t Location-Scale Distribution
Overview
Thetlocation-scale distribution is useful for modeling data distributions with heavier tails (more prone to outliers) than the normal distribution. It approaches the normal distribution asνapproaches infinity, and smaller values ofνyield heavier tails.
Parameters
Thetlocation-scale distribution uses the following parameters.
Parameter | Description | 金宝app |
---|---|---|
μ | Location parameter | –∞ < μ < ∞ |
σ | Scale parameter | σ > 0 |
ν | Shape parameter | ν > 0 |
To estimate distribution parameters, usemle
. Alternatively, fit atLocationScaleDistribution
object to data usingfitdist
or theDistribution Fitter应用程序。
Probability Density Function
The probability density function (pdf) of thetlocation-scale distribution is
where Γ( • ) is the gamma function,µis the location parameter,σis the scale parameter, andνis the shape parameter .
To compute the probability density function, usepdf
and specify'tLocationScale'
. Alternatively, you can create atLocationScaleDistribution
object usingfitdist
ormakedist
, then use thepdf
to work with the object.
Cumulative Distribution Function
To compute the probability density function, usecdf
and specify'tLocationScale'
. Alternatively, you can create atLocationScaleDistribution
object usingfitdist
ormakedist
, then use thecdf
to work with the object.
Descriptive Statistics
The mean of thetlocation-scale distribution is
whereμis the location parameter. The mean is only defined for shape parameter valuesν> 1. For other values ofν, the mean is undefined.
The variance of thetlocation-scale distribution is
whereμis the location parameter andνis the shape parameter. The variance is only defined for values ofν> 2. For other values ofν, the variance is undefined.
To compute the mean and variance, create atLocationScaleDistribution
object usingfitdist
ormakedist
. You can also use theDistribution Fitter应用程序。
Relationship to Other Distributions
Ifxhas atlocation-scale distribution, with parametersµ,σ, andν, then
has a Student'stdistribution withνdegrees of freedom.