Coronavirus data II
One other issue in looking at the spread of coronavirus is the problem of insufficient testing. If testing is done only on needs basis (because test kits are scarce) then mostly people with severe and covid19-like symptoms are being tested, while people like me aren't tested and probably will never be (although I'm coughing, coming out of what feels like a mild flu).

In this testing scenario, if the epidemic is spreading we expect an increase of the number of tests performed, together with an increase of detected cases. We expect that a constant fraction of persons being tested are positive. If the epidemic has stopped spreading (a large number of infected people is now infecting a small number of still-healthy people), we expect a diminishing number of tests but still a constant fraction of positive cases.

If tests are so scarce that not everyone with severe symptoms is tested, we still expect to see exactly the same results: an increasing number of tests and a proportionally increasing number of positive cases.

A different scenario is when tests are done routinely: say, everyone coming into a hospital or an airport will be tested for covid19, regardless of symptoms. In that case, the number of tests grows according to the number of test kits available. The fraction of positive cases can grow or decrease, depending on whether the epidemic is spreading faster or slower than the number of test kits available.

Summary: we can't conclude whether the epidemic is spreading fast or is slowing down unless we know, week after week in a given region, 1) how many test kits are being made available to doctors, 2) who is being selected for testing (by severe symptoms or routinely), 3) how many actual tests are being made.