Unintentional warning after each first-time step of any simulation?

[navidcy]

This warning appears after the initial time-step of any simulation...

https://github.com/CliMA/OceanTurbulenceParameterEstimation.jl/blob/30d5cfbedad1cab328885bce485d19cb5b6afe03/src/Observations.jl#L310-L316

[ Info: Initializing simulation...
[ Info:     ... simulation initialization complete (830.369 ms)
[ Info: Executing initial time step...
[ Info:     ... initial time step complete (8.342 ms).
┌ Warning: Current time 20 seconds not found in
│                      time collector times ["0 seconds", "1 hour", "2 hours", "3 hours", "4 hours", "5 hours", "6 hours", "7 hours", "8 hours", "9 hours", "10 hours"]

I don't think that was the intention, right?

cc @glwagner

[glwagner]

There's something amiss. We shouldn't "call" the FieldTimeSeriesCollector at times that we don't intend to collect data. I think its a problem with initializing SpecifiedTimes.

[glwagner]

basically every callback is called at model.clock.iteration=0. This means we get the first time. I don't know why the callback is called again at model.clock.iteration=1.

[glwagner]

We don't have to throw a warning but we probably should have a way of validating that the data collected by FieldTimeSeriesCollector is right. Maybe we can keep track of the "collection times" independently from the data itself. Then after a run is done, we can verify that collector.collection_times == collector.times.

[navidcy]

#220 resolved this, right?

[glwagner]

I guess #220 means that we can now take out the warning without worrying anything bad might happen. That change would resolve this issue.