author: Piotr Sulewski, Pomeranian University
The goal of the PSGoft package is to put into practice the (a,b)modified Lilliefors goodness-of-fit normality test. This modificationconsists in varying a formula of calculating the empirical distributionfunction. Values of constants a, b in the formula depend on values ofsample skewness and excess kurtosis, which is recommended in order toincrease the power of the LF test. To read more about the package pleasesee (and cite :)) papers:
Sulewski P. (2019) Modified Lilliefors Goodness-of-fit Test forNormality, Communications in Statistics - Simulation and Computation,51(3), 1199-1219
You can install the released version ofPSGoft fromCRAN with:
install.packages("PSGoft")You can install the development version ofPSGoftfromGitHub with:
library("remotes")remotes::install_github("PiotrSule/PSGoft")This package includes two real data sets
The first one,data1, consist of 72 observations forDozer Cycle Times.
The second one,data2, is the height of 99five-year-old British boys in cm
library(PSGoft)length(data1)#> [1] 72head(data2)#> [1] 96.1 97.1 97.1 97.2 99.2 99.4MLF.stat
This function returns the value of the Modified Lillieforsgoodness-of-fit statistic
MLF.stat(data1)#> [1] 0.05488005MLF.stat(rnorm(33,mean =0,sd =2))#> [1] 0.09910243MLF.pvalue
This function returns the p-value for the test
MLF.pvalue(data1)#> [1] 0.81592MLF.pvalue(rnorm(33,mean =0,sd =2))#> [1] 0.66459MLF.stat
This function returns the value of the Modified Lilliefors statisticand the p-value for the test.
MLF.test(data1)#>#> Modified Lilliefors goodness-of-fit normality test#>#> data: data1#> D = 0.05488, p-value = 0.816MLF.test(rnorm(33,mean =0,sd =2))#>#> Modified Lilliefors goodness-of-fit normality test#>#> data: rnorm(33, mean = 0, sd = 2)#> D = 0.083871, p-value = 0.748