In a recent Upstream blog post we explored where data connected to papers funded by several U.S. Federal Agencies are published. Different data sharing practices across these agencies led to very different distributions of datasets across various repositories. We used CHORUS reports that combine linked article and dataset metadata as input for that work.