# Controlled-grid full-vector linewidth optimization

- Step accepted: **True**
- Initial/trial/retained linewidth Q: **967.292047 / 968.651934 / 968.651934**
- Retained Q factor: **1.00140587**
- Initial/trial fit R²: **0.999913210 / 0.999914042**
- Trial complex-field overlap: **0.99998889**
- Trial inner/gap/PML energy: **0.917009 / 0.691717 / 0.157066**
- Failed gates: **none**
- Withheld initial/trial Q: **939.889575 / 940.838657**
- Peak resident memory: **1.848 GB**
- Runtime: **555.75 s**

The objective is the fitted linewidth of the real-axis, centered-gap Maxwell
response.  No complex-pole extrapolation is used to accept this step.  The
result is controlled-grid evidence only; transverse and longitudinal grid
refinement are still required before treating the Q value as predictive.