probability that they will meet for lunch?
Hint: P(Y − X < 1/4) + P(X − Y < 1/4).
Dr. Suresh Kumar 46 / 50
Q.6 The following table shows the quality and meal price
ratings (1 lowest to 3 highest) of 300 restaurants in a metro
city:
Quality/Meal Price 1 2 3 Total
1 42 39 3 84
2 33 63 54 150
3 3 15 48 66
Total 78 117 105 300

Dr. Suresh Kumar 47 / 50

Develop a bivariate probability distribution for quality X and
meal price Y of a randomly selected restaurant in the metro
city. Determine Cov(X , Y ) and Cor(X , Y ). Based on your
results, do you suppose it is likely to find a low cost restaurant
with high meal quality.
X/Y 1 2 3 fX (x)
1 0.14 0.13 0.01 0.28
2 0.11 0.21 0.18 0.50
3 0.01 0.05 0.05 0.22
fY (y ) 0.26 0.29 0.35 1
Dr. Suresh Kumar 48 / 50
E (X ) = 1.94, E (Y ) = 2.09, V (X ) = 0.4964,
V (Y ) = 0.6019, Cov(X , Y ) = 0.2854 and
Cor(X , Y ) = 0.5221.

Dr. Suresh Kumar 49 / 50

Q.7 Let T1, T2, ..., Tk be independent exponential random vari-
ables with mean values 1/λ1, 1/λ2, ..., 1/λk , respectively. De-
note Tmin = min(T1, T2, ..., Tk ). Show that Tmin has an expo-
nential distribution. What is the mean of Tmin?
Sol. Since T1, T2, ..., Tk are independent,
P(Tmin > t) =
P(T1 > t)P(T2 > t)....P(Tk > t)

Dr. Suresh Kumar 50 / 50