Digital Signal Processing Reference
In-Depth Information
For the interval
1< x <a
, we have:
PfX xg¼
0
for
1< x <a:
(2.29)
From (
2.27
)to(
2.29
), we arrive at:
8
<
0
for
1< x <
2
;
x
2
4
F
X
ðxÞ¼
(2.30)
for
2
x
6
;
:
1
for
6
< x <1:
This distribution is shown in Fig.
2.10b
. Note that this distribution function
does not have jumps like the distribution of a discrete variable (
2.24
).
The distribution of a mixed random variable has one or more jumps and the
continuous part. To this end, it is possible to write the distribution of a mixed
random variable [THO71, pp. 50-51] as:
F
X
ðxÞ¼aF
c
ðxÞþF
d
ðxÞ;
(2.31)
where
F
c
is a distribution of the continuous variable and
F
d
is a distribution of the
discrete random variable, and
0
a
1
;
(2.32)
a ¼
1
X
i
PfX ¼ x
i
g:
(2.33)
The proof can be found in [THO71, pp. 50-51].
A distribution where
a ¼
0 corresponds to a discrete r.v., while a distribution
where
a ¼
1 corresponds to a continuous r.v.
Example 2.2.4
In this example, we will illustrate the distribution of a mixed
random variable
X
; which has one discrete value in
x ¼
2 with the probability
P
{2}
¼
1/3 and the continuous interval 2
< x <
6 (as shown in Fig.
2.11a
).
Fig. 2.11
Illustration of the distribution in Example 2.2.4
Search WWH ::
Custom Search