You are on page 1of 2

Factorial of Negative One Half

Ben Hall
December 28, 2015

The Gamma Functions Relation to the Factorial

Recall the Gamma Function:

xt1 ex dx

(t)

(1)

Integrating by parts:
(t) = x

Z

e + (t 1)

t1 x

xt2 ex dx

(2)

Evaluating the first term:


xt1
(t) = lim x lim xt1 ex + (t 1)
x0
x
e

xt2 ex dx

(3)

The middle term clearly goes to zero. The first term can be seen to go to zero by applying
LHospitals Rule t 1 times:
dt1 xt1
t1
lim ddx
t1 ex
x
dxt1

= lim
x

(t 1)(t 2)...1
=0
ex

(4)

Hence only the third term of (3) remains, allowing us to see the following recursive relation:
Z
(t) = (t 1)
xt2 ex dx = (t 1)(t 1)
(5)
0

Carrying out the recursive relation:


(t) = (t 1)(t 1) = (t 1)(t 2)(t 2) = ...
= (t 1)(t 2)...(1) = (t 1)(t 2)...1 = (t 1)!

(6)

We discover that the Gamma Function maps its argument to the factorial of its argument minus
one:
(n) = (n 1)!
(7)
Hence we have discovered that the Gamma Function generalizes the factorial function. We can
now investigate the factorial of all real numbers.

The Specific Case for the Factorial of Negative One


Half

2.1

Connection to the Gaussian Integral

We now use the Gamma Functions relation to the factorial (7) to see that the factorial of
negative one half is given by:
 


  Z
1
1
1
1
x 2 ex dx
(8)
!=
1 !=
=
2
2
2
0
We now use the following substitution:
1

u x2
1 1
du = x 2 dx
2

(9)

In terms of x:
x = u2

(10)

dx = 2x 2 du = 2udu
Substituting in our change of variables (10) to our integral (8):
Z
Z
2
21 x
x e dx = 2
eu du
0

(11)

Since eu is an even function, we obtain:


 
Z
1
2
!=
eu du
2

2.2

(12)

Evaluating the Gaussian Integral

We can evaluate the Gaussian integral by evaluating its square:


2 Z
  2 Z
Z
Z Z
1
2
2
u2
x2
y 2
e du =
! =
e dx
e dy =
e(x +y ) dxdy
2


Converting to polar coordinates:
ZZ
e

(x2 +y 2 )

dxdy =

all space

rer drd

(13)

(14)

Evaluating the integral:


  2
Z
1
2
rer dr
! = 2
2
 0 
1 r2
= 2
e
2
0


0
= e
e =
Hence:

1
!=
2
2

(15)

(16)

You might also like