My raw data is simply time in seconds. Each row is an event that is either real or not. I need to separate the "real" events from the "non-real" events. "Real" events are either: 1. the first event 2. any event that occurs >20 seconds after another "real" event.

I made a column that simply subtracts the time of a suspected event from the value above. This identifies the majority of "real" events, but misses other "real" events.

I can check the "non-real" events by hand, but I have data sets with thousands of suspected events. I was hoping to find some kind of formula to speed this analysis.

Any help is appreciated! I've attached a data set including my raw data (column A), my current analysis (columns B-D), and identified which events were missed by my initial analysis (columns E and F).

Attached Files

Short example dataset for practice.xlsx (12.7 KB, 14 views)