useR! 2024

Because publications by public data providers focus on a broader audience, their datasets are often not convenient to use for research.
To mitigate this problem, the opentimeseries R package provides the time series and official statistics communities with reusable code to conveniently source data from public sources. By splitting data and metadata into two different files, a long format CSV file for the data and a JSON file for multi-lingual metainformation, the package generates output that is inclusive to humans (and their favorite spreadsheet software) _and_ convenient to ingest for machines.
This data output is the starting point not only for intertemporal comparisons but also for versioning of time series, as it is needed for real-time analysis or evaluation of forecasts. The package open-sources a data ingestion framework, proven through its longtime usage in monitoring the Swiss economy at the KOF Swiss Economic Institute at ETH Zurich, for the first time. We explicitly chose the R ecosystem with its great documentation and boiler plating tools to encourage dataset maintenance and community contributions across different fields that use public data for research.