“…MNMs are fitted via the gpcm() function of the R‐package mixture (Browne, ElSherbiny, & McNicholas, ), M t Ms are fitted via the teigen() function of the R‐package teigen (Andrews & McNicholas, ), MCNMs are fitted via the CNmixt() function of the R‐package ContaminatedMixt (Punzo, Mazza, & McNicholas, ), MPEMs are fitted via the mpe() function of an R package available at http://onlinelibrary.wiley.com/doi/10.1111/biom.12351/suppinfo Dang et al (Dang, Browne, & McNicholas, ), whereas a specific R code, available as Supplementary Material for Review, has been implemented to fit MLNMs. To allow for a direct comparison of the competing models, all the algorithms are initialized by providing the initial quantities , : nine times using a random initialization and once with a k ‐means initialization (as implemented by the kmeans() function for R).…”