Discrete Probability Distribution

Discrete Uniform Probability Distribution

We may have a situation where the probabilities of each

event are the same. For example, if we roll a fair die, we
assume that the probability of obtaining each number is 1
6
If X is the random variable (r.v.) “ the number showing” ,
the probability distribution table is
x 1 2 3 4 5 6
1 1 1 1 1 1
P(X = x)
6 6 6 6 6 6
The probability distribution function (p.d.f.) is
1
P( X  x)  , x  1, 2, 3, 4, 5, 6
6
The distribution with equal probabilities is called “uniform”
A diagram for the distribution
1
P( X  x)  , x  1, 2, 3, 4, 5, 6
6
looks like this:
p
1
6

1 2 3  4 5 6 x

The mean value of X is given by the average of the 1st and last
values of x, so,
1 6
  35
2
However, we could also use the formula for the mean of any
discrete distribution of a random variable:
   xf ( x)
1
For f ( x)  P( X  x)  , x  1, 2, 3, 4, 5, 6
6
we would get
1 1 1
  1  2   . . .  6 
6 6 6
1
 21
6
 35
The Variance of the Uniform Distribution
We can find the variance for any discrete random variable X using

Var ( X )  
2
  x 2 f ( x)   2
e.g. The random variable X has p.d.f. given by
1
f ( x)  P( X  x)  , x  1, 2, 3, 4, 5, 6
6
1 1 1
So, Var ( X )  1   2  ... 6   
2 2 2 2

6 6 6
We found earlier that   3·5, so
91
Var ( X )   3  52  2  92
6
The Bernoulli Process
Bernoulli process must possess the following
properties:
1. The experiment consists of repeated trials.

2. Each trial results in an outcome that may be classified as a

success or a failure.

3. The probability of success, denoted by p, remains constant

from trial to trial.

4. The repeated trials are independent.

 Binomial Distribution

The number X of successes in n Bernoulli trials is called a binomial

random variable.
The probability distribution of this discrete random variable is called
the binomial distribution and its values will be denoted by b(x; n, p)
where n is number of trials and p is probability of success at each
trial
A Bernoulli trial can result in a success with probability p and a
failure with probability q = 1−p. Then the probability distribution
of the binomial random variable X, the number of successes in n
independent trials, is
n x
f ( x)    p (1  p) n  x
 x
 Examples
– A coin is flipped 10 times. What is the probability that
exactly we will get 4 head?
– Flip a fair coin 10 times. X=4 heads

10   1   1   10   1 
4 6 10

f ( x)              210 *.00097  0.205

 4  2   2   4  2 

– Die rolled for 3 times. What is the probability that at least

once 6 will result?

  1   5 
0 3
3
f ( x)  1         1  0.5787  0.4213
0 6   6 
 Examples
– Twelve pregnant women selected at random, take a home
pregnancy test. This test give correct result with 0.8
probability. What is the probability that 10 women find a
correct result?
12 
f ( x)     0.8   0.2   66* 0.107 * 0.04  0.283
10 2

10 

– Random guessing on a multiple choice exam. 25

questions. 4 answers per question. A person get pass
marks if he correctly guesses at least 15. What is the
probability that a person who does not know correct
answer of any question will get a pass marks?
 25   25 
f ( x)     0.25   0.75   ...     0.25   0.75 
15 10 25 0

15   25 
• The binomial distribution derives its name from the fact that
the n + 1 terms in the binomial expansion of (q+p)n
correspond to the various values of b(x; n, p) for x = 0, 1, 2, . . .
, n. That is
 Finding Mean and Variance
• Let the outcome of jth trial be represented by indicator
variable Ij which assumes the value 0 and 1 with probabilities
q and p
• In binomial experiment number of success can be written as
the sum of the n independent indicator variable
• X= I1 + I2 +…+ In
• Mean of Ij = [x • P(x)] = 0.q + 1.p = p
• µ = E(X) = E(I1)+E(I2) +… +E(In)=p+p+…+p=np
2
•  Ij =E[(Ij – p)2]=E(Ij2) – p2 = (02)q+(12)p- p2 =p(1-
p)=pq
• For n independent variable variance would be
pq+pq+…+pq=npq
 Python for Binomial Distribution
• You are tossing 20 times, probability of head is 0.5. Now
you are repeating this experiment for 50 times. How many
head appears each time?
from numpy import random, mean
x = random.binomial(n=20, p=0.5, size=50)
mean_head = mean(x)
print(x)
print (mean_head)
[10 11 10 9 8 10 9 9 7 12 12 8 10 12 9 11 12 8 11 14 10 11 8 9
8 11 14 10 11 9 6 10 11 9 11 10 13 15 10 13 9 10 7 10 6 13 6 7
16 9]
10.08
 Visualization
from numpy import random
import matplotlib.pyplot as plt
import seaborn as sns
sns.distplot(random.binomial(n=20, p=0.5, size=50), kde=False)
plt.show()
 Areas of Application
– Quality control measure in industrial process

– In epidemiology: Yes/no outcomes (dead/alive,

treated/untreated, sick/well, etc.)

– In military application: like if a missile can successfully hit a

location with probability p then what is the probability that
missile hits the target k times out of n times.
 Effect of n and p on shape
• For small n and small p, the binomial distribution is what we call skewed
right / positively skewed. That is, the bulk of the probability falls in the
smaller numbers 0, 1, 2,..., and the distribution tails off to the right. For
example, here's a picture of the binomial distribution when n = 15 and p =
0.2:
• For small n and large p, the binomial distribution is what we call skewed
left / negatively skewed. That is, the bulk of the probability falls in the
larger numbers n, n−1, n−2,... and the distribution tails off to the left. For
example, here's a picture of the binomial distribution when n = 15 and p =
0.8:
• For p = 0.5 and large and small n, the binomial distribution is what we
call symmetric. That is, the distribution is without skewness. For example,
here's a picture of the binomial distribution when n = 15 and p = 0.5:
• For large n and small p, the binomial distribution approaches symmetry.
For large n, the distribution is nearly symmetric. For example, here's a
picture of the binomial distribution when n = 40 and p = 0.2:
 Multinomial Distribution
• The binomial experiment becomes a multinomial experiment if we let
each trial have more than two possible outcomes.
• In general, if a given trial can result in any one of k possible outcomes E1,
E2, . . . , Ek with probabilities p1, p2, . . . , pk, then the multinomial
distribution will give the probability that E1 occurs x1 times, E2 occurs x2
times, . . ., and Ek occurs xk times in n independent trials, where
x1 + x2 + · · · + xk = n.
• We shall denote this joint probability distribution by
f(x1, x2, . . . , xk; p1, p2, . . . , pk, n).
• Clearly, p1 + p2 + · · · + pk = 1, since the result of each trial must be one of
the k possible outcomes.
• For a certain airport with three runways, it is known that in the ideal
setting the following are the probabilities that the individual runways are
accessed by a randomly arriving
• commercial jet:
• Runway 1: p1 = 2/9,
• Runway 2: p2 = 1/6,
• Runway 3: p3 = 11/18.
• What is the probability that 6 randomly arriving airplanes are distributed
in the following fashion?
• Runway 1: 2 airplanes,
• Runway 2: 1 airplane,
• Runway 3: 3 airplanes
Distribution of event when probability
of events are given
from numpy import random
x = random.multinomial(n=100, pvals=[1/6, 1/6, 1/6, 1/6, 1/6, 1/6])
print(x)
[ 9 13 16 19 19 24]
 Visualization
from numpy import random
import matplotlib.pyplot as plt
import seaborn as sns
sns.distplot(random.multinomial(n=100, pvals=[1/6, 1/6, 1/6, 1/6, 1/6,
1/6]), kde=False)
plt.show()
 Negative Binomial and Geometric
Distributions
• Experiment is repeated until fixed number of success
occurs.
• Instead of k success in nth trial, we are now
interested at finding number of trials needed for k
success. Out of k success last success must occur at
last trial.
• consider the use of a drug that is known to be effective in 60% of the
cases. Find out the probability that we will have 5th success at 7th attempt?

• Probability of 5 success and 2 failures = (0.6)5*(0.4)2

• here are many possible arrangements of success and failure, however, last
attempt must be a success. So from the first six attempts there must be 4
success and 2 failures.

• P(X=7) = 6C4* (0.6)5*(0.4)2=0.1866

• What Is the Negative Binomial Random Variable?
• The number X is the number of trials required to produce k successes in a
negative binomial experiment is called a negative binomial random
variable and its probability distribution is called the negative binomial
distribution

• If repeated independent trials can result in a success with probability p

and a failure with probability q = 1 − p, then the probability distribution of
the random variable X, the number of the trial on which the kth success
occurs, is
• In an NBA (National Basketball Association) championship series, the team
that wins four games out of seven is the winner. Suppose that teams A and
B face each other in the championship games and that team A has
probability 0.55 of winning a game over team B.

• (a) What is the probability that team A will win the series in 6 games?

• (b) What is the probability that team A will win the series?
a) b∗(6; 4, 0.55) =5C3*0.554(1 − 0.55)6−4 = 0.1853

b) P(team A wins the championship series) is

b∗(4; 4, 0.55) + b∗(5; 4, 0.55) + b∗(6; 4, 0.55) + b∗(7; 4, 0.55)
= 0.0915 + 0.1647 + 0.1853 + 0.1668 = 0.6083.
 Number of failures before nth success
when success probability is p
from numpy import random, mean
x = random.negative_binomial(n=5, p=0.2, size=50)
mean_trial = mean(x)
print(x)
print (mean_trial)
• [12 17 11 45 18 39 5 13 10 19 33 50 27 12 27 13
14 3 27 20 20 15 21 40 13 50 9 9 49 31 22 8 14 22
8 17 33 7 27 15 5 13 18 27 24 36 6 22 10 16]
20.44
• Average Number of experiments = 20.44+1=21.44
 Visualization
• from numpy import random
• import matplotlib.pyplot as plt
• import seaborn as sns
• sns.distplot(random.negative_binomial(n=5, p=0.2, size=50), kde=False)
• plt.show()
 Geometric Distribution
• If we consider the special case of the negative binomial distribution where
k = 1, we have a probability distribution for the number of trials required
before a single success.
• b∗(x; 1, p) = pqx−1, x= 1, 2, 3, . . . .
• Since the successive terms constitute a geometric progression, it is
customary to refer to this special case as the geometric distribution and
denote its values by g(x; p)

• If repeated independent trials can result in a success with probability p

and a failure with probability q = 1 − p, then the probability distribution of
the random variable X, the number of the trial on which the first success
occurs, is
• g(x; p) =pqx−1, x= 1, 2, 3, . . . .
• For a certain manufacturing process, it is known that, on an average, 1 in
every 100 items is defective. What is the probability that the fifth item
inspected is the first defective item found?

• Using the geometric distribution with x = 5 and p = 0.01, we have

• g(5; 0.01) = (0.01)(0.99)4 = 0.0096.
 Number of experiments required for
first success when success prob is p
from numpy import random, mean
x = random.geometric(p=0.3, size=30)
mean_experiment = mean(x)
print(x)
print (mean_experiment)
[ 3 1 1 2 2 4 4 2 2 5 6 1 6 10 1 1 1 4 1 4 2 2 7 6 2 2 1 5 9 7]
3.466666666666667
• Draw one thousand values from the geometric
distribution, with the probability of an individual success
equal to 0.35. What is the probability of success after a
single run?
• import numpy as np
• z = np.random.geometric(p=0.35, size=1000)
• sum_item = 0
• for item in z:
• if item == 1:
• sum_item +=1;
• prob = sum_item/(1000.0)
• print(prob)
0.354
Visualization of geometric distribution
• from numpy import random
• import matplotlib.pyplot as plt
• import seaborn as sns
• sns.distplot(random.geometric(p=0.3, size=30), kde=False)
• plt.show()
 Properties of the Poisson Process
• The number of outcomes occurring in one time interval or specified region of
space is independent of the number that occur in any other disjoint time
interval or region. In this sense we say that the Poisson process has no
memory.

• The probability that a single outcome will occur during a very short time
interval or in a small region is proportional to the length of the time interval or
the size of the region.

• The probability that more than one outcome will occur in such a short time
interval or fall in such a small region is negligible.

• The number X of outcomes occurring during a Poisson experiment is called a

Poisson random variable, and its probability distribution is called the Poisson
distribution.
• The mean number of outcomes is computed from μ = λ, Since the
probabilities depend on λ, the rate of occurrence of outcomes, we
shall denote them by p(x; λ).
• The probability distribution of the Poisson random variable X,
representing the number of outcomes occurring in t time interval
with mean rate λ

e (   t ) ( t ) x
p( x, t ) 
x!

• where λ is the average number of outcomes for the unit time

interval, distance, area, or volume.
• Arrivals of bus at a bus-stop follow a Poisson distribution with an
average of 18 bus every hour. Calculate the probability of fewer
than 3 arrivals in a quarter of an hour.

e (   t ) ( t ) x
p ( x,  t ) 
x!

• λt=4.5, P(0,4.5)= 0.01111, p(1,4.5)=0.04999 and p(2,4.5)=0.11248

• So the probability of fewer than 3 arrivals in a quarter of an hour

time is 0.01111+ 0.04999 + 0.11248 =0.17358
• During a laboratory experiment, the average number of
radioactive particles passing through a counter in 1
millisecond is 4. What is the probability that 6 particles enter
the counter in a given millisecond?

• p(6; 4) =e−446/6!= 0.1042

• Both the mean and the variance of the Poisson distribution

p(x; λ) are λ.
 Average outcome due to Poisson
process
from numpy import random, mean
x = random.poisson(lam=2, size=100)
mean_experiment = mean(x)
print(x)
print (mean_experiment)
[3 1 1 1 2 6 1 1 1 2 1 2 3 2 2 3 1 2 3 3 2 1 0 2 3 4 2 3
55121041323301305412203102222
32235112201324413231513215011
1 0 2 2 3 1 3 1 1 2 1 1 1 1]
2.02
• from numpy import random
import matplotlib.pyplot as plt
import seaborn as sns
sns.distplot(random.poisson(lam=2, size=1000), kde=False)
plt.show()
Poisson Density Function for different
Mean
 Approximation of Binomial
Distribution by a Poisson Distribution
• If n is large and p is close to 0, the Poisson distribution can be used,
with μ = np, to approximate binomial probabilities
• If p is close to 1, we can still use the Poisson distribution to
approximate binomial probabilities by interchanging what we have
defined to be a success and a failure, thereby changing p to a value
close to 0.
• Binomial situation, n= 100, p=0.075. Calculate the probability
of fewer than 10 successes.
• Using binomial distribution value is 0.7832687
• The Poisson approximation to the Binomial states that  will
be equal to np, i.e. 100 x 0.075 so =7.5
• Using Poisson distribution value is 0.7764076
• In a certain industrial facility, accidents occur infrequently. It is
known that the probability of an accident on any given day is 0.005
and accidents are independent of each other.
a) What is the probability that in any given period of 400 days there
will be an accident on one day?
b) What is the probability that there are at most three days with an
accident?
• Let X be a binomial random variable with n = 400 and p = 0.005.
Thus, np = 2. Using the Poisson approximation,
a) P(1,2) = e−221 = 0.271 and
b) P(X ≤ 3,2) = 3𝑥=0 e−22x/x! =0.857.
 Area of Application

• Number of deaths due to some epidemic

• Number of faulty item produced in a year

• Number of days it is going to rain

 Exercise
• The probability that a student pilot passes the written test for
a private pilot’s license is 0.7. Find the probability that a given
student will pass the test
• (a) on the third try;
• (b) before the fourth try.
• Using the geometric distribution
a) P(X = 3) = g(3; 0.7) = (0.7)(0.3)2 = 0.0630.
− =
b) P(X < 4) = 3𝑥=1 𝑔 𝑥, 0.7 = 3𝑥=1 0.7 0.3 𝑥 1 0.973
 Exercise
• On an average, 3 traffic accidents per month occur at a certain
intersection. What is the probability that in any given month
at this intersection
• exactly 5 accidents will occur?
• fewer than 3 accidents will occur?
• Using the Poisson distribution with x = 5 and λt = 3,
p(5; 3) =e-3*35/5!=0.1008
• P(X < 3) = P(X ≤ 2) = p(0,3)+p(1,3)+p(2,3)=0.4232.