Probability Generating Functions
Probability Generating Functions
Topic assessment
1. The probability generating function of the discrete random variable X is given by
GX (t ) k (1 12 t )3
(i) Find the value of k. [2]
(ii) Find P(X = 2) [2]
(iii) Find E(X) and Var(X). [6]
(iv) Find the probability generating function of 3X + 2. [2]
2. The random variable X has the Poisson distribution with parameter , so that
e x
P( X x) , x = 0, 1, 2, ...
x!
(i) Derive the probability generating function of X. [3]
(ii) Hence obtain the mean and variance of X. [5]
(iii) The random variable Y is independent of X and has the Poisson distribution with
parameter . Use probability generating functions to find the distribution of X + Y.
[4]
(iv) Use the distributions of X, Y and X + Y to find the conditional probability that X = x
given that X + Y = n, where n is a non-negative integer. Deduce that the conditional
distribution of X given that X + Y = n is binomial with parameters n and . [6]
(MEI)
Total 44 marks
1 of 4 17/01/19 © MEI
integralmaths.org
Edexcel FS1 PGFs Assessment solutions
1. (i) G X (t ) k(1 21 t )3
GX (1) 1 k(1 21 )3 1 27
8 k 1 k 8
27
[2]
(ii) G X (t ) 8
27 (1 t ) 1
2
3
8
27 1 3( 1
2 t ) 3( 21 t )2 ( 21 t )3
The coefficient of t 2 is 8
27 3( 21 )2 8
27 43 2
9
so P( X 2) 2
9 [2]
(iii) G X (t ) 8
27 (1 21 t )3
GX (t ) 8
27 3(1 21 t )2 21 49 (1 21 t )2
GX (t ) 4
9 2(1 21 t ) 21 49 (1 21 t )
GX (1) 49 (1 21 )2 4
9 ( 23 )2 1
GX (1) 49 (1 21 ) 4
9 23 2
3
1
E( X ) G(1)
Var( X ) G(1) G(1)
[G(1)]
2
2
3 11
2
3
[6]
(iv) G3 X 2(t ) t 27
8
(1 21 t )
2 3 3
8
27 t 2(1 21 t 3 )3
[2]
e ( 1 1)
G(1)
2 of 4 17/01/19 © MEI
integralmaths.org
Edexcel FS1 PGFs Assessment solutions
E( X ) G(1)
Var( X ) G(1) G(1)
[G(1)]
2
2 2
[5]
( 1 t )
(iii) G X (t ) e
GY (t ) e ( 1 t )
G X Y (t ) G X (t ) GY (t )
e ( 1 t )e ( 1 t )
e ( )( 1 t )
This is the p.g.f. for a Poisson distribution with parameter
so, as p.g.f.s are unique, X + Y ~ Poisson( ).
[6]
P( X x and X Y n )
(iv) P( X x | X Y n )
P( X Y n )
P( X x and Y n x )
P( X Y n )
e x e n x
x ! (n x )!
( )
e ( )n
n!
n! x n x
x !(n x )! ( )n
n x n x
x ( ) ( )
x n x
x n x
n
1
x
This probability distribution is binomial with parameters n and .
[6]
3 of 4 17/01/19 © MEI
integralmaths.org
Edexcel FS1 PGFs Assessment solutions
pt
G(t )
1 t (1 p )
p[ 1 t (1 p )] pt (1 p ) p
G( t )
[ 1 t (1 p )]2
[ 1 t (1 p )]2
p p 1
G(1) 2
[ 1 (1 p )] 2
p p
1
E( X ) G(1)
p
Var( X ) G (1) G(1)
[G(1)]
2
2(1 p ) 1 1
2
p p p2
2 2p p 1
p2
1p
p2
[9]
(ii) X1 is the number of trials to the success 1st
X2 is the number of trials from the 1st success to the 2nd success
and so on, until Xn is the number of trials from the (n-1)th success to the nth
success.
Therefore Y = X1 + X2 + ... + Xn is the total number of trials until the
nth success.
The Xi are independent, because the original sequence of trials is.
So GY (t ) G X (t ) p nt n [ 1 t (1 p )]n
n
n
E(Y ) E( X 1 ) E( X 2 ) ... E( Xn )
p
n(1 p )
Var(Y ) Var( X 1 ) Var( X 2 ) ... Var( X n )
p2
[5]
4 of 4 17/01/19 © MEI
integralmaths.org