Number of readings per minute

[carlosparadis]

It would be interesting to revisit this metric but rather than using it from a "missing data" perspective, compare it to the sample rate we left devices on. Ideally, the plot should have some line indicating the correct amount, for example, every hour, of readings, and how many it is getting. Or, the plot can convey the difference between the expected number of readings and the actual ones being obtained.

The plot mentioned here was the one created on #1

[kathrynparadis]

See issue #6. This should be what this issue is asking for.