Skip to contents

Create an Uncertainty Budget

Source: R/uncertainty.R
measure_uncertainty_budget.Rd

Combines multiple uncertainty components into a complete uncertainty budget following ISO GUM methodology. Calculates combined standard uncertainty, effective degrees of freedom (Welch-Satterthwaite), and expanded uncertainty.

Usage

measure_uncertainty_budget(..., .list = NULL, k = 2, result_value = NULL)

Arguments

...

uncertainty_component() objects to include in the budget.

.list

Optional list of uncertainty components.

k

Coverage factor for expanded uncertainty. Default is 2 (approximately 95% coverage for normal distribution).

result_value

Optional. The measurement result value, used for calculating relative uncertainty.

Value

A measure_uncertainty_budget object containing:

  • components: List of input uncertainty components

  • combined_u: Combined standard uncertainty

  • effective_df: Effective degrees of freedom (Welch-Satterthwaite)

  • coverage_factor: The k value used

  • expanded_U: Expanded uncertainty (k * combined_u)

  • result_value: The measurement result (if provided)

  • relative_u: Relative standard uncertainty (if result provided)

Details

Combined Standard Uncertainty

Calculated as the root sum of squares of contributions: $$u_c = \sqrt{\sum_i (c_i \cdot u_i)^2}$$

Welch-Satterthwaite Effective Degrees of Freedom

$$\nu_{eff} = \frac{u_c^4}{\sum_i \frac{(c_i \cdot u_i)^4}{\nu_i}}$$

This is used to determine the appropriate coverage factor for a given confidence level.

Expanded Uncertainty

$$U = k \cdot u_c$$

With k=2, this provides approximately 95% coverage.

See also

uncertainty_component() for creating components, tidy.measure_uncertainty_budget() for extracting results, autoplot.measure_uncertainty_budget() for visualization.

Examples

# Create components
u_repeat <- uncertainty_component("Repeatability", 0.05, type = "A", df = 9)
u_cal <- uncertainty_component("Calibrator", 0.02, type = "B", df = 50)
u_temp <- uncertainty_component("Temperature", 0.03, type = "B")

# Create budget
budget <- measure_uncertainty_budget(u_repeat, u_cal, u_temp, k = 2)
print(budget)
#> <measure_uncertainty_budget>
#>   Components: 3 (1 Type A, 2 Type B)
#>   Combined u: 0.06164
#>   Effective df: 21
#>   Coverage k: 2
#>   Expanded U: 0.1233

# With result value for relative uncertainty
budget <- measure_uncertainty_budget(
  u_repeat, u_cal, u_temp,
  result_value = 10.5
)