Puneet Varma (Editor)

Inverse gamma distribution

Updated on
Edit
Like
Comment
Share on FacebookTweet on TwitterShare on LinkedInShare on Reddit
Inverse-gamma distribution

Parameters
  
α > 0 {displaystyle alpha >0} shape (real) β > 0 {displaystyle eta >0} rate (real)

Support
  
x ∈ ( 0 , ∞ ) {displaystyle xin (0,infty )!}

PDF
  
β α Γ ( α ) x − α − 1 exp ⁡ ( − β x ) {displaystyle { rac {eta ^{alpha }}{Gamma (alpha )}}x^{-alpha -1}exp left(-{ rac {eta }{x}}ight)}

CDF
  
Γ ( α , β / x ) Γ ( α ) {displaystyle { rac {Gamma (alpha ,eta /x)}{Gamma (alpha )}}!}

Mean
  
β α − 1 {displaystyle { rac {eta }{alpha -1}}!} for α > 1 {displaystyle alpha >1}

Mode
  
β α + 1 {displaystyle { rac {eta }{alpha +1}}!}

In probability theory and statistics, the inverse gamma distribution is a two-parameter family of continuous probability distributions on the positive real line, which is the distribution of the reciprocal of a variable distributed according to the gamma distribution. Perhaps the chief use of the inverse gamma distribution is in Bayesian statistics, where the distribution arises as the marginal posterior distribution for the unknown variance of a normal distribution, if an uninformative prior is used, and as an analytically tractable conjugate prior, if an informative prior is required.

Contents

However, it is common among Bayesians to consider an alternative parametrization of the normal distribution in terms of the precision, defined as the reciprocal of the variance, which allows the gamma distribution to be used directly as a conjugate prior. Other Bayesians prefer to parametrize the inverse gamma distribution differently, as a scaled inverse chi-squared distribution.

Probability density function

The inverse gamma distribution's probability density function is defined over the support x > 0

f ( x ; α , β ) = β α Γ ( α ) x − α − 1 exp ⁡ ( − β x )

with shape parameter α and rate parameter β . Here Γ ( ⋅ ) denotes the gamma function.

Unlike the Gamma distribution, which contains a somewhat similar exponential term, β is a scale parameter as the distribution function satisfies:

f ( x ; α , β ) = f ( x / β ; α , 1 ) β

Cumulative distribution function

The cumulative distribution function is the regularized gamma function

F ( x ; α , β ) = Γ ( α , β x ) Γ ( α ) = Q ( α , β x )

where the numerator is the upper incomplete gamma function and the denominator is the gamma function. Many math packages allow you to compute Q, the regularized gamma function, directly.

Characteristic function

K α ( ⋅ ) in the expression of the characteristic function is the modified Bessel function of the 2nd kind.

Properties

For α > 0 and β > 0 ,

E [ ln ⁡ ( X ) ] = ln ⁡ ( β ) − ψ ( α )

and

E [ X − 1 ] = α β ,

where ψ ( α ) is the digamma function.

The Kullback-Leibler divergence of Inverse-Gamma(αp, βp) from Inverse-Gamma(αq, βq) is the same as the KL-divergence of Gamma(αp, βp) from Gamma(αq, βq):

D K L ( α p , β p ; α q , β q ) = E [ log ⁡ ρ ( X ) π ( X ) ] = E [ log ⁡ ρ ( 1 / Y ) π ( 1 / Y ) ] = E [ log ⁡ ρ G ( Y ) π G ( Y ) ] ,

where ρ , π are the pdfs of the Inverse-Gamma distributions and ρ G , π G are the pdfs of the Gamma distributions, Y is Gamma(αp, βp) distributed.

Differential equation:

{ x 2 f ′ ( x ) + f ( x ) ( − β + α x + x ) = 0 , f ( 1 ) = e − β β α Γ ( α ) }

  • If X ∼ Inv-Gamma ( α , β ) then k X ∼ Inv-Gamma ( α , k β )
  • If X ∼ Inv-Gamma ( α , 1 2 ) then X ∼ Inv- χ 2 ( 2 α ) (inverse-chi-squared distribution)
  • If X ∼ Inv-Gamma ( α 2 , 1 2 ) then X ∼ Scaled Inv- χ 2 ( α , 1 α ) (scaled-inverse-chi-squared distribution)
  • If X ∼ Inv-Gamma ( 1 2 , c 2 ) then X ∼ Levy ( 0 , c ) (Lévy distribution)
  • If X ∼ Gamma ( k , θ ) (Gamma distribution) then 1 X ∼ Inv-Gamma ( k , θ ) (see derivation in the next paragraph for details)
  • Inverse gamma distribution is a special case of type 5 Pearson distribution
  • A multivariate generalization of the inverse-gamma distribution is the inverse-Wishart distribution.
  • For the distribution of a sum of independent inverted Gamma variables see Witkovsky (2001)

    • Derivation from Gamma distribution

    The pdf of the gamma distribution with shape parameter α and rate parameter β is

    f ( x ) = β α Γ ( α ) x α − 1 e − β x

    and define the transformation Y = g ( X ) = 1 X then the resulting transformation is

    f Y ( y ) = f X ( g − 1 ( y ) ) | d d y g − 1 ( y ) | = β α Γ ( α ) ( 1 y ) α − 1 exp ⁡ ( − β y ) 1 y 2 = 1 θ k Γ ( k ) ( 1 y ) k + 1 exp ⁡ ( − 1 θ y ) = 1 θ k Γ ( k ) y − k − 1 exp ⁡ ( − 1 θ y ) = β α Γ ( α ) x − α − 1 exp ⁡ ( − β x ) α = k , β = θ − 1 , x = y .

    References

    Inverse-gamma distribution Wikipedia
    (Text) CC BY-SA


    Similar Topics