AutoCrossKernel
- class pmrf.covariance_kernels.AutoCrossKernel(auto: AbstractCovarianceKernel, cross: AbstractCovarianceKernel, num_outputs: int)
Bases:
AbstractCovarianceKernelKernel that routes between a auto-correlation and cross-correlation kernels.
Constructs a block matrix where diagonal elements evaluate the auto kernel and off-diagonal elements evaluate the cross kernel.
This can be used to model reflection (auto) and transmission (cross) discrepancy separately. In this case, set num_outputs to the number of ports.
- Parameters:
auto (AbstractCovarianceKernel) – The covariance kernel describing the auto terms (e.g. S11, S22).
cross (AbstractCovarianceKernel) – The covariance kernel describing the cross terms (e.g. S21, S43).
- __call__(x1, x2, key=None)
Evaluate the kernel between two points.
- Parameters:
x1 (jnp.ndarray) – First input point.
x2 (jnp.ndarray) – Second input point.
key (jax.random.PRNGKey, optional) – Random key for stochastic kernels.
- Returns:
Kernel covariance scalar.
- Return type:
jnp.ndarray
- cross: AbstractCovarianceKernel
- num_outputs: int