gauge_repeatability

gauge_repeatability ¤

GaugeRepeatabilityEvents ¤

GaugeRepeatabilityEvents(
    dataframe: DataFrame,
    signal_uuid: str,
    *,
    part_column: str = "value_string",
    value_column: str = "value_double",
    operator_column: str | None = None,
    event_uuid: str = "quality:gauge_rr",
    time_column: str = "systime"
)

Bases: Base

Quality: Gauge Repeatability & Reproducibility (Gauge R&R)

Measurement System Analysis (MSA) for inline sensors. Evaluates repeatability (within-part, same sensor) and reproducibility (across operators/stations) using repeated measurements of reference parts or check standards.

Methods: - repeatability: Equipment Variation (EV) per part. - reproducibility: Appraiser Variation (AV) across operators. - gauge_rr_summary: Full Gauge R&R table with %GRR and ndc. - measurement_bias: Compare measurements to known reference values.

repeatability ¤

repeatability(n_trials: int | None = None) -> pd.DataFrame

Equipment Variation (EV) — within-part variation.

Groups measurements by part, computes range-based or pooled standard deviation estimate of repeatability.

When operator_column is set, ranges and standard deviations are taken within each operator x part cell and averaged per part (AIAG average-and-range method). Pooling every operator's trials into one range folded operator bias (reproducibility) into EV.

Parameters:

Name	Type	Description	Default
`n_trials`	`int \| None`	Expected number of trials per part. If None, auto-detected from data.	`None`

Returns:

Type	Description
`DataFrame`	DataFrame with columns: part, mean, range, repeatability_std, EV.

reproducibility ¤

reproducibility() -> pd.DataFrame

Appraiser Variation (AV) — between-operator variation.

Groups by operator_column, computes variation of part means across operators. Requires operator_column to be set.

AIAG average-and-range method: AV = sqrt((X_diff * K2)^2 - EV^2 / (n * r)) with X_diff the range of operator means, K2 = 1/d2* for the number of operators, n parts and r trials. The EV^2 / (n * r) term removes the repeatability noise already contained in the operator means; a negative radicand gives AV = 0.

Returns:

Type	Description
`DataFrame`	DataFrame with columns: operator, mean, reproducibility_std, AV.

gauge_rr_summary ¤

gauge_rr_summary(
    tolerance_range: float | None = None,
) -> pd.DataFrame

Full Gauge R&R summary table.

Computes EV, AV, GRR, PV, TV, %GRR, %PV, and number of distinct categories (ndc).

Parameters:

Name	Type	Description	Default
`tolerance_range`	`float \| None`	Total tolerance range (USL - LSL). If provided, %GRR is also computed relative to tolerance.	`None`

Returns:

Type	Description
`DataFrame`	DataFrame with single row: EV, AV, GRR, PV, TV, pct_GRR,
`DataFrame`	pct_PV, ndc, and optionally pct_GRR_tolerance.
`DataFrame`	Canonical twin columns: `grr_pct` (after `pct_GRR`), `pv_pct` (after
`DataFrame`	`pct_PV`), `grr_tolerance_pct` (after `pct_GRR_tolerance`); each holds
`DataFrame`	the same values as the deprecated column right before it, which is removed
`DataFrame`	in the next major release.

measurement_bias ¤

measurement_bias(
    reference_values: dict[str, float],
) -> pd.DataFrame

Compare average measurements to known reference values.

Parameters:

Name	Type	Description	Default
`reference_values`	`dict[str, float]`	Dict mapping part identifier to its true reference value, e.g. {"part_A": 10.0, "part_B": 20.0}.	required

Returns:

Type	Description
`DataFrame`	DataFrame with columns: part, measured_mean, reference,
`DataFrame`	bias, bias_pct.