Continuous & Discrete Random Variables

Ch 4 and 5, Ross

9th October 2022

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 1 / 26
Continuous RVs

While a discrete random variable takes either finite or countably

infinite possible values, a continuous random variable takes
uncountable possible values
A function f (x) with values defined over a set of real numbers is
called a probability density function of the continuous variable X if
and only if
Zb
P(a ≤ X ≤ b) = f (x)dx
a

for any real constants a and b with a ≤ b.

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 2 / 26
Theorem: Probability Density

A function can be considered as the probability density of a continuous

random variable X if and only if its values, f (x), satisfy the conditions
1 f (x) ≥ 0 for −∞ < x < ∞
R∞
2 f (x)dx = 1
−∞

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 3 / 26
So tell me,

How does a probability distribution function of a discrete random

variable look like?

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 4 / 26
So tell me,

How does a probability distribution function of a discrete random

variable look like?

Correspondingly, how would a probability ’density’ function of a

continuous random variable look like?

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 4 / 26
Probability Functions

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 5 / 26
Probability Functions

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 5 / 26
Theorem: Courtesy Continuity

If X is a continuous random variable and a and b are real constants with

a ≤ b, then

P(a ≤ X ≤ b) = P(a ≤ X < b)

= P(a < X ≤ b)
= P(a < X < b)

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 6 / 26
So tell me,

What is a distribution function (cumulative distribution function) of a

discrete random variable?

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 7 / 26
So tell me,

What is a distribution function (cumulative distribution function) of a

discrete random variable?

Correspondingly, what is a distribution function (cumulative

distribution function) of a continuous random variable?

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 7 / 26
Distribution Function

If X is a continuous random variable, and the probability density at x is

f (x), then the function F
Za
F (a) = P(X ≤ a) = f (x)dx
−∞

is called the distribution function or the cumulative distribution function of

X.

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 8 / 26
So tell me,

What is cdf-pdf relation for a discrete random variable?

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 9 / 26
So tell me,

What is cdf-pdf relation for a discrete random variable?

Correspondingly, what is cdf-pdf relation for a continuous random

variable?

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 9 / 26
Theorem: pdf – cdf relation

If f (x) and F (x) are the values of the probability density and distribution
function of X at x then
1 P(a ≤ X ≤ b) = F (b) − F (a) for any real constants a and b with
a≤b
dF (x)
2 f (x) = dx where the derivative exists.
Intuition?

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 10 / 26
Uniform Random Variable

Think of two experiments: toss of a fair coin and roll of a fair die.

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 11 / 26
Uniform Random Variable

Think of two experiments: toss of a fair coin and roll of a fair die.
What do these have in common?

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 11 / 26
Uniform Random Variable

Think of two experiments: toss of a fair coin and roll of a fair die.
What do these have in common?
Both have a fixed number of outcomes. Coin flip: 2. Die roll: 6

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 11 / 26
Uniform Random Variable

Think of two experiments: toss of a fair coin and roll of a fair die.
What do these have in common?
Both have a fixed number of outcomes. Coin flip: 2. Die roll: 6
Probability of each outcome is the same: 1/2 or 1/6

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 11 / 26
Uniform Random Variable

Think of two experiments: toss of a fair coin and roll of a fair die.
What do these have in common?
Both have a fixed number of outcomes. Coin flip: 2. Die roll: 6
Probability of each outcome is the same: 1/2 or 1/6

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 11 / 26
Uniform Random Variable: Discrete

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 12 / 26
Uniform Random Variable: Discrete

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 13 / 26
Uniform Random Variable: Discrete

The shuttle is supposed to leave campus at 11.30 a.m. We know

from experience that it leaves between 11.25 and 11.35 randomly.

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 14 / 26
Uniform Random Variable: Discrete

The shuttle is supposed to leave campus at 11.30 a.m. We know

from experience that it leaves between 11.25 and 11.35 randomly.
The shuttle departure time during any 1-min interval is just as likely
as the next 1-min interval.

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 14 / 26
Uniform Random Variable: Discrete

The shuttle is supposed to leave campus at 11.30 a.m. We know

from experience that it leaves between 11.25 and 11.35 randomly.
The shuttle departure time during any 1-min interval is just as likely
as the next 1-min interval.
This is a discrete uniform distribution.

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 14 / 26
Uniform Random Variable: Discrete

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 15 / 26
Discrete Uniform Distribution: PMF

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 16 / 26
Uniform Random Variable: Continuous
A random variable is said to be uniformly distributed over the interval
(α, β) if it has “equal chances” of getting any value within this interval.
The probability of X falling in a subinterval of (α, β) is equal to the length
of the subinterval, divided by the length of the interval (α, β).

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 17 / 26
Uniform Random Variable: Continuous
A random variable is said to be uniformly distributed over the interval
(α, β) if it has “equal chances” of getting any value within this interval.
The probability of X falling in a subinterval of (α, β) is equal to the length
of the subinterval, divided by the length of the interval (α, β).
Parameters α, β > 0
 1
β−α for α<x <β
PDF f (x) =
0 otherwise
 0 for x < α
x−α
CDF F (x) = for α<x <β
 β−α
1 for x > β
α+β
E (X ) 2

(β−α)2
Var (X ) 12

e tβ −e tα
MGF of X t(β−α)

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 17 / 26
Uniform Random Variable

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 18 / 26
Uniform Random Variable

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 18 / 26
Uniform Random Variable

Zb
1 b−a
P(a < X < b) = dx =
β−α β−α
a

1. If X is uniformly distributed over the interval [0, 10], compute the

probability that (a) 2 < X < 9; (b) X < 5 (c) X > 6
Ans: (a) 7/10; (b) 5/10; (c) 4/10
2. Buses arrive at a stop at 15-min intervals starting at 7 a.m. If a
passenger arrives at the stop at a time that is uniformly distributed
between 7 and 7.30, find the probability that he waits (a) < 5 mins; at
least 12 mins for a bus.
Ans: X is a uniform random variable over [0,30]. X is the time in minutes
past 7 a.m. that the passenger arrives. (a) He will have to wait < 5 mins
if he arrives between 7.10 -7.15 & bet 7.25 - 7.30.
5 5
Hence P(10 < X < 15) + P(25 < X < 30) = 30 + 30 = 31

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 19 / 26
Normal Random Variables
This distribution is most easily recognized, aesthetically pleasing and
famous for its bell-shape. The density function is symmetric about its
mean. The curve is unimodal. It describes many real world phenomenon
like distribution of height, distribution of weight, some psychological test
scores, etc.

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 20 / 26
Normal Random Variables
This distribution is most easily recognized, aesthetically pleasing and
famous for its bell-shape. The density function is symmetric about its
mean. The curve is unimodal. It describes many real world phenomenon
like distribution of height, distribution of weight, some psychological test
scores, etc.
Parameters µ ∈ R, σ 2 > 0
2 2
PDF 1
f (x) = √2πσ e −(x−µ) /2σ −∞<x <∞

CDF Complicated
E (X ) µ

Var (X ) σ2

σ2 t 2
MGF of X exp{µt + 2 }

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 20 / 26
Moment Generating Function of a Normal Random
Variable

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 21 / 26
Mean and Variance of a Normal Random Variable

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 22 / 26
Standard Normal Distribution

If X is normal with mean µ and variance σ 2 , then Y = αX + β is

normal with mean αµ + β and variance α2 σ 2

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 23 / 26
Standard Normal Distribution

If X is normal with mean µ and variance σ 2 , then Y = αX + β is

normal with mean αµ + β and variance α2 σ 2
If X ∼ N(µ, σ 2 ), then Z = X σ−µ is a normal variable with mean 0 and
variance 1. This random variable Z has a standard, or unit normal
distribution.

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 23 / 26
Standard Normal Distribution

If X is normal with mean µ and variance σ 2 , then Y = αX + β is

normal with mean αµ + β and variance α2 σ 2
If X ∼ N(µ, σ 2 ), then Z = X σ−µ is a normal variable with mean 0 and
variance 1. This random variable Z has a standard, or unit normal
distribution.
Probability Distribution Function of a standard normal variable:
2
f (x) = √12π e −(x) /2 −∞<x <∞
We can write probability statements about X in terms of probability
statements about Z.

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 23 / 26
Using Standard Normal Tables

Anita scored 25 in her exam and Ajay scored 1150. Which of the two
scored better on their tests?

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 24 / 26
Using Standard Normal Tables

Anita scored 25 in her exam and Ajay scored 1150. Which of the two
scored better on their tests?
What proportion of students scored worse than Anita and Ajay?
Assume both tests have nearly normal distribution.

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 24 / 26
Using Standard Normal Tables

Anita scored 25 in her exam and Ajay scored 1150. Which of the two
scored better on their tests?
What proportion of students scored worse than Anita and Ajay?
Assume both tests have nearly normal distribution.
Anita’s exam scores: X ∼ N(µ = 20, σ = 4)

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 24 / 26
Using Standard Normal Tables

Anita scored 25 in her exam and Ajay scored 1150. Which of the two
scored better on their tests?
What proportion of students scored worse than Anita and Ajay?
Assume both tests have nearly normal distribution.
Anita’s exam scores: X ∼ N(µ = 20, σ = 4)
Ajay’s exam scoress : Y ∼ N(µ = 1000, σ = 150)

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 24 / 26
Using Standard Normal Tables

Anita scored 25 in her exam and Ajay scored 1150. Which of the two
scored better on their tests?
What proportion of students scored worse than Anita and Ajay?
Assume both tests have nearly normal distribution.
Anita’s exam scores: X ∼ N(µ = 20, σ = 4)
Ajay’s exam scoress : Y ∼ N(µ = 1000, σ = 150)
convert these into z-scores: Anita: 1.25 and Ajay: 1.0

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 24 / 26
Using Standard Normal Tables

Anita scored 25 in her exam and Ajay scored 1150. Which of the two
scored better on their tests?
What proportion of students scored worse than Anita and Ajay?
Assume both tests have nearly normal distribution.
Anita’s exam scores: X ∼ N(µ = 20, σ = 4)
Ajay’s exam scoress : Y ∼ N(µ = 1000, σ = 150)
convert these into z-scores: Anita: 1.25 and Ajay: 1.0

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 24 / 26
Using Standard Normal Tables

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 25 / 26
Exponential Random Variable
Exponential distribution is often associated with the amount of time until
some event occurs. For instance, starting from now until earthquake
occurs, or until a war breaks out, etc.

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 26 / 26
Exponential Random Variable
Exponential distribution is often associated with the amount of time until
some event occurs. For instance, starting from now until earthquake
occurs, or until a war breaks out, etc.
Parameters λ>0 
λe −λx for x ≥0
PDF f (x) =
 0 otherwise
0 for x < 0
CDF F (x) = −λx
1−e for x ≥ 0
1
E (X ) λ

1
Var (X ) λ2

λ
MGF of X λ−t

Ch 4 and 5, Ross Continuous & Discrete Random Variables 9th October 2022 26 / 26