Gage study model

`daspi.anova.model.GageStudyModel(source, target, reference, tolerance, resolution, u_cal, u_bi=None, u_lin=None, u_rest=None, k=2, tolerance_ratio=0.2, q_ms_limit=0.15, cg_limit=1.33, cgk_limit=1.33, resolution_ratio_limit=0.05, alpha=0.05, bias_corrected=False)` ¶

Bases: LinearModel

Calculates uncertainties for a measurement system (MSA Type 1 study), supporting one or multiple GageEstimator instances. If multiple are provided, the uncertainty for linearity is also calculated.

PARAMETER	DESCRIPTION
`source`	Pandas DataFrame as tabular data in a long format used for the model. TYPE: `pandas DataFrame`
`target`	Column name for source data holding the measurement values. TYPE: `str`
`reference`	Column name holding the reference values for the measured parts. This column is also used to identify which measured values belong to which reference part. If the column has missing values, the pandas method `ffill()` is used to ensure that the column is filled. The uncertainty u_lin is determined if the column contains several reference values, unless a known measurement uncertainty is given with the argument `u_lin`. TYPE: `str`
`tolerance`	The specification limits for the measurement system. This can be a float, a `SpecLimits` instance or a `Specification` instance. If a float is given, it is interpreted as the tolerance (e.g., 0.1 for ±0.05). TYPE: `float \| SpecLimits \| Specification`
`resolution`	The resolution of the measurement system. If None, the resolution is estimated from the data. If a float is given, it is interpreted as the resolution (e.g., 0.01 for a resolution of 0.01). TYPE: `float \| None`
`u_cal`	The measurement uncertainty of the gage used to measure the reference value. This parameter quantifies the uncertainty associated with the measurement device itself, reflecting how much the measured value could vary due to the inherent limitations of the gage. If a float is specified, it is assumed to be the expanded uncertainty with a coverage factor of `k = 2`, which typically corresponds to a 95% confidence level. Please note that the coverage factor of the calibration is independent of the coverage factor selected here. TYPE: `MeasurementUncertainty \| float`
`u_bi`	The uncertainty for bias `u_BI` can be specified here if it is known; otherwise, it will be determined from the data. This parameter represents the systematic error that may affect the measurement results, indicating how much the measured values deviate from the true value due to consistent inaccuracies. The default is None. TYPE: `MeasurementUncertainty \| None` DEFAULT: `None`
`u_lin`	The measurement uncertainty for linearity `u_LIN` is determined if multiple reference values are provided. This parameter assesses how well the measured values conform to a linear relationship with the reference values. However, it will not be determined if a known measurement uncertainty is provided here. The default is None. TYPE: `MeasurementUncertainty \| None` DEFAULT: `None`
`u_rest`	The uncertainty for further uncertainties not covered by the parameters above. This parameter accounts for all additional sources of uncertainty that may affect the overall measurement but are not specifically addressed by the default uncertainties provided here. It provides a more comprehensive assessment of measurement uncertainty by considering factors that may arise from environmental conditions, operator influences, or other unknown variables. If known, it can be specified here; otherwise, the default value is None. TYPE: `MeasurementUncertainty \| None` DEFAULT: `None`
`k`	The coverage factor for expanded uncertainty. It is used as a multiplier to determine the expanded uncertainty based on the standard uncertainty. The value of `k` is typically set to reflect the desired confidence level in the measurement results. Default is 2, typical values are: - k=2 corresponds to a confidence interval of 95.45% - k=3 corresponds to a confidence interval of 99.73% TYPE: `int \| float` DEFAULT: `2`
`tolerance_ratio`	The ratio of the tolerance to the standard deviation of the measurement system. If the ratio is below this limit, the measurement system is considered unacceptable. Default is 0.2. TYPE: `float` DEFAULT: `0.2`
`q_ms_limit`	The limit for the Q_MS value. If the Q_MS is below this limit, the measurement system is considered acceptable. Default is 0.15. TYPE: `float` DEFAULT: `0.15`
`cg_limit`	The limit for the Gage R&R study's Cg value. If the Cg value below this limit, the measurement system is considered unacceptable. Default is 1.33. TYPE: `float` DEFAULT: `1.33`
`cgk_limit`	The limit for the Gage R&R study's Cgk value. If the Cgk value is below this limit, the measurement system is considered unacceptable. Default is 1.33. TYPE: `float` DEFAULT: `1.33`
`resolution_ratio_limit`	The ratio of the resolution to the standard deviation of the measurement system. If the ratio is below this limit, the measurement system is considered unacceptable. Default is 0.05. TYPE: `float` DEFAULT: `0.05`
`bias_corrected`	Indicates whether the bias is corrected for the Gage R&R study. If True, the bias is not included in the measurement uncertainty; otherwise, it is included. Default is False. TYPE: `bool` DEFAULT: `False`

Examples:

Run the following command in a jupyter notebook to get the html output of gage or you can also use print(repr(gage)) instead:

import daspi as dsp
df = dsp.load_dataset('grnr_layer_thickness')
gage = dsp.GageStudyModel(
    source=df,
    target='result_gage',
    reference='reference',
    u_cal=df['U_cal'][0],
    tolerance=df['tolerance'][0],
    resolution=df['resolution'][0],)
chart = dsp.GageStudyCharts(
        gage, stretch_figsize=1.5
    ).plot(
    ).stripes(
    ).label(
    ) # .save('path/to/file.png')
gage # or print(repr(gage))

Notes

If only one reference is provided, the uncertainty for linearity (LIN) will be 0.0. If multiple references are provided, the uncertainty for linearity will be calculated based on the standard deviation of the biases of the GageEstimator instances.

`gage` `property` ¶

Returns the GageEstimator instances (read-only).

`ref_gages` `property` ¶

Returns a list of GageEstimator instances used in the model (read_only).

If only one reference is used, it returns a list containing one GageEstimator of that reference. If multiple references are used, it returns a list of GageEstimator instances for each reference part measured.

`bias_corrected` `property` ¶

Whether the bias is corrected for the Gage R&R study. If True, the bias itself is not included in the measurement uncertainty; otherwise, it is.

`bias` `property` ¶

Returns the bias of the Gage study (read-only).

`n_samples` `property` ¶

The number of samples used in the Gage study (read-only).

`n_references` `property` ¶

Get the amount of references used in the study (read-only).

`n_replications` `property` ¶

Get the number of replications for each reference (read-only).

`tolerance` `property` ¶

Returns the tolerance of the first GageEstimator instance (read-only).

`resolution` `property` ¶

Returns the resolution of the first GageEstimator instance (read-only).

`k` `property` `writable` ¶

Get the coverage factor k used in uncertainty calculations.

This property returns the coverage factor, which is a multiplier used to determine the expanded uncertainty based on the standard uncertainty. The value of k is typically set to reflect the desired confidence level in the measurement results.

Set the coverage factor with a positive number, typical values are: - k=2 corresponds to a confidence interval of 95.45% - k=3 corresponds to a confidence interval of 99.73%

`u_cal` `property` ¶

The expanded uncertainty of the calibration.

`u_re` `property` ¶

The uncertainty of the resolution of the testing system (read-only).

`u_bi` `property` ¶

The uncertainty of the bias of the testing system (read-only).

`u_lin` `property` ¶

The uncertainty of linearity of the measurement system (read-only).

`u_evr` `property` ¶

The uncertainty of the expanded variance ratio of the testing system (read-only).

`u_rest` `property` ¶

The uncertainty of the repeatability of the testing system (read-only).

`u_ms` `property` ¶

The uncertainty of the measurement system (read-only).

`q_ms_limit` `property` ¶

Get the provided limit for the Q_MS as float between 0 and 1 (read-only).

`T_min_UMS` `property` ¶

The minimum allowed tolerance for this testing system based on the uncertainty (read-only).

`from_gage_estimators(gages, k=2, bias_corrected=False)` `staticmethod` ¶

Create a GageStudyModel from a list of GageEstimator instances. This method is useful when you already have GageEstimator instances and want to create a GageStudyModel without needing to provide the source DataFrame and other parameters again.

PARAMETER DESCRIPTION

gages

A list of GageEstimator instances to create the model from.

TYPE: GageEstimator | List[GageEstimator]

k

The coverage factor for expanded uncertainty. It is used as a multiplier to determine the expanded uncertainty based on the standard uncertainty. The value of k is typically set to reflect the desired confidence level in the measurement results. Default is 2, typical values are: - k=2 corresponds to a confidence interval of 95.45% - k=3 corresponds to a confidence interval of 99.73%

TYPE: int | float DEFAULT: 2

RETURNS	DESCRIPTION
`GageStudyModel`	A new GageStudyModel instance created from the provided GageEstimator instances.

`references_analysis()` ¶

Returns a DataFrame with the analysis of the reference parts.

The analysis includes the GageEstimator statistics for each reference part, such as Cg, Cgk, resolution ratio, and bias.

RETURNS	DESCRIPTION
`DataFrame`	Analysis of the reference parts with GageEstimator statistics.

Examples:

import daspi as dsp
df = dsp.load_dataset('grnr_layer_thickness')
gage = dsp.GageStudyModel(
    source=df,
    target='result_gage',
    reference='reference',
    u_cal=df['U_cal'][0],
    tolerance=df['tolerance'][0],
    resolution=df['resolution'][0],)
print(gage.references_analysis())

     Ref     mean     Bias         s      R
1  0.101  0.10066 -0.00034  0.000688  0.003

`capabilities()` ¶

Returns a DataFrame with the capabilities of the measurement system.

The capabilities include Cg, Cgk, and the ratio of resolution the .

RETURNS	DESCRIPTION
`DataFrame`	Capabilities for Cp, Cpk, Cg, Cgk and the ratio of resolution to standard deviation.

Examples:

import daspi as dsp
df = dsp.load_dataset('grnr_layer_thickness')
gage = dsp.GageStudyModel(
    source=df,
    target='result_gage',
    reference='reference',
    u_cal=df['U_cal'][0],
    tolerance=df['tolerance'][0],
    resolution=df['resolution'][0],)
print(gage.capabilities())

            Value  Limit  Capable     T_min
Cg        2.179005   1.33     True  0.018311
Cgk       1.932051   1.33     True  0.021711
RE_ratio  0.033333   0.05     True  0.020000
U_MS      0.096371   0.15     True  0.019274
p_BI      0.001024   0.05    False       NaN

`uncertainties()` ¶

Returns a DataFrame with the uncertainties for the measurement system.

The table contains the following rows:

CAL: Calibration uncertainty
RE: Resolution uncertainty
BI: Bias uncertainty
LIN: Linearity uncertainty
EVR: Equipment Variation on the Reference
REST: Other known and provided uncertainty
MS: Measurement System uncertainty

The table contains the following columns:

u: The measurement uncertainty for the respective components
U: The expanded uncertainty as k * u
Q: The Quality Indicator serves as a quality indicator for the measurement process, reflecting how well the measurement system performs in relation to the specified requirements and tolerances.
rank: The rank of the uncertainty component, where a lower rank indicates a more significant contribution to the overall uncertainty.

RETURNS	DESCRIPTION
`DataFrame`	Uncertainties for CAL, RE, BI, LIN, EVR, REST, MS.

Examples:

import daspi as dsp
df = dsp.load_dataset('grnr_layer_thickness')
gage = dsp.GageStudyModel(
    source=df,
    target='result_gage',
    reference='reference',
    u_cal=df['U_cal'][0],
    tolerance=df['tolerance'][0],
    resolution=df['resolution'][0],)
print(gage.uncertainties())

             u         U         Q  rank
CAL   0.000100  0.000200  0.013333   4.0
RE    0.000289  0.000577  0.038490   2.0
BI    0.000196  0.000393  0.026173   3.0
LIN   0.000000  0.000000  0.000000   NaN
EVR   0.000688  0.001377  0.091785   1.0
REST  0.000000  0.000000  0.000000   NaN
MS    0.000723  0.001446  0.096371   NaN

Gage study model

daspi.anova.model.GageStudyModel(source, target, reference, tolerance, resolution, u_cal, u_bi=None, u_lin=None, u_rest=None, k=2, tolerance_ratio=0.2, q_ms_limit=0.15, cg_limit=1.33, cgk_limit=1.33, resolution_ratio_limit=0.05, alpha=0.05, bias_corrected=False) ¶

gage property ¶

ref_gages property ¶

bias_corrected property ¶

bias property ¶

n_samples property ¶

n_references property ¶

n_replications property ¶

tolerance property ¶

resolution property ¶

k property writable ¶

u_cal property ¶

u_re property ¶

u_bi property ¶

u_lin property ¶

u_evr property ¶

u_rest property ¶

u_ms property ¶

q_ms_limit property ¶

T_min_UMS property ¶

from_gage_estimators(gages, k=2, bias_corrected=False) staticmethod ¶

references_analysis() ¶

capabilities() ¶

uncertainties() ¶

`daspi.anova.model.GageStudyModel(source, target, reference, tolerance, resolution, u_cal, u_bi=None, u_lin=None, u_rest=None, k=2, tolerance_ratio=0.2, q_ms_limit=0.15, cg_limit=1.33, cgk_limit=1.33, resolution_ratio_limit=0.05, alpha=0.05, bias_corrected=False)` ¶

`gage` `property` ¶

`ref_gages` `property` ¶

`bias_corrected` `property` ¶

`bias` `property` ¶

`n_samples` `property` ¶

`n_references` `property` ¶

`n_replications` `property` ¶

`tolerance` `property` ¶

`resolution` `property` ¶

`k` `property` `writable` ¶

`u_cal` `property` ¶

`u_re` `property` ¶

`u_bi` `property` ¶

`u_lin` `property` ¶

`u_evr` `property` ¶

`u_rest` `property` ¶

`u_ms` `property` ¶

`q_ms_limit` `property` ¶

`T_min_UMS` `property` ¶

`from_gage_estimators(gages, k=2, bias_corrected=False)` `staticmethod` ¶

`references_analysis()` ¶

`capabilities()` ¶

`uncertainties()` ¶