To keep things more organized in terms of tasks, I'm creating this issue to track the remaining tasks under the despiking test umbrella.
We got a lot of discussion in #543. However, the work is beyond just biological/physical variables, and eve burst/nonburst data. For example, we need to care for time sampling quality, time shifts, unequal bursts, etc. There are also the issue of using results of previously run QC tests (e.g. in/out of the water), and despiking beyond just a single burst (e.g. windowed burst despiking).
Feel free to start/continue the discussion again here, but note that I may be editing the tasks below.
[x] Code to analyse the TIME sampling quality and detect several cases of bogus time sampling handling in any timeseries data
[x] Review some work from the burstQCstats branch.
[x] Take into consideration other QC flags and/or avoid rewriting qc flags from other tests (e.g. in/out of water).
[x] Enable despiking tests to work on individual bursts & ensemble of bursts.
[x] Remove the fence in ImosTimeSeriesSpikeQC for burst data/variables.
To keep things more organized in terms of tasks, I'm creating this issue to track the remaining tasks under the despiking test umbrella.
We got a lot of discussion in #543. However, the work is beyond just biological/physical variables, and eve burst/nonburst data. For example, we need to care for time sampling quality, time shifts, unequal bursts, etc. There are also the issue of using results of previously run QC tests (e.g. in/out of the water), and despiking beyond just a single burst (e.g. windowed burst despiking).
Feel free to start/continue the discussion again here, but note that I may be editing the tasks below.