Smoke render: optimisation and MCMC are shrunk for CI speed, so figures and diagnostics are not publication-fidelity. The production docs are built with full-length inference.

GPJax

FunctionalSample#

gpjax.typing.FunctionalSample#

Type alias for functions representing \(B\) samples from a model, to be evaluated on any set of \(N\) inputs (of dimension \(D\)) and returning the evaluations of each (potentially approximate) sample draw across these inputs.

alias of Callable[[Array, 'N D'] | Float[ndarray, 'N D']], Array, 'N B'] | Float[ndarray, 'N B']]