Creating time-indexed datasets with clusters of texts as inputs and timeseries as targets. - View it on GitHub