VectorExpansion

class pmrf.models.surrogates.expansion.VectorExpansion(frequency: Any, coefficients_real: Any, coefficients_imag: Any, basis: Any, offset: Any, *, name: str | None = None, metadata: Any = None, domain: str = 's', z0: Any = 50.0)

Bases: AbstractSingleDiscreteDomain

(experimental) A model where the output is a linear expansion of vector/matrix basis functions with an optional offset.

The S-parameters are returned as offset + coefficients @ basis, where the coefficients are the model parameters.

Parameters:

• coefficients_real (Param) – The real coefficients parameters

• coefficients_imag (Param) – The imaginary coefficients parameters

• basis (jnp.ndarray) – The fixed basis functions

• offset (jnp.ndarray) – An optional fixed offset

discrete_matrix() → Array

Return the static discrete matrix.

Returns:

The tabulated intrinsic domain matrix.

Return type:

jax.numpy.ndarray

inverse(sample: Array) → Array

basis: Array

The fixed basis functions

coefficients_imag: AbstractVariable | Inexact[jaxlib._jax.Array, '...']

The imaginary coefficients parameters

coefficients_real: AbstractVariable | Inexact[jaxlib._jax.Array, '...']

The real coefficients parameters

offset: Array

An optional fixed offset