Estimating vocal repertoires from finite samples in which we expect undercounting - View it on GitHub