Every row in both event lists has an Acc / Rej toggle. Use it when an event is simply wrong and should not be counted at all.
Rejecting
Flip the toggle to Rej and you get a confirmation:
Reject Event? Are you sure you want to reject this event? It will be excluded from all calculations and metrics.
It shows the event’s time range, duration, and confidence so you can double-check you have the right row. Click Reject and it is done — Event rejected successfully.
The row fades to half opacity and can no longer be clicked.
Accepting again
Nothing is deleted. Flip the toggle back and you get:
Accept Event? Are you sure you want to accept this previously rejected event? It will be included in all calculations and metrics.
Refresh afterwards
Rejecting changes the numbers, so a card appears at the bottom right of the page:
↻ Statistics outdated — Refresh to update calculations
Click Refresh and every metric, card, and chart recalculates without the rejected event. You can reject several events first and refresh once at the end.
When to reject
Do reject:
- An “empty bunk” event where the photos show the bunk clearly has feed in it.
- A delivery that is plainly a vehicle passing through, not feeding.
- Events during a period when the camera was fogged, blocked, or knocked, where you cannot trust anything it saw.
Do not reject just because a number looks wrong. If deliveries are being under-counted, rejecting the ones that were found makes it worse. Rejection removes; it never adds.
Reject or re-tag?
They solve different problems:
- Reject — this event did not happen, or should not be counted.
- Re-tag — this event happened, but it is the wrong kind. A push-up recorded as a delivery should be re-tagged, not rejected, otherwise you lose it from the push-up count too.
Re-tagging keeps the information. Prefer it wherever it applies.
If lots of events are wrong
Correcting them one at a time is the slow way round. If deliveries and push-ups are being confused, tag your feed vehicles instead — one minute, and it re-checks the whole batch.