Click or drag to resize
Accord.NET (logo)

BetaDistributionGradient Method (Double, Double, Double, Double, Double, Double)

Computes the Gradient of the Log-Likelihood function for estimating Beta distributions.

Namespace:  Accord.Statistics.Distributions.Univariate
Assembly:  Accord.Statistics (in Accord.Statistics.dll) Version: 3.8.0
Syntax
public static double[] Gradient(
	double sum1,
	double sum2,
	double n,
	double alpha,
	double beta,
	double[] g
)
Request Example View Source

Parameters

sum1
Type: SystemDouble
The sum of log(y), where y refers to all observed values.
sum2
Type: SystemDouble
The sum of log(1 - y), where y refers to all observed values.
n
Type: SystemDouble
The total number of observed values.
alpha
Type: SystemDouble
The current alpha value.
beta
Type: SystemDouble
The current beta value.
g
Type: SystemDouble
A bi-dimensional vector to store the gradient.

Return Value

Type: Double
A bi-dimensional vector containing the gradient w.r.t to alpha in its first position, and the gradient w.r.t to be in its second position.
See Also