All demonstrationsMEASUREMENT UNCERTAINTY / RUST CPU

How certain is your answer?

Follow uncertain inputs through a formula. Compare first-order propagation with seeded Monte Carlo, see which inputs contribute most, and catch incompatible units before calculating.

propagate_uncertaintyTry the real call

Concept artwork · not calculation evidence

01 / UNDERSTAND THE METHOD

Give uncertainty a dimension.

INPUT PREVIEW · RUN TO CALCULATE

Choose a case. Then bring it to life.

Preset buttons change the inputs. Calculation starts only when you ask.

THE MEASUREMENT BENCHINPUT PREVIEW

PROPAGATE · COMPARE · QUESTION

How much trust fits inside a number?

x1 + x2 + x3 + x4

Preparing the measurement bench…

A physical view of inputs, sensitivities and coverage intervals.

Input preview · equal-height holders are not uncertainties

Drag to inspect the instrument

01 Describe the inputs.02 Run the Rust engine.03 Compare the intervals.

Loading a preset changes the request. Use the calculation button to obtain a result. This visual does not run or simulate the model.

The model’s inputs

4 declared variables

x10 1

normal · standard uncertainty: 1 1

x20 1

normal · standard uncertainty: 1 1

x30 1

normal · standard uncertainty: 1 1

x40 1

normal · standard uncertainty: 1 1

The bench illustrates a supplied measurement model. Rust calculates the result; the scene displays it. No sample cloud or probability density is invented from the returned intervals.

02 / ASK THE ACTUAL TOOL

Make it your experiment.

Measurement formulaCoverage probability (0.5–0.999)Sample countRandom seed

Complete tool arguments

Arrays, coordinates and nested contracts remain editable here. Units and limits are checked by the real tool.

Runs only when you press the button. Input edits discard the previous displayed result. Public compute limits apply.

KNOW WHAT THE RESULT MEANS

Useful evidence needs boundaries.

Reported uncertainty is conditional on the supplied formula, distributions and correlations. Unmodelled biases are not covered.
A first-order interval may be inadequate for a nonlinear model. More Monte Carlo draws do not repair an invalid physical model.
A convergence diagnostic concerns the sampling calculation; it is not a guarantee about a real measurement.
Signed correlation contributions are shown separately. Independent variance shares must not be read as the complete correlated budget.
The illustration is conceptual. Only a completed call supplies numerical intervals and contribution heights.

Choose another demonstration Call the same tool from your AI