Algorithm Validation

Algorithm Validation

Cross-validation with reference software, public data, and analytical solutions

This section records algorithm comparisons performed with fixed inputs to assess agreement between the software and independent references. Each record provides the shared inputs, configuration evidence, result comparison, and acceptance criteria.

ValidationIndependent referenceComparisonMain result
Broadband AR Cross-ValidationEssential Macleod400–700 nm forward reflectance and four-layer thickness optimizationOverlapping spectra; maximum optimized-thickness difference of 0.782%
DBR Cross-ValidationEssential Macleod400–800 nm reflectance, transmittance, reflection phase, and a 510 nm angle sweepOverlapping R/T; converted phase agreement; matching angular curves for three polarizations

Agreement applies only to the stated model scope and inputs; it does not replace measurement, fabrication-tolerance analysis, or material-uncertainty analysis. For the assumptions that scope shares with every transfer-matrix solver, and the situations in which real samples break them, see Limits of the TMM Model.

Copyright © 2026 Dreapex