EDFtest (version 0.1.0)

AD: Anderson-Darling statistic

Description

Compute the Anderson-Darling goodness-of-fit statistic \(A^2\) for an i.i.d sample, x, to test for the given distribution with parameters unknown.Estimate parameters by ML using EDFtest MLE function by default.

Usage

AD.uniform(x, parameter = estimate.uniform(x))
AD.normal(x, parameter = estimate.normal(x))
AD.gamma(x, parameter = estimate.gamma(x))
AD.logistic(x, parameter = estimate.logistic(x))
AD.laplace(x, parameter = estimate.laplace(x))
AD.weibull(x, parameter = estimate.weibull(x))
AD.extremevalue(x)
AD.exp(x, parameter = estimate.exp(x))
AD(z)

Arguments

x

A random sample.

parameter

Parameters of the given distribution, MLE by default.

z

A standard uniform random sample.

Value

Anderson-Darling statistic of the given sample.

See Also

estimate for estimating distribution parameters by ML; CvM for calculating Cram<U+00E9>r-von Mises statistic; Watson for calculating Watson statistic; AD.pvalue for calculating P-value of Anderson-Darling statistic.

Examples

# NOT RUN {
x0=runif(n=100,min=-1,max=1)
AD.uniform(x0)

x1=rnorm(n=100,mean=0,sd=1)
AD.normal(x1)

x2=rgamma(n=100,shape=1,scale=1)
AD.gamma(x2)

x3=rlogis(n=100,location=0,scale=1)
AD.logistic(x3)

x4= rmutil::rlaplace(n=100,m=0,s=1)
AD.laplace(x4)

x5=rweibull(n=100,shape=1,scale=1)
AD.weibull(x5)
x5_log=log(x5)
AD.extremevalue(x5_log)

x6=rexp(n=100,rate=1/2)
AD.exp(x6)
# }