create.kernel.object: Create a kernel object
In mattdneal/gaussianProcess: Gaussian Process Fitting and Prediction

Kernel objects provide a unifying wrapper for the various things which can act as kernels (built-in kernels (represented by strings), model trees, and arbitrary R functions).

1 2	create.kernel.object(kernel, grad_function = NULL, hess_function = NULL, hyperparam_names = character(0), additional_params = list())

`grad_function`	if `kernel_function` is a function, then this should be a function returning the grad w.r.t. the hyperparameters, otherwise this should be null
`hyperparam_names`	A character vector of the names of the hyperparameters
`additional_params`	A list of additional parameters
`kernel_function`	the kernel function (either a string, a function, or a ModelTree)