Inverse Gamma Distribution

< Back to Probability Distribution List

The inverse gamma distribution (or inverted gamma distribution) is commonly used for Bayesian analysis. It has the same distribution of the reciprocal of the gamma distribution. The shorthand X~inverted gamma(α, β), or IG(α, β), means that a random variable X has an inverse gamma distribution with positive parameters α and β.

There are more than two different parameterizations of the inverse gamma distribution. One common one for the probability density function (PDF) is:

Graph of the inverse gamma distribution PDF for varying values of α and β.

The PDF has two positive parameters (α and β):

  1. The shape parameter α > 0 controls the height. In general, the greater the alpha, the taller the probability density function (PDF); higher values for this parameter will also result in thinner tails.
  2. The scale parameter β > 0 controls the spread.
  3. Γ (α) is the gamma function.

The expected value of the inverse gamma distribution is [1]

The variance is:

You may see different formulas depending on which parameterization the author chooses for the PDF. For example, some Bayesians parametrize the distribution as a scaled inverse chi-squared distribution.

The inverse chi-squared is a special case of the inverse gamma distribution with α = ν /2 and β = ½, where ν = degrees of freedom [2] and a special case of type 5 Pearson distribution.  


The main use of the inverse gamma distribution is in Bayesian probability as a marginal posterior (a way to summarize uncertain quantities) or as a conjugate prior (a prior is a probability distribution that represents your beliefs about a quantity, without taking any evidence into account). The distribution is also used in machine learning, reliability theory (a general theory about systems failure), and survival analysis [3].


Graph of inverse gamma distribution: IkamusumeFan, <> CC BY-SA 4.0</a>, via Wikimedia Commons

[1] Inverted Gamma Distribution.

[2] Gelman, A. et al. (2003). Bayesian Data Analysis, Second Edition. Taylor & Francis.

[3] Stephanie Glen. “Inverse Gamma Distribution: Definition, Mean, Variance, PDF” From Elementary Statistics for the rest of us!