Run Analysis
Run Analysis is the main one-dimensional precision workflow. It reads the controller state, builds the active timing model, computes ideal and instrument-specific theory curves, optionally validates them with Monte Carlo, and refreshes the diagnostics and estimator-tracking widgets with the same sweep context.
Theory path
For specialists
The desktop calls DigitalTwinService.run_precision(). The service patches a temporary PhysicsConfig, builds the x-grid, computes the ideal reference, computes the instrument theory path, and packages the payloads needed by the precision, accuracy, and diagnostics widgets.
The current theory path separates:
- conditional precision on collected photons
- photon survival / reporting basis
- detector-transfer approximation versus explicit event-driven simulation
Monte Carlo validation
When enabled, Monte Carlo validation runs the estimator repeatedly across the same x-grid and compares the empirical estimator behaviour to the theory path. The payload now includes conditional F, survival eta, compatible/incompatible flags, and optional bootstrap confidence intervals.
Photon basis and rescaling
The collected-photon result is canonical. Toggling Compute F-value on changes the reporting basis, not the collected-photon estimator. In many cases this means cached results can be rescaled without recomputing the underlying sweep.
Outputs
| Output | Where it appears | Meaning |
|---|---|---|
| Ideal reference | Precision widget | Reference ceiling for the active sweep/grid interpretation. |
| Theory curve | Precision widget | Expected performance of the configured instrument. |
| Monte Carlo points and CIs | Precision widget | Empirical validation of the theory path. |
| Estimator tracking | MLE Accuracy widget | Estimated-vs-truth view of estimator bias and spread. |
| Reference PDF and gates | Instrument Diagnostics | Timing-domain explanation of the current result. |