multi_sensor_validation

multi_sensor_validation ¤

MultiSensorValidationEvents ¤

MultiSensorValidationEvents(
    dataframe: DataFrame,
    sensor_uuids: List[str],
    *,
    value_column: str = "value_double",
    event_uuid: str = "quality:multi_sensor",
    time_column: str = "systime"
)

Bases: Base

Quality: Multi-Sensor Cross-Validation

Cross-validate redundant inline sensors measuring the same process variable. Detects disagreement, identifies drifting sensors, and assesses measurement consensus.

Methods: - detect_disagreement: Flag windows where sensor spread exceeds threshold. - pairwise_bias: Mean difference between each sensor pair per window. - consensus_score: Per-window measurement consensus across sensors. - identify_outlier_sensor: Find the sensor furthest from the group.

detect_disagreement ¤

detect_disagreement(
    threshold: float, window: str = "5m"
) -> pd.DataFrame

Flag windows where sensor spread exceeds threshold.

Parameters:

Name	Type	Description	Default
`threshold`	`float`	Maximum acceptable spread (max - min) across sensors.	required
`window`	`str`	Resample window size.	`'5m'`

Returns:

Type	Description
`DataFrame`	DataFrame with columns: window_start, window_end, max_spread,
`DataFrame`	sensor_high, sensor_low, duration.

pairwise_bias ¤

pairwise_bias(window: str = '1h') -> pd.DataFrame

Mean difference between each sensor pair per window.

Parameters:

Name	Type	Description	Default
`window`	`str`	Resample window size.	`'1h'`

Returns:

Type	Description
`DataFrame`	DataFrame with columns: window_start, sensor_a, sensor_b,
`DataFrame`	bias, abs_bias.

consensus_score ¤

consensus_score(window: str = '1h') -> pd.DataFrame

Per-window measurement consensus across sensors.

Score = 1.0 means perfect agreement, 0.0 means high disagreement.

Parameters:

Name	Type	Description	Default
`window`	`str`	Resample window size.	`'1h'`

Returns:

Type	Description
`DataFrame`	DataFrame with columns: window_start, consensus_mean,
`DataFrame`	spread_std, consensus_score.

identify_outlier_sensor ¤

identify_outlier_sensor(
    window: str = "1h", method: str = "median"
) -> pd.DataFrame

Identify the sensor furthest from the group per window.

Parameters:

Name	Type	Description	Default
`window`	`str`	Resample window size.	`'1h'`
`method`	`str`	'median' or 'mean' — central tendency for comparison.	`'median'`

Returns:

Type	Description
`DataFrame`	DataFrame with columns: window_start, outlier_sensor,
`DataFrame`	deviation, other_sensors_mean.