Lectures 5 & 6

Distribution of function of random variables

* Given that 𝑋 is a random variable with 𝑝𝑑𝑓 𝑓(𝑥). Let 𝑌 = 𝑢(𝑋) be a function

of 𝑋. What is the 𝑝𝑑𝑓 of 𝑌 ?

* Given that 𝑋1 and 𝑋2 are two random variables with joint 𝑝𝑑𝑓 𝑓(𝑥1 , 𝑥2 ). Let

𝑌1 = 𝑢1 (𝑋1 , 𝑋2 ) be a function of 𝑋1 and 𝑋2 . What is the 𝑝𝑑𝑓 of 𝑌1 ?

Define another function of 𝑋1 and 𝑋2 (𝑌2 = 𝑢2 (𝑋1 , 𝑋2 ), say). What is the

joint 𝑝𝑑𝑓 of 𝑌1 and 𝑌2 ?

* Given that 𝑋1 , 𝑋2 , … , 𝑋𝑛 are 𝑛 random variables with joint 𝑝𝑑𝑓

𝑓(𝑥1 , 𝑥2 , … , 𝑥𝑛 ). Define 𝑌 = 𝑢1 (𝑋1 , 𝑋2 , … , 𝑋𝑛 ) to be a function of 𝑋1 ,

𝑋2 , … , 𝑋𝑛 . What is the 𝑝𝑑𝑓 of 𝑌 ?

There are three methods to answer such of the above questions:

1. Distribution Function Method

2. Transformation Method

3. Moment generating Function (MGF) Method

Distribution Function Method

Definition: If 𝑋 is a random variable with 𝑝𝑑𝑓 𝑓(𝑥) then the distribution

function (or cumulative distribution function, 𝑐𝑑𝑓) of 𝑋 is defined by

𝐹𝑋 (𝑥) = 𝑃(𝑋 ≤ 𝑥), ∀𝑥

* If 𝑋 is a continuous random variable then

1
 𝑑𝐹𝑋 (𝑥)
𝑓(𝑥) = = 𝐹𝑋′ (𝑥).
𝑑𝑥

Example 1: Let 𝑋 be a random variable with 𝑝𝑑𝑓

2𝑥 , 0 < 𝑥 < 1
𝑓(𝑥) = {
0 , 𝑜. 𝑤

Find

1. 𝑃(𝑋 ≤ −1) 2. 𝑃(𝑋 ≤ 0.6) 3. 𝑃(𝑋 ≤ 3) 4. 𝐹𝑋 (𝑥)

5. the 𝑝𝑑𝑓 of 𝑌 = 2𝑋 + 1 by using the distribution function method.
6. the 𝑝𝑑𝑓 of 𝑌 = 1 − 2𝑋 by using the distribution function method.
Solution:

−1 −1
1. 𝑃(𝑋 ≤ −1) = ∫−∞ 𝑓(𝑥)𝑑𝑥 = ∫−∞ 0 𝑑𝑥 = 0.

0.6
2. 𝑃(𝑋 ≤ 0.6) = ∫−∞ 𝑓(𝑥)𝑑𝑥
0 0.6
= ∫ 0 𝑑𝑥 + ∫ 2𝑥𝑑𝑥
−∞ 0
0.6
= 0 + ∫ 2𝑥𝑑𝑥
0

= 0.62 = 0.36
−3
3. 𝑃(𝑋 ≤ 3) = ∫−∞ 𝑓(𝑥)𝑑𝑥
0 1 3
= ∫ 0𝑑𝑥 + ∫ 2𝑥𝑑𝑥 + ∫ 0𝑑𝑥
−∞ 0 1
1
= 0 + ∫ 2𝑥𝑑𝑥 + 0
0

=1

4. 𝐹𝑋 (𝑥) = 𝑃(𝑋 ≤ 𝑥), ∀ 𝑥 ∈ (−∞, ∞)

2
 0, 𝑥<0
𝐹𝑋 (𝑥) = {𝑥 2 , 0 ≤ 𝑥 < 1
1, 𝑥≥1

5. 𝑌 = 2𝑋 + 1. If 0 < 𝑥 < 1 then 1<𝑦<3

Note that: If you take all values of 𝑥 ∈ (−∞, ∞) then 𝑦 ∈ (−∞, ∞).
To find the 𝑝𝑑𝑓 of 𝑌 by using the distribution function method, we need first
to find the distribution function of 𝑌. Let 𝐺𝑌 (𝑦) be the distribution function of
𝑌 then,
𝐺𝑌 (𝑦) = 𝑃(𝑌 ≤ 𝑦), ∀ 𝑦 ∈ (−∞, ∞)
= 𝑃(2𝑋 + 1 ≤ 𝑦)
𝑦−1
= 𝑃 (𝑋 ≤ )
2
0 , 𝑦<1
𝑦−1 2
= ( ) , 1≤𝑦<3
2
{ 1 , 𝑦≥3
Now, the 𝑝𝑑𝑓 of 𝑌 can be obtained by deriving 𝐺𝑌 (𝑦) with respect to 𝑦. The
𝑝𝑑𝑓 of 𝑌 is,
𝑑𝐺𝑌 (𝑦)
𝑔𝑌 (𝑦) = = 𝐺𝑌′ (𝑦)
𝑑𝑦
𝑦−1
={ 2 , 1<𝑦<3
0 𝑜. 𝑤

6. 𝑌 = 1 − 2𝑋. If 0 < 𝑥 < 1 then −1 < 𝑦 < 1

Again, if 𝑥 ∈ (−∞, ∞) then 𝑦 ∈ (−∞, ∞).
Let 𝐺𝑌 (𝑦) be the distribution function of 𝑌 then,
𝐺𝑌 (𝑦) = 𝑃(𝑌 ≤ 𝑦), ∀ 𝑦 ∈ (−∞, ∞)
= 𝑃(1 − 2𝑋 ≤ 𝑦)
1−𝑦
= 𝑃 (𝑋 ≥ )
2

3
 0 , 𝑦 < −1
1
= ∫1−𝑦2𝑥𝑑𝑥 , −1 ≤ 𝑦 < 1
2
{ 1 , 𝑦≥1
0 , 𝑦 < −1
1−𝑦 2
= 1−( ) , −1 ≤ 𝑦 < 1
2
{ 1 , 𝑦≥1
The 𝑝𝑑𝑓 of 𝑌 is,
𝑑𝐺𝑌 (𝑦)
𝑔𝑌 (𝑦) = = 𝐺𝑌′ (𝑦)
𝑑𝑦
1−𝑦
= { 2 , −1 < 𝑦 < 1
0 , 𝑜. 𝑤
Example 2: Let 𝑋 be a continuous random variable with 𝑝𝑑𝑓 𝑓𝑋 (𝑥) and

distribution function 𝐹𝑋 (𝑥). Find the 𝑝𝑑𝑓 of 𝑌 = 𝑋 2 in terms of 𝑓𝑋 by using the

distribution function method.

Solution: Let 𝐺𝑌 (𝑦) be the distribution function of 𝑌 then,

𝐺𝑌 (𝑦) = 𝑃(𝑌 ≤ 𝑦)
= 𝑃(𝑋 2 ≤ 𝑦)
= 𝑃(−√𝑦 ≤ 𝑋 ≤ √𝑦)

= 𝑃(𝑋 ≤ √𝑦) − 𝑃(𝑋 ≤ −√𝑦)

= 𝐹𝑋 (√𝑦) − 𝐹𝑋 (−√𝑦)
The 𝑝𝑑𝑓 of 𝑌 is,
𝑔𝑌 (𝑦) = 𝐺𝑌′ (𝑦)
1 −1
= 𝑓𝑋 (√𝑦) − 𝑓𝑋 (−√𝑦)
2√𝑦 2 √𝑦
1
= (𝑓𝑋 (√𝑦) + 𝑓𝑋 (−√𝑦))
2 √𝑦

4
Example 3: Let 𝑋 be a random variable with 𝑝𝑑𝑓

1
𝑓𝑋 (𝑥) = {2 , −1 < 𝑥 < 1
0 , 𝑜. 𝑤

Find the 𝑝𝑑𝑓 of 𝑌 = 𝑋 2 .

Solution: If −1 < 𝑥 < 1 then 0 < 𝑦 < 1.

Note that: 𝑥 ∈ (−1,1) = (−1,0) ∪ (0,1). If 𝑥 ∈ (−1,0) then 𝑦 ∈ (0,1) and if

𝑥 ∈ (0,1) then 𝑦 ∈ (0,1).

Let 𝑔𝑌 (𝑦) be the 𝑝𝑑𝑓 of 𝑌, then (by using the result of the last example)

1
𝑔𝑌 (𝑦) = (𝑓𝑋 (√𝑦) + 𝑓𝑋 (−√𝑦)) , 0<𝑦<1
2 √𝑦

1 1 1
= ( + )
2 √𝑦 2 2

1
=
2 √𝑦

Therefore,

1
, 0<𝑦<1
𝑔𝑌 (𝑦) = {2√𝑦
0 , 𝑜. 𝑤

Example 4: Let 𝑋 be a random variable with 𝑝𝑑𝑓

𝑒 −𝑥 , 𝑥 > 0
𝑓𝑋 (𝑥) = { . Note that: 𝑋~exp(1)
0 , 𝑜. 𝑤

Find the 𝑝𝑑𝑓 of 𝑌 = 𝑋 2 .

Solution: If 𝑥 > 0 then 𝑦 > 0.

Let 𝑔𝑌 (𝑦) be the 𝑝𝑑𝑓 of 𝑌, then

5
 1
𝑔𝑌 (𝑦) = (𝑓𝑋 (√𝑦) + 𝑓𝑋 (−√𝑦)) , 𝑦>0
2 √𝑦
1
= (𝑒 −√𝑦 + 0)
2 √𝑦
1
= 𝑒 −√𝑦
2 √𝑦
Therefore,
1
𝑒 −√𝑦 , 𝑦 > 0
𝑔𝑌 (𝑦) = {2√𝑦
0 , 𝑜. 𝑤

Example 5: Let 𝑋 be a random variable with 𝑝𝑑𝑓

𝑥2
𝑓𝑋 (𝑥) = { 3 , −1 < 𝑥 < 2
0 , 𝑜. 𝑤

Find the 𝑝𝑑𝑓 of 𝑌 = 𝑋 2 .

Solution: If −1 < 𝑥 < 2 then 0 < 𝑦 < 4.

Note (1): 𝑥 ∈ (−1,2) = (−1,0) ∪ (0,2). If 𝑥 ∈ (−1,0) then 𝑦 ∈ (0,1) and if

𝑥 ∈ (0,2) then 𝑦 ∈ (0,4). Therefore, if 𝑥 ∈ (−1,2) = (−1,0) ∪

(0,2) then 𝑦 ∈ (0,1) ∪ (0,4) = (0,4)

Let 𝑔𝑌 (𝑦) be the 𝑝𝑑𝑓 of 𝑌, then

1
𝑔𝑌 (𝑦) = (𝑓𝑋 (√𝑦) + 𝑓𝑋 (−√𝑦)) , 0<𝑦<4
2 √𝑦

Note (2): If 0 < 𝑦 < 4 then 0 < √𝑦 < 2 and −2 < −√𝑦 < 0.
(√𝑦)2
Note (3): If 0 < √𝑦 < 2 then 𝑓𝑋 (√𝑦) = .
3

6
 0, −2 < −√𝑦 < −1
Note (4): If −2 < −√𝑦 < 0 then 𝑓𝑋 (−√𝑦) = {(−√𝑦)2
, −1 < −√𝑦 < 0
3

Therefore, we need to divide the range of 𝑦 and the 𝑝𝑑𝑓 of 𝑌 is given as

follows:
2 2
1 (√𝑦) (−√𝑦)
( + ), 0 < 𝑦 < 1
2 √𝑦 3 3
𝑔𝑌 (𝑦) = 2
1 (√𝑦)
( + 0) , 1<𝑦<4
2 √𝑦 3
{ 0 , 𝑜. 𝑤
√𝑦
, 0<𝑦<1
3
= √𝑦
, 1<𝑦<4
6
{0 , 𝑜. 𝑤
Example 6: Let 𝑋1 , 𝑋2 be a random sample (𝑋1 , 𝑋2 is a random sample means

𝑋1 , 𝑋2 are independent and identically distributed (iid)) from 𝑒𝑥𝑝(1).

1. Find 𝑃(𝑋1 + 𝑋2 ≤ 2)

2. Use the distribution function method to find the 𝑝𝑑𝑓 of 𝑌1 = 𝑋1 + 𝑋2 .

3. Find 𝑃(𝑋1 − 𝑋2 ≤ −3)

4. Find 𝑃(𝑋1 − 𝑋2 ≤ 4)

5. Use the distribution function method to find the 𝑝𝑑𝑓 of 𝑌2 = 𝑋1 − 𝑋2

Solution: You need to remember that the joint 𝑝𝑑𝑓 of the two random variables
𝑋1 , 𝑋2 (refer to STAT 211) is,
𝑓(𝑥1 , 𝑥2 ) = 𝑓(𝑥1 )𝑓(𝑥2 )
= 𝑒 −𝑥1 𝑒 −𝑥2
= 𝑒 −(𝑥1+𝑥2) , 𝑥1 > 0, 𝑥2 > 0

7
1. 𝒙𝟐

2 X1+X2 = 2

X1+X2≤ 2

2 𝒙𝟏

𝑃(𝑋1 + 𝑋2 ≤ 2) = ∫ ∫ 𝑓(𝑥1 , 𝑥2 )𝑑𝑥1 𝑑𝑥2

𝑥1 +𝑥2 ≤2
2 2−𝑥2
=∫ ∫ 𝑒 −(𝑥1 +𝑥2) 𝑑𝑥1 𝑑𝑥2
0 0
3
=1− = 0.593994
ⅇ2

2. 𝑦1 = 𝑥1 + 𝑥2 . Because 𝑥1 > 0 and 𝑥2 > 0 then 𝑦1 > 0. Let 𝐺𝑌1 (𝑦1 ) be

the distribution function of 𝑌1 , then

. 𝒙𝟐

y1 X1+X2 = y1, y1 > 0

X1+X2≤ y1

y1 𝒙𝟏

8
 𝐺𝑌1 (𝑦1 ) = 𝑃(𝑌1 ≤ 𝑦1 ),
= 𝑃(𝑋1 + 𝑋2 ≤ 𝑦1 )

= ∫∫ 𝑓(𝑥1 , 𝑥2 )𝑑𝑥1 𝑑𝑥2

𝑥1 +𝑥2 ≤𝑦1
𝑦1 𝑦1 −𝑥2
=∫ ∫ 𝑒 −(𝑥1+𝑥2) 𝑑𝑥1 𝑑𝑥2
0 0

= 𝑒−𝑦1 (−1 − 𝑦1 + 𝑒𝑦1 ) , 𝑦1 > 0 .

The 𝑝𝑑𝑓 of 𝑌1 is,

𝑑𝐺𝑌1 (𝑦1 )
𝑔𝑌1 (𝑦1 ) =
𝑑𝑦1
𝑦 𝑒−𝑦1 , 𝑦1 > 0
={ 1
0 , 𝑜. 𝑤

Note that: 𝑌1 ~ Γ(2, 1)

3. 𝒙𝟐

X1-X2 = -3

-3 𝒙𝟏

𝑃(𝑋1 − 𝑋2 ≤ −3) = ∫ ∫ 𝑓(𝑥1 , 𝑥2 )𝑑𝑥1 𝑑𝑥2

𝑥1 −𝑥2 ≤−3
∞ −3+𝑥2
=∫ ∫ 𝑒 −(𝑥1 +𝑥2 ) 𝑑𝑥1 𝑑𝑥2
3 0
1
= = 0.0248935
2ⅇ 3

9
 4. 𝒙𝟐 X1 = 4

X1-X2 = 4

𝒙𝟐 = 𝟎
4 𝒙𝟏

-4
𝒙𝟏 = 0

𝑃(𝑋1 − 𝑋2 ≤ 4) = ∫ ∫ 𝑓(𝑥1 , 𝑥2 )𝑑𝑥1 𝑑𝑥2

𝑥1 −𝑥2 ≤4
∞ 4 ∞ 4+𝑥2
−(𝑥1 +𝑥2 )
=∫ ∫ 𝑒 𝑑𝑥1 𝑑𝑥2 + ∫ ∫ 𝑒 −(𝑥1+𝑥2) 𝑑𝑥1 𝑑𝑥2
0 0 0 4
1
=1− = 0.990842
2ⅇ 4

5. 𝑦2 = 𝑥1 − 𝑥2 . Because 𝑥1 > 0 and 𝑥2 > 0 then −∞ < 𝑦2 < ∞. Let 𝐺𝑌2 (𝑦2 )

be the distribution function of 𝑌2 , then (Refer to Parts 3 and 4 of this example

to further understand the steps below).

10
 𝒙𝟐

X1-X2 = y1, y1<0

-y1 X1-X2 = y1, y1>0

y1 y1 𝒙𝟏

- y1

𝐺𝑌2 (𝑦2 ) = 𝑃(𝑌2 ≤ 𝑦2 ),

= 𝑃(𝑋1 − 𝑋2 ≤ 𝑦2 )
= ∫ ∫𝑥 𝑓(𝑥1 , 𝑥2 )𝑑𝑥1 𝑑𝑥2 , −∞ < 𝑦2 < ∞.
1 −𝑥2 ≤𝑦2
∞ 𝑦2 +𝑥2
∫ ∫ 𝑒 −(𝑥1+𝑥2) 𝑑𝑥1 𝑑𝑥2 , 𝑦2 < 0
−𝑦2 0
= ∞ 𝑦2 ∞ 𝑦2 +𝑥2
−(𝑥1 +𝑥2 )
∫ ∫ 𝑒 𝑑𝑥1 𝑑𝑥2 + ∫ ∫ 𝑒 −(𝑥1 +𝑥2 ) 𝑑𝑥1 𝑑𝑥2 , 𝑦2 > 0
{ 0 0 0 𝑦2
𝑦2
𝑒
, 𝑦2 < 0
2
={ −𝑦 .
𝑒 2
1− , 𝑦2 > 0
2
The 𝑝𝑑𝑓 of 𝑌2 is,
𝑑𝐺𝑌1 (𝑦1 )
𝑔𝑌1 (𝑦1 ) =
𝑑𝑦1
𝑦
𝑒 2
, 𝑦2 < 0
= { −𝑦22
𝑒
, 𝑦2 > 0
2
|𝑦 |
𝑒 2
= , −∞ < 𝑦2 < ∞
2

11
 Questions

Q1: Let 𝑋 be a random variable with 𝑝𝑑𝑓 𝑁(0,1). Find the 𝑝𝑑𝑓 of 𝑌 = 𝑋 2 .

Q2: Let 𝑋1 , 𝑋2 be a random sample from 𝑈(0,1).

1. Find 𝑃(𝑋1 + 𝑋2 ≤ 2) 2. Find 𝑃(𝑋1 + 𝑋2 ≤ 0)

3. Find 𝑃(𝑋1 + 𝑋2 ≤ 0.5) 4. Find 𝑃(𝑋1 + 𝑋2 ≤ 1.5)

5. Use the distribution function method to find the 𝑝𝑑𝑓 of 𝑌1 = 𝑋1 + 𝑋2 .

6. Find 𝑃(𝑋1 − 𝑋2 ≤ −0.5) 7. Find 𝑃(𝑋1 − 𝑋2 ≤ 0)

8. Find 𝑃(𝑋1 − 𝑋2 ≤ 0.4)

9. Use the distribution function method to find the 𝑝𝑑𝑓 of 𝑌2 = 𝑋1 − 𝑋2

10. Use the distribution function method to find the 𝑝𝑑𝑓 of 𝑌3 = 𝑋1 𝑋2

Q3: The joint 𝑝𝑑𝑓 of 𝑋1 , 𝑋2 is

0.5 , 0 < 𝑥1 < 1, 0 < 𝑥2 < 2

𝑓(𝑥) = {
0 , 𝑜. 𝑤

Use the distribution function method to find the 𝑝𝑑𝑓 of 𝑌1 = 𝑋1 + 𝑋2 .

12