lumopt2.fom#

The figure of merit module defines the objective functions for optimization.

Figure of merit (FOM) module for inverse design optimization.

This module provides classes for defining and computing figures of merit based on simulation results from field and port monitors. The FOM determines what is being optimized in the inverse design problem.

Classes#

Fom: Factory function that returns FieldFom or PortFom based on monitor types.
BaseFom: Base class providing shared FOM infrastructure (advanced use).
FieldFom: FOM calculator for field-based (DFT monitor) simulations.
PortFom: FOM calculator for port-based (waveguide mode) simulations.
FieldResults: Container for field monitor simulation results.
PortResults: Container for port monitor simulation results.

Examples#

Port transmission optimization (using factory):

>>> from lumopt2.fom import Fom, PortResults
>>> port_out = PortResults('port_output')
>>> def fom_function(port):
...     return port.transmission.values
>>> fom = Fom(port_out, fct=fom_function)  # Returns PortFom instance

Field intensity optimization (using factory):

>>> from lumopt2.fom import Fom, FieldResults
>>> import autograd.numpy as anp
>>> field = FieldResults('focus_region')
>>> fom = Fom(field, fct=lambda f: anp.sum(anp.abs(f.values) ** 2))  # Returns FieldFom instance