Definition 2.1. Let S be a sample space and B a σ -field of subsets of S.

A function X : S → IR is
called a random variable if the inverse images under X of all semi-closed intervals of the form (−∞,
x],
where x ∈ IR are events in B, i.e.
X −1 (−∞, x] = { ω ∈ S : X( ω ) ≤ x} ∈ B

Definition 2.2. Let (S, B, P) be a probability space and X be a random variable defined on S. The
function F : IR → IR defined by
F(x) = P(−∞, x] = P { ω ∈ S : X( ω ) ≤ x} for all x ∈ IR
(2.1)
is called the distribution function or the cumulative distribution function (CDF) of the random
variable
X.
Remark 2.4. It is a common practice to express Definition (2.1) in an abbreviated form as
F(x) = P{X ≤ x}

Theorem 2.2. Let F be the distribution function of a random variable X defined on a probability
space
(S, B, P). Then the following properties hold:
(i) F is non-decreasing, i.e. F(x) ≤ F(y) whenever x < y.(ii) F is right-continuous, i.e.
F(x + 0) = lim F(x + h) = F(x) for all x ∈ IR.
h→0
(iii) F(−∞) = lim F(x) = 0 and F(+∞) = lim F(x) = 1
x→−∞ x→+∞

Theorem 2.3. If F : IR → IR is a function which is non-decreasing, right-continuous, F(−∞) = 0 and

F(+∞) = 1, then there exists a unique probability function P defined on the Borel σ -field B
generated
by the semi-closed intervals of IR such that
F(x) = P{X ≤ x} = P{ ω : X( ω ) ≤ x}.
Combining Theorems 2.2 and 2.3, we have the following characterizing result for distribution func-
tions (CDF).
Theorem 2.4. A function F : IR → IR is a distribution function for some probability space (S, B, P)
if
and only if F satisfies the following properties:
(i) F is non-decreasing.
(ii) F is right-continuous.
(iii) F(−∞) = 0 and F(+∞) = 1.

Theorem 2.5. Let F be the distribution function of a random variable X defined on a probability
space
(S, B, P). Then,
F(x) − F(x − 0) = P{X = x}
(2.2)
where F(x − 0) = lim F(x − h)
h→0s

Theorem 2.6. A necessary and sufficient condition for a distribution function F of a random variable
X to be continuous is that
P{X = x} = 0 for all x ∈ IR

Theorem 2.7. The set of points of discontinuity of a distribution function F is countable.

Proof. By Theorem 2.2, the distribution function F is a non-decreasing function on the real line. It is
a standard result in Real Analysis that the set of points of discontinuity of every monotone function
(non-decreasing or non-increasing) is countable.

Definition 2.3. A random variable (RV) X defined on a sample space S is called discrete if X can
take
on at most a countable number of possible values, say, x 1 , x 2 , . . . , x n , . . . with p i = P(X = x i )
> 0 for
each i and satisfying
∑ p i = ∑ P{X = x i } = 1
i
i
The points x i of S having positive probability (mass) are said to be mass points of the discrete
random
variable X.

Definition 2.4. Let X be a discrete RV defined on a sample space S with mass points x 1 , x 2 , . . . ,
xn,...
and p i = P(X = x i ) for i = 1, 2, 3, . . .. Then the function f : IR → IR defined by
(
P(X = x i ) = p i if x = x i for some i
f (x) =
0
if x 6 = x i for any i
is called the probability mass function (PMF) of the discrete RV X. (Sometimes, we also express f
as
f X when it is necessary to emphasize that f is the PMF of the discrete RV X.)
Definition 2.5. If X is a discrete RV defined on a sample space S with mass points x 1 , x 2 , . . . , x
n,...
and p i = P(X = x i ) for i = 1, 2, 3, . . ., then the set of ordered pairs
{(x 1 , p 1 ), (x 2 , p 2 ), . . . , (x n , p n )}
is called the probability distribution of the discrete RV X.

Theorem 2.8. Let X be a discrete random variable defined on a sample space S. Then the
distribution
function F of X can be obtained from the probability mass function f of X, and conversely.
Proof. Let X be a discrete random variable with mass points x 1 , x 2 , . . . , x n , . . .
If f is the probability mass function of X, then
P(X = x i ) = f (x i ) for i = 1, 2, . . .
Hence, the distribution function of X is obtained as
F(x) = P{X ≤ x} =
∑
f (x i )
x i ≤x
Conversely, if F is the distribution function of X, then the probability mass function f of X is
obtained as
1⁄2
F(x i ) − F(x i − 0) if x = x i for some i
f (x) =
0
if x 6 = x i for any i
where, by Theorem 2.5, it follows that
P(X = x) = F(x) − F(x − 0) for any x ∈ IR

Theorem 2.9. If f : IR → IR is a function with range [0, 1] having the properties:

(i) f (x) = 0 for all x except for a countable collection of points x 1 , x 2 , . . . , x n , . . . for which f
(x) > 0
(ii) ∑ f (x i ) = 1
i
then f will be a probability mass function for some discrete RV X.

Definition 2.6. Let X be a random variable with distribution function (CDF) F. Then X is called a
continuous random variable if F is a continuous function on IR.

Definition 2.6. Let X be a random variable with distribution function (CDF) F. Then X is called a
continuous random variable if F is a continuous function on IR.
Remark 2.6. By Theorem 2.6, a necessary and sufficient condition for the distribution function F to
be continuous on IR is that
P{X = x} = 0 for each x ∈ IR
(2.3)
This shows that P{X = x} = 0 whenever X is a continuous random variable. Thus, for a continuous
random variable, probability is not concentrated at any “mass points” as in the case of a discrete
random
variable. Another consequence of Eq. (2.3) is that for a continuous random variable X, the
probabilities
P {a < X ≤ b} , P {a ≤ X < b} , P {a ≤ X ≤ b} , P {a < X < b} are all equal.
Definition 2.7. Let X be a continuous random variable with the distribution function F. We say that
the function f : IR → IR is a probability density function (PDF) for F (or X) if f (x) ≥ 0 for all x ∈
IR
such that
Zx
f (t) dt
F(x) =
−∞
(2.4)

Definition 2.9. Let X be a random variable with the expected value (or mean) μ . Then the variance
of
X, denoted by Var(X), is defined as
¤
£
Var(X) = E (X − μ ) 2
and the standard deviation of X, denoted by σ X or σ , is defined as
p
σ X = + Var(X).