AlarmManagementEvents¤

Alarm frequency, alarm-on durations, chattering and standing alarms (ISA-18.2 style) from one boolean alarm signal (value_bool, True = active).

ts_shape.events.production.alarm_management · Production guide · API reference

Example¤

import numpy as np
import pandas as pd
from ts_shape.events.production.alarm_management import AlarmManagementEvents

rng = np.random.default_rng(42)
df = pd.DataFrame({
    "systime": pd.date_range("2024-01-01 06:00", periods=500, freq="30s"),
    "uuid": "line1:temp_high_alarm",
    "value_bool": rng.random(500) < 0.15,
})
alarms = AlarmManagementEvents(df, alarm_uuid="line1:temp_high_alarm")
print(alarms.alarm_frequency(window="1h").head()[["start", "end", "alarm_count"]])
chattering = alarms.chattering_detection(min_transitions=5, window="10min")
standing = alarms.standing_alarms(min_duration="1h")

Methods¤

Method	Returns
`alarm_frequency(window='1h')`	activations per window: `start`, `end`, `duration_seconds`, `alarm_count`
`alarm_duration_statistics()`	one row: `alarm_count`, `min_`/`avg_`/`max_`/`total_duration_seconds` of the alarm-on intervals
`chattering_detection(min_transitions=5, window='10min')`	windows with at least `min_transitions` state changes: `start`, `end`, `transition_count`
`standing_alarms(min_duration='1h')`	alarm-on intervals longer than `min_duration`: `start`, `end`, `duration_seconds`

Notes¤

An activation is a False to True transition. With include_initial=True (default) an alarm already active in the first sample counts as an activation; pass include_initial=False to ignore it.
An alarm-on interval ends at the next False sample, or at the last sample if the alarm is still active at the end of the data.
Missing values count as False.