New Data: CA COVID-19 Hospital Data

[elaguerta]

County or Counties: All
Data source: https://data.chhs.ca.gov/dataset/california-covid-19-hospital-data-and-case-statistics/resource/6cd8d424-dfaa-4bdd-9410-a3d656e1176e?inner_span=True
Output to:
Update frequency/time: Daily, likely in the morning, updated to present day

[benghancock]

Hi @elaguerta - I'd love to help with this issue, if you're still looking for help on it. It looks like the CSV data is pretty tidy, and as noted, pulls the hospital data for all counties. The columns are:

['county', 'todays_date', 'hospitalized_covid_confirmed_patients',
       'hospitalized_suspected_covid_patients', 'hospitalized_covid_patients',
       'all_hospital_beds', 'icu_covid_confirmed_patients',
       'icu_suspected_covid_patients', 'icu_available_beds']

How do you envision this integrating with the county scrapers? Would each county scraper also pull from this dataset, and filter for the selected county? Or would it not be integrated -- i.e. the hospital data would be scraped separately?

[elaguerta]

Hi @benghancock, for now I think we should keep it separate because it is coming from a completely different source than all the county sources. Any integration that might be needed by the front-end design, or by our future API users, can happen downstream. To make that easier, I'd suggest you structure your top method to get just one county by a name or other identifier passed as an argument.

[benghancock]

Got it, that makes sense. I'll plan to start working on this later today. I've forked the project and will create a branch for this issue. Thanks!

[benghancock]

I've got the pagination working to loop through and collect all records returned by the API. Still need to collect the notes for each field, and document the output data model. By default the response returns all the records going back to March, which is a lot (5,000+ for all counties) and will obviously continue to grow. Do we want to limit the time period or just pull them all in so we have complete data each time?

[benghancock]

Hi @elaguerta - I've put in a PR that hopefully delivers what's needed for this.