Testing data changes - Githubissues

Mayor's office spreadsheet no longer updating county sites, only city sites.

City's Google spreadsheet
LA County RShiny - persons tested by date Use persons tested or tests conducted for apples-to-apples comparison?

Can take LA County's RShiny data and use that against city sites.

County tests performed - includes city, county sites and private healthcare providers (Kaiser, etc)
City tests performed - city sites only, no private healthcare providers

Sanity checks:

city tests < county tests (this doesn't seem to be true for 7/17 and 7/18...are numbers just not updated yet?)
city site tests < city + county site tests

Thoughts:

Benchmark of 45 tests per 100k residents is likely targeting # tests conducted, rather than # persons tested. Ideally, it takes only 1 test per person's infection to detect, but known problems of the optimal time to test while asymptomatic/symptomatic and false negative rates from oral swabs makes this problematic.
This benchmark represents a goal toward reopening, which means essential workers are likely tested repeatedly as a way to monitor when new outbreaks flare-up.
We know mayor's spreadsheet, capturing city site tests, will not capture persons tested, but only tests conducted.
Persons tested is interesting metric to know, and is available using LA County's data.
LA County's data will not allow us to break out city numbers. But, the benchmarks are already set at county level, and city is just half that value. Getting rid of city numbers doesn't seem problematic. More crucial to have apples-to-apples comparison, rather than covering county and city metrics even knowing it's apples-to-oranges.
We need to better understand the time lag in testing data being reported. Should we be looking 3 days ago? 5 days ago? How long does it take for testing date associated with given date to hit its maximum?

CityOfLosAngeles / covid19-indicators