LuasKernel.K#
- LuasKernel.K(hp: Any, x_l1: Array, x_l2: Array, x_t1: Array, x_t2: Array, **kwargs) Array [source]#
Generates the full covariance matrix K formed from the sum of two kronecker products:
\[K = K_l \otimes K_t + S_l \otimes S_t\]Not needed for any calculations with the
LuasKernel
but useful for creating aGeneralKernel
object with the same kernel function as aLuasKernel
.- Parameters:
hp (Pytree) – Hyperparameters needed to build the covariance matrices
Kl
,Kt
,Sl
,St
. Will be unaffected if additional mean function parameters are also included.x_l1 (JAXArray) – The first array containing wavelength/vertical dimension regression variable(s) for the observed locations. May be of shape
(N_l,)
or(d_l,N_l)
ford_l
different wavelength/vertical regression variables.x_l2 (JAXArray) – Second array containing wavelength/vertical dimension regression variable(s).
x_t1 (JAXArray) – The first array containing time/horizontal dimension regression variable(s) for the observed locations. May be of shape
(N_t,)
or(d_t,N_t)
ford_t
different time/horizontal regression variables.x_t2 (JAXArray) – Second array containing time/horizontal dimension regression variable(s).
- Returns:
The full covariance matrix K of shape
(N_l*N_t, N_l*N_t)
.- Return type:
JAXArray