thin_film.optimization.needle

Needle Synthesis for thin film stack design.

Iterative layer insertion algorithm based on: Tikhonravov & Trubetskov, “Development of the needle optimization technique and new features of OptiLayer design software”, SPIE Vol. 2253, 1994.

This module uses a numerical approach: trial needles are inserted at sampled positions and the full merit function is evaluated, rather than computing analytical variational derivatives.

Classes

NeedleResult(iteration, layer_index, ...)

Result of a single needle insertion.

NeedleSynthesis(stack, candidate_materials)

Needle synthesis algorithm for discovering optimal thin film stack configurations by iteratively inserting thin layers at optimal positions.

NeedleSynthesisResult(success, ...)

Overall result of needle synthesis.
class NeedleResult(iteration: int, layer_index: int, position_fraction: float, material_name: str, thickness_nm: float, merit_before: float, merit_after: float, improvement: float)[source]

Result of a single needle insertion.

improvement: float
iteration: int
layer_index: int
material_name: str
merit_after: float
merit_before: float
position_fraction: float
thickness_nm: float
class NeedleSynthesis(stack: ThinFilmStack, candidate_materials: list[BaseMaterial], needle_thickness_nm: float = 1.0, min_thickness_nm: float = 1.0, max_iterations: int = 50, target_merit: float | None = None, num_positions_per_layer: int = 10, optimizer_method: OptimizerMethod = 'L-BFGS-B', optimizer_max_iter: int = 200)[source]

Needle synthesis algorithm for discovering optimal thin film stack configurations by iteratively inserting thin layers at optimal positions.

Parameters:

  • stack – Starting thin film stack design.

  • candidate_materials – Materials to try inserting as needles.

  • needle_thickness_nm – Trial needle thickness in nm used for position screening. Smaller values (0.5–2 nm) give a better finite-difference approximation of the variational gradient.

  • min_thickness_nm – Layers thinner than this are removed during cleanup.

  • max_iterations – Maximum number of needle insertion iterations.

  • target_merit – Stop early if merit falls below this value.

  • num_positions_per_layer – Number of internal sampling points per layer.

  • optimizer_method – Scipy method for thickness re-optimization.

  • optimizer_max_iter – Max iterations for thickness re-optimization.

add_spectral_target(property: Literal['R', 'T', 'A'], wavelengths_nm: list[float], target_type: Literal['equal', 'below', 'over'], value: float, weight: float = 1.0, aoi_deg: float = 0.0, polarization: str = 'u', tolerance: float = 1e-06) → NeedleSynthesis[source]

Add a spectral target over multiple wavelengths.

Parameters:

  • property – Optical property (‘R’, ‘T’, or ‘A’).

  • wavelengths_nm – List of wavelengths in nanometers.

  • target_type – Type of target (‘equal’, ‘below’, ‘over’).

  • value – Target value (single scalar applied at all wavelengths).

  • weight – Weight for this target.

  • aoi_deg – Angle of incidence in degrees.

  • polarization – Polarization state (‘s’, ‘p’, ‘u’).

  • tolerance – Tolerance for ‘equal’ targets.

Returns:

self for method chaining.

add_target(property: Literal['R', 'T', 'A'], wavelength_nm: float | list[float], target_type: Literal['equal', 'below', 'over'], value: float | list[float], weight: float = 1.0, aoi_deg: float | list[float] = 0.0, polarization: str = 'u', tolerance: float = 1e-06) → NeedleSynthesis[source]

Add an optimization target.

Parameters:

  • property – Optical property to target (‘R’, ‘T’, or ‘A’).

  • wavelength_nm – Wavelength(s) in nanometers.

  • target_type – Type of target (‘equal’, ‘below’, ‘over’).

  • value – Target value(s).

  • weight – Weight for this target.

  • aoi_deg – Angle(s) of incidence in degrees.

  • polarization – Polarization state (‘s’, ‘p’, ‘u’).

  • tolerance – Tolerance for ‘equal’ targets.

Returns:

self for method chaining.

run(verbose: bool = False) → NeedleSynthesisResult[source]

Execute the needle synthesis algorithm.

Parameters:

verbose – If True, print progress information.

Returns:

NeedleSynthesisResult with the optimized stack and history.

class NeedleSynthesisResult(success: bool, num_iterations: int, num_layers_added: int, initial_merit: float, final_merit: float, history: list[NeedleResult], stack: ThinFilmStack)[source]

Overall result of needle synthesis.

final_merit: float
history: list[NeedleResult]
initial_merit: float
num_iterations: int
num_layers_added: int
stack: ThinFilmStack
success: bool