Isotropic covariance functions
==============================
------------------------------------------
``gpyreg.isotropic_covariance_functions``
------------------------------------------

The isotropic kernels use a single length scale for every input dimension, so they have two hyperparameters whatever the dimensionality. Each is implemented as a subclass of :ref:`\`\`AbstractIsotropicKernel\`\``, which fixes those two hyperparameters and the bounds recommended for them (``hyperparameter_count``, ``hyperparameter_info`` and ``get_bounds_info``), and of the kernel of the same family in :ref:`\`\`gpyreg.covariance_functions\`\``, from which it inherits the rest, such as the degree of the Matern kernel; each defines its own ``compute``. An isotropic kernel is not periodic: given the ``periods`` that the kernels of :ref:`\`\`gpyreg.covariance_functions\`\`` take, its constructor raises ``ValueError``. An instance is passed to ``gpyreg.GP`` at initialization in the same way as an anisotropic one.

``AbstractIsotropicKernel``
---------------------------
.. autoclass:: gpyreg.isotropic_covariance_functions.AbstractIsotropicKernel
    :members:
    :undoc-members:
    :show-inheritance:

``MaternIsotropic``
-------------------
.. autoclass:: gpyreg.isotropic_covariance_functions.MaternIsotropic(degree: int)
    :members:
    :undoc-members:
    :show-inheritance:

``SquaredExponentialIsotropic``
-------------------------------
.. autoclass:: gpyreg.isotropic_covariance_functions.SquaredExponentialIsotropic()
    :members:
    :undoc-members:
    :show-inheritance: