Mathematical Definitions
Introduction
This page provides the mathematical definitions of the expanded gamma distribution, detailing its probability density function, cumulative distribution function, and key statistical measures. These definitions serve as a foundation for understanding the distribution’s properties and for applying it to real-world data analysis scenarios.
Parameters
The expanded gamma distribution is characterized by three parameters:
\(\alpha\) (alpha): The shape parameter, which must be positive (:math:alpha > 0).
\(\beta\) (beta): The scale parameter, which must be non-zero (:math:beta neq 0). It determines the direction of skewness; positive values lead to right-skewness, while negative values result in left-skewness.
\(\delta\) (delta): The location parameter, which can be any real number and shifts the distribution along the x-axis.
The parameters \(\alpha\) and \(\beta\) primarily control the shape and scale of the distribution, respectively. The parameter \(\delta\) shifts the distribution along the x-axis.
Note that while the distribution is strictly either left- or right-skewed, as the parameter \(\alpha\) approaches infinity, the shape of the distribution converges to that of a normal distribution. Thus, for practical purposes, an expanded gamma distribution defined with a high \(\alpha\) value, such as \(1e9\), can approximate symmetrical scenarios quite well.
Distribution Functions
The probability density function (PDF) and cumulative distribution function (CDF) are defined as follows:
Probability Density Function (PDF):
where \(\Gamma(\alpha)\) is the gamma function \(\displaystyle\int_{0}^{\infty}t^{\alpha-1}e^{-t}\,dt\)
Cumulative Density Function (CDF):
where \(\gamma \left(\alpha, \frac{x-\delta}{\beta}\right)\) is the lower incomplete gamma function \(\displaystyle\int_{0}^{\frac{x-\delta}{\beta}}t^{\alpha-1}e^{-t}\,dt\\.\)
Please note that the cumulative distribution function (CDF) of the gamma distribution does not have a closed-form solution. Consequently, the computation of the inverse CDF, or percent-point function, cannot be performed directly and requires the use of numerical techniques or approximation methods for its calculation.
Statistical measures
Expected value:
Variance:
Skewness:
Kurtosis:
Mode: