The Johns Hopkins University Center for Systems Science and Engineering (CSSE) maintains the COVID-19 Data Repository, which is updated daily (JHU). Every day at 6 AM UTC, the data is updated, which is immediately after the raw JHU data.

It is available in raw form like files with the prefix RAW and in a convenient form the files prefixed with CONVENIENT).

The data consists of:

• Country-level confirmed cases and fatalities
• By US county, confirmed cases and fatalities
• Some metadata that is present in the unprocessed JHU data.

The RAW version is provided exactly as it was in the initial dataset.

Analyzing the CONVENIENT version should be simpler. Instead of by row, the data is grouped by column. The metadata is extracted and placed in a different file. Additionally, it switched from cumulative totals to daily change.

You can discuss any problems you notice with the data in this thread of discussion. I'll try to address the topics that received the most votes.

Please post your requests for modifying or enhancing this data in this discussion thread. I'll try once more to respond to the requests with the most votes.

I have a notebook that refreshes immediately upon each data dump update, providing a quick summary of the most recent data. If you want to see how to read the CONVENIENT data into a pandas data frame, it's also a helpful resource.