LaplaceConfig

C++: phynexis::ml::shape::NetLaplace::LaplaceConfig
Python: phynexis.ml.shape.LaplaceConfig
Header: src/ml/shape/net_laplace.hpp

Description

Configuration struct for Laplace field training and surface extraction. Inherits all fields from NetField.SurfaceConfig.

Creates a configuration with default values.

Property	Type	Default	Description
`extraction_method`	`ExtractionMethod`	`SphericalProjection`	Surface extraction method
`projection_tolerance`	`float`	`1e-5`	Surface projection tolerance
`projection_max_iterations`	`int`	`100`	Maximum projection iterations
`alpha`	`float`	`0.0`	Alpha shape parameter
`use_alpha_shape`	`bool`	`False`	Use alpha shape
`grid_resolution`	`int`	`64`	Grid resolution for marching cubes
`iso_value`	`float`	`0.0`	Isosurface value

Property	Type	Default	Description
`expansion_factor`	`float`	`1.5`	AABB expansion factor for bounds
`num_domain_points`	`int`	`5000`	Number of domain points for PDE training
`num_interior_points`	`int`	`100`	Number of interior points (deprecated)
`num_boundary_points`	`int`	`200`	Number of boundary points for BC training
`interior_weight_scale`	`float`	`50.0`	Weight scale for interior point loss
`use_siren`	`bool`	`True`	Use SIREN activation
`omega_0`	`float`	`30.0`	SIREN frequency parameter
`epochs`	`int`	`2000`	Training epochs
`lr`	`float`	`0.001`	Learning rate
`lambda_pde`	`float`	`1.0`	PDE loss weight
`lambda_bc`	`float`	`10.0`	Boundary condition loss weight