It is a package that helps to estimate the number of real clusters in data as well as to identify atypical units. The underlying methods are based on distances rather than on unit x variables.
