q.label.concurrency counts the label intervals active at each observation. Event lifetimes include both their start and end observations, matching the endpoint convention used by triple-barrier labels.
Concurrent labels share future returns and are therefore not independent observations.
import pandas as pd
import qrt as q
observations = pd.date_range("2026-01-01", periods=7, name="datetime")
end_times = pd.Series(
observations[[3, 5, 6]],
index=observations[[0, 1, 4]].rename("event_time"),
name="touch_time",
)
active = q.label.concurrency(observations, end_times)
pd.DataFrame({"concurrency": active})| concurrency | |
|---|---|
| datetime | |
| 2026-01-01 | 1 |
| 2026-01-02 | 2 |
| 2026-01-03 | 2 |
| 2026-01-04 | 2 |
| 2026-01-05 | 2 |
| 2026-01-06 | 2 |
| 2026-01-07 | 1 |
The first event is active from positions 0 through 3, the second from 1 through 5, and the third from 4 through 6. Counts outside all event lifetimes are zero.
Back to top