R Packages for density based empirical likelihood based hypothesis tests
Shepherd, Lori Ann
MetadataShow full item record
This thesis presents the functionality of two new R packages for performing goodness-of-fit tests based on density based empirical likelihood methods: dbEmpLikeGOF and dbEmpLikeNorm. The dbEmpLikeGOF function dbEmpLikeGOF tests a given vector of data for normality, uniformity, or if another vector of data is given, distribution equality. Expanding on this concept, dbEmpLikeNorm function dbELnorm performs a joint assessment for normality of a set of k independent samples with different means and variances. Each test reports a test statistic and associated p -value obtained from a Monte-Carlo approach. This thesis presents relevant concepts to density based empirical likelihood ratio test for goodness-of-fit, discusses the derivations of test statistics, and presents the packages and code with applications to real data.