NPMLEmix: Two-Groups Mixture Model with Covariates

We develop three procedures for estimation in a two-groups model with covariates. One of them is a nonparametric maximum likelihood estimation based approach when the signal distribution is an infinite Gaussian location mixture and the signal proportion is a logistic function of the available covariates. Two other functions - marg1() and marg2() have also been implemented for inference in the above framework. All these methods can be used for inference in multiple hypotheses testing. For more information, see the paper: Deb, Saha, Guntuboyina and Sen (2019), ''Two-component Mixture Model in the Presence of Covariates'' <arXiv:1810.07897v2>.

install.packages("NPMLEmix")

Try the NPMLEmix package in your browser