gpflow.covariances#

Functions#

gpflow.covariances.Kuf#

This function uses multiple dispatch, which will depend on the type of argument passed in:

gpflow.covariances.Kuf( InducingPoints, Kernel, object )
# dispatch to -> gpflow.covariances.kufs.Kuf_kernel_inducingpoints(...)

gpflow.covariances.kufs.Kuf_kernel_inducingpoints(inducing_variable, kernel, Xnew)[source]#

Parameters:

inducing_variable (InducingPoints) –
kernel (Kernel) –
Xnew (Union[ndarray[Any, Any], Tensor, Variable, Parameter]) –

Return type:

Tensor

gpflow.covariances.Kuf( Multiscale, SquaredExponential, object )
# dispatch to -> gpflow.covariances.kufs.Kuf_sqexp_multiscale(...)

gpflow.covariances.kufs.Kuf_sqexp_multiscale(inducing_variable, kernel, Xnew)[source]#

Parameters:

inducing_variable (Multiscale) –
kernel (SquaredExponential) –
Xnew (Union[ndarray[Any, Any], Tensor, Variable, Parameter]) –

Return type:

Tensor

gpflow.covariances.Kuf( InducingPatches, Convolutional, object )
# dispatch to -> gpflow.covariances.kufs.Kuf_conv_patch(...)

gpflow.covariances.kufs.Kuf_conv_patch(inducing_variable, kernel, Xnew)[source]#

Parameters:

inducing_variable (InducingPatches) –
kernel (Convolutional) –
Xnew (Union[ndarray[Any, Any], Tensor, Variable, Parameter]) –

Return type:

Tensor

gpflow.covariances.Kuf( InducingPoints, MultioutputKernel, object )
# dispatch to -> gpflow.covariances.multioutput.kufs.Kuf_generic(...)

gpflow.covariances.multioutput.kufs.Kuf_generic(inducing_variable, kernel, Xnew)[source]#

Parameters:

inducing_variable (InducingPoints) –
kernel (MultioutputKernel) –
Xnew (Union[ndarray[Any, Any], Tensor, Variable, Parameter]) –

Return type:

Tensor

gpflow.covariances.Kuf( SharedIndependentInducingVariables, SharedIndependent, object )
# dispatch to -> gpflow.covariances.multioutput.kufs.Kuf_shared_shared(...)

gpflow.covariances.multioutput.kufs.Kuf_shared_shared(inducing_variable, kernel, Xnew)[source]#

Parameters:

inducing_variable (SharedIndependentInducingVariables) –
kernel (SharedIndependent) –
Xnew (Tensor) –

Return type:

Tensor

gpflow.covariances.Kuf( SeparateIndependentInducingVariables, SharedIndependent, object )
# dispatch to -> gpflow.covariances.multioutput.kufs.Kuf_separate_shared(...)

gpflow.covariances.multioutput.kufs.Kuf_separate_shared(inducing_variable, kernel, Xnew)[source]#

Parameters:

inducing_variable (SeparateIndependentInducingVariables) –
kernel (SharedIndependent) –
Xnew (Union[ndarray[Any, Any], Tensor, Variable, Parameter]) –

Return type:

Tensor

gpflow.covariances.Kuf( SharedIndependentInducingVariables, SeparateIndependent, object )
# dispatch to -> gpflow.covariances.multioutput.kufs.Kuf_shared_separate(...)

gpflow.covariances.multioutput.kufs.Kuf_shared_separate(inducing_variable, kernel, Xnew)[source]#

Parameters:

inducing_variable (SharedIndependentInducingVariables) –
kernel (SeparateIndependent) –
Xnew (Union[ndarray[Any, Any], Tensor, Variable, Parameter]) –

Return type:

Tensor

gpflow.covariances.Kuf( SeparateIndependentInducingVariables, SeparateIndependent, object )
# dispatch to -> gpflow.covariances.multioutput.kufs.Kuf_separate_separate(...)

gpflow.covariances.multioutput.kufs.Kuf_separate_separate(inducing_variable, kernel, Xnew)[source]#

Parameters:

inducing_variable (SeparateIndependentInducingVariables) –
kernel (SeparateIndependent) –
Xnew (Union[ndarray[Any, Any], Tensor, Variable, Parameter]) –

Return type:

Tensor

gpflow.covariances.Kuf( FallbackSeparateIndependentInducingVariables, LinearCoregionalization, object )
# dispatch to -> gpflow.covariances.multioutput.kufs.Kuf_fallback_separate_linear_coregionalization(...)

gpflow.covariances.multioutput.kufs.Kuf_fallback_separate_linear_coregionalization(inducing_variable, kernel, Xnew)[source]#

Parameters:

inducing_variable (FallbackSeparateIndependentInducingVariables) –
kernel (LinearCoregionalization) –
Xnew (Union[ndarray[Any, Any], Tensor, Variable, Parameter]) –

Return type:

Tensor

gpflow.covariances.Kuf( FallbackSharedIndependentInducingVariables, LinearCoregionalization, object )
# dispatch to -> gpflow.covariances.multioutput.kufs.Kuf_fallback_shared_linear_coregionalization(...)

gpflow.covariances.multioutput.kufs.Kuf_fallback_shared_linear_coregionalization(inducing_variable, kernel, Xnew)[source]#

Parameters:

inducing_variable (FallbackSharedIndependentInducingVariables) –
kernel (LinearCoregionalization) –
Xnew (Union[ndarray[Any, Any], Tensor, Variable, Parameter]) –

Return type:

Tensor

gpflow.covariances.Kuf( SharedIndependentInducingVariables, LinearCoregionalization, object )
# dispatch to -> gpflow.covariances.multioutput.kufs.Kuf_shared_linear_coregionalization(...)

gpflow.covariances.multioutput.kufs.Kuf_shared_linear_coregionalization(inducing_variable, kernel, Xnew)[source]#

Parameters:

inducing_variable (SharedIndependentInducingVariables) –
kernel (LinearCoregionalization) –
Xnew (Union[ndarray[Any, Any], Tensor, Variable, Parameter]) –

Return type:

Tensor

gpflow.covariances.Kuf( SeparateIndependentInducingVariables, LinearCoregionalization, object )
# dispatch to -> gpflow.covariances.multioutput.kufs.Kuf_separate_linear_coregionalization(...)

gpflow.covariances.multioutput.kufs.Kuf_separate_linear_coregionalization(inducing_variable, kernel, Xnew)[source]#

Parameters:

inducing_variable (SeparateIndependentInducingVariables) –
kernel (LinearCoregionalization) –
Xnew (Union[ndarray[Any, Any], Tensor, Variable, Parameter]) –

Return type:

Tensor

gpflow.covariances.Kuu#

This function uses multiple dispatch, which will depend on the type of argument passed in:

gpflow.covariances.Kuu( InducingPoints, Kernel )
# dispatch to -> gpflow.covariances.kuus.Kuu_kernel_inducingpoints(...)

gpflow.covariances.kuus.Kuu_kernel_inducingpoints(inducing_variable, kernel, *, jitter=0.0)[source]#

Parameters:

inducing_variable (InducingPoints) –
kernel (Kernel) –
jitter (float) –

Return type:

Tensor

gpflow.covariances.Kuu( Multiscale, SquaredExponential )
# dispatch to -> gpflow.covariances.kuus.Kuu_sqexp_multiscale(...)

gpflow.covariances.kuus.Kuu_sqexp_multiscale(inducing_variable, kernel, *, jitter=0.0)[source]#

Parameters:

inducing_variable (Multiscale) –
kernel (SquaredExponential) –
jitter (float) –

Return type:

Tensor

gpflow.covariances.Kuu( InducingPatches, Convolutional )
# dispatch to -> gpflow.covariances.kuus.Kuu_conv_patch(...)

gpflow.covariances.kuus.Kuu_conv_patch(inducing_variable, kernel, jitter=0.0)[source]#

Parameters:

inducing_variable (InducingPatches) –
kernel (Convolutional) –
jitter (float) –

Return type:

Tensor

gpflow.covariances.Kuu( InducingPoints, MultioutputKernel )
# dispatch to -> gpflow.covariances.multioutput.kuus.Kuu_generic(...)

gpflow.covariances.multioutput.kuus.Kuu_generic(inducing_variable, kernel, *, jitter=0.0)[source]#

Parameters:

inducing_variable (InducingPoints) –
kernel (MultioutputKernel) –
jitter (float) –

Return type:

Tensor

gpflow.covariances.Kuu( FallbackSharedIndependentInducingVariables, SharedIndependent )
# dispatch to -> gpflow.covariances.multioutput.kuus.Kuu_shared_shared(...)

gpflow.covariances.multioutput.kuus.Kuu_shared_shared(inducing_variable, kernel, *, jitter=0.0)[source]#

Parameters:

inducing_variable (FallbackSharedIndependentInducingVariables) –
kernel (SharedIndependent) –
jitter (float) –

Return type:

Tensor

gpflow.covariances.Kuu( FallbackSharedIndependentInducingVariables, SeparateIndependent )
gpflow.covariances.Kuu( FallbackSharedIndependentInducingVariables, IndependentLatent )
# dispatch to -> gpflow.covariances.multioutput.kuus.Kuu_fallback_shared(...)

gpflow.covariances.multioutput.kuus.Kuu_fallback_shared(inducing_variable, kernel, *, jitter=0.0)[source]#

Parameters:

inducing_variable (FallbackSharedIndependentInducingVariables) –
kernel (Union[SeparateIndependent, IndependentLatent]) –
jitter (float) –

Return type:

Tensor

gpflow.covariances.Kuu( FallbackSeparateIndependentInducingVariables, SharedIndependent )
# dispatch to -> gpflow.covariances.multioutput.kuus.Kuu_fallback_separate_shared(...)

gpflow.covariances.multioutput.kuus.Kuu_fallback_separate_shared(inducing_variable, kernel, *, jitter=0.0)[source]#

Parameters:

inducing_variable (FallbackSeparateIndependentInducingVariables) –
kernel (SharedIndependent) –
jitter (float) –

Return type:

Tensor

gpflow.covariances.Kuu( FallbackSeparateIndependentInducingVariables, SeparateIndependent )
gpflow.covariances.Kuu( FallbackSeparateIndependentInducingVariables, LinearCoregionalization )
# dispatch to -> gpflow.covariances.multioutput.kuus.Kuu_fallbace_separate(...)

gpflow.covariances.multioutput.kuus.Kuu_fallbace_separate(inducing_variable, kernel, *, jitter=0.0)[source]#

Parameters:

inducing_variable (FallbackSeparateIndependentInducingVariables) –
kernel (Union[SeparateIndependent, LinearCoregionalization]) –
jitter (float) –

Return type:

Tensor