Right now, the user has to download all of the data and unzip it ahead of time in a particular directly. This should be handled JIT. As the user requests a particular data split, e.g., California 2018 1-Year, we should download (if needed) just that particular part of the dataset and cache it (in an optionally user-specified location).
Right now, the user has to download all of the data and unzip it ahead of time in a particular directly. This should be handled JIT. As the user requests a particular data split, e.g., California 2018 1-Year, we should download (if needed) just that particular part of the dataset and cache it (in an optionally user-specified location).