ExtDist (version 0.3.3)

bestDist: Best distribution for (weighted) sample.

Description

A function to choose the best fitted distibution based on specified criteria.

Usage

bestDist(X, w = rep(1, length(X))/length(X), candDist = c("Beta_ab",
  "Laplace", "Normal"), criterion = "AICc")

Arguments

X

obersevations.

w

weights of sample.

candDist

a vector of names of candidate distributions.

criterion

the criterion based on which the best fitted distribution is choosen.

Value

• the name of best fitted distribution and the parameter estimates.

Details

Details

Examples

X <- rBeta_ab(30, a = 0, b = 1, shape1 = 2, shape2 = 10 )
bestDist(X, candDist = c("Beta_ab","Laplace","Normal"), criterion = "logLik")
bestDist(X, candDist = c("Beta_ab","Laplace","Normal"), criterion = "AIC")
bestDist(X, candDist = c("Beta_ab","Laplace","Normal"), criterion = "AICc")
bestDist(X, candDist = c("Beta_ab","Laplace","Normal"), criterion = "BIC")
bestDist(X, candDist = c("Beta_ab","Laplace","Normal"), criterion = "MDL")

w <- c(0.32, 1.77, 1.22, 0.64, 0.38, 0.93, 1.63, 1.34, 0.57, 1.73, 1.67, 0.67,
0.09, 1, 1.55, 0.53, 0.76, 1.06, 1.13, 1.31, 1.18, 1.64, 0.07, 1.41, 1.18,
0.69, 0.28, 1.27, 0.9, 1.08)
bestDist(X, w, candDist = c("Beta_ab","Laplace","Normal"), criterion = "logLik")
bestDist(X, w, candDist = c("Beta_ab","Laplace","Normal"), criterion = "AIC")
bestDist(X, w, candDist = c("Beta_ab","Laplace","Normal"), criterion = "AICc")
bestDist(X, w, candDist = c("Beta_ab","Laplace","Normal"), criterion = "BIC")
bestDist(X, w, candDist = c("Beta_ab","Laplace","Normal"), criterion = "MDL")

# parameter for best distribution
best_dist <- bestDist(X, candDist = c("Beta_ab","Laplace","Normal"), criterion = "logLik")
attributes(best_dist)$best.dist.par