An implementation of 14 parsimonious mixture models for model-based clustering or model-based classification. Gaussian, Student's t, generalized hyperbolic, variance-gamma or skew-t mixtures are available. All approaches work with missing data. Celeux and Govaert (1995)