Class EstimateByLogLikelihood
- java.lang.Object
-
- dev.nm.stat.evt.evd.univariate.fitting.EstimateByLogLikelihood
-
public class EstimateByLogLikelihood extends Object
Result from maximum likelihood fitting algorithm, which contains:- the log-likelihood function,
- the fitted parameters for the target model,
- the variance-covariance matrix,
- the standard errors,
- the confidence intervals.
-
-
Constructor Summary
Constructors Constructor Description EstimateByLogLikelihood(Vector fittedParameters, RealScalarFunction logLikelihoodFunction)
-
Method Summary
All Methods Instance Methods Concrete Methods Modifier and Type Method Description ConfidenceInterval
confidenceInterval(double confidenceLevel)
Compute the \((1 - \alpha)100\%\) confidence intervals for each element of the fitted parameter, given the required confidence level.Matrix
covarianceMatrix()
Get the covariance matrix, which is estimated as the inverse of negative Hessian matrix of the log-likelihood function valued at the fitted parameter.ImmutableVector
getFittedParameters()
Get the fitted parameters.RealScalarFunction
getLogLikelihoodFunction()
Get the log-likelihood function.double
logLikelihood()
Compute the log-likelihood at the fitted value.Vector
standardError()
Get the standard errors of the fitted parameters.
-
-
-
Constructor Detail
-
EstimateByLogLikelihood
public EstimateByLogLikelihood(Vector fittedParameters, RealScalarFunction logLikelihoodFunction)
-
-
Method Detail
-
getFittedParameters
public ImmutableVector getFittedParameters()
Get the fitted parameters. That is, the parameters that evaluate to the maximum log-likelihood.- Returns:
- the fitted parameters for the model
-
getLogLikelihoodFunction
public RealScalarFunction getLogLikelihoodFunction()
Get the log-likelihood function. That is, \[ \ell(\theta | X_1,\ldots,X_n) = \sum_{i=1}^n \log f(X_i| \theta) \] where \(\theta\) is the parameter, \(X_i\) are the observations, \(f(.)\) is the probability density function.- Returns:
- the log-likelihood function
-
logLikelihood
public double logLikelihood()
Compute the log-likelihood at the fitted value. That is, the maximum log-likelihood.- Returns:
- the maximum log-likelihood computed at the fitted value
-
covarianceMatrix
public Matrix covarianceMatrix()
Get the covariance matrix, which is estimated as the inverse of negative Hessian matrix of the log-likelihood function valued at the fitted parameter.- Returns:
- the covariance matrix
-
standardError
public Vector standardError()
Get the standard errors of the fitted parameters.- Returns:
- the standard errors
-
confidenceInterval
public ConfidenceInterval confidenceInterval(double confidenceLevel)
Compute the \((1 - \alpha)100\%\) confidence intervals for each element of the fitted parameter, given the required confidence level. That is, \[ CI = (\hat{\theta} \pm z_{\alpha/2} \hat{\sigma}_{\hat{\theta}}) \] where \(\hat{\theta}\) is the fitted parameter, \(\hat{\sigma}_{\hat{\theta}}\) is the standard error of the estimate.- Parameters:
confidenceLevel
- the required confidence level- Returns:
- the confidence interval
-
-