In this repository, we provide the code for RadlER, a novel solution to produce clean samples from data containing duplicates according to a target distribution.
We also provide the datasets used in our experimental evaluation, made available in the dedicated "datasets" folder.