boax.core.models.kernels.transformations.scaled

Contents

boax.core.models.kernels.transformations.scaled#

boax.core.models.kernels.transformations.scaled(inner, amplitude)#

Scales a given kernel.

Computes k(x, y) = amplitude * inner(x, y).

Example

>>> inner = rbf(jnp.array([0.2, 3.0]))
>>> kernel = scale(inner, 3.0)
>>> Kxx = kernel(xs, xs)
Parameters:

  • kernel – The kernel to be scaled.

  • amplitude (Union[Array, ndarray, bool, number, float, int]) – The parameter controlling the maximum of the kernel.

Return type:

Kernel

Returns:

A scaled Kernel.

Contents