Record forecast component effects in Nightscout LoopStatus

[ps2]

To help with remote interpretation of Loop decision making, we need to be able to see magnitude of momentum and retrospective correction effects in the Nightscout LoopStatus object.

COB and IOB are already present and give an idea of those effects, but we could also express those in terms of BG effect (mg/dL), like RC and momentum.

The final stat I'd want to include is some measure of observed insulin counteraction effects. Something like last 5 minutes, last 15 and last 30 minutes (maybe even 1,6,12,24 hours?).

[dm61]

I'd suggest also including the modeled insulin activity expressed in (mg/dL)/1min or (mg/dL)/5min. It's probably best to use the same units for both insulin counteraction effects and for insulin activity, and report both to NS over last 5 min whenever Loop has a chance to upload status to NS. That would allow experimenting with various retrospective data analyses over longer periods of time.

[ps2]

@dm61 I understand the use case(s) for that data, but device status is about explaining a particular loop run. For historical analysis, I think you'd want to depend as little as possible on derived data, and instead recompute it in the analysis where you can modify parameters involved in deriving those terms like ISF, insulin curve, etc, and incorporate any updates to the source data involved.

[github-actions[bot]]

This issue is stale because it has been open for 30 days with no activity.

[github-actions[bot]]

This issue was closed because it has been inactive for 14 days since being marked as stale.