Gaussian-process mechanism (default)

Models each node's conditional mean as an exact Gaussian process of its parents, with a squared-exponential (ARD) kernel - one lengthscale per parent, which captures non-linearity and interactions automatically. Kernel hyperparameters (lengthscales, signal and noise variance) are set by type-II maximum likelihood; there is no MCMC. A node with no parents is modelled by its marginal (a Normal). This is Kagu's default and only mechanism.

Structure discovery uses the exact GP log marginal likelihood, which - in contrast to fast basis/eigenfunction approximations - is well calibrated: for a genuinely unidentifiable (e.g. linear-Gaussian) edge it does not manufacture spurious confidence about direction.

Posterior function samples for effect propagation are drawn by pathwise / decoupled sampling (a random-feature prior plus the exact data update), so each draw is a coherent function evaluable at any point - giving correctly correlated uncertainty for do-calculus contrasts.

Super class

Mechanism -> GPMechanism

Public fields

num_results: Integer - number of posterior draws to keep (default 1000).
n_features: Integer - random Fourier features for pathwise sampling (default 300).
jitter: Numeric - diagonal jitter for numerical stability (default 1e-6).

Methods

`GPMechanism$new()`

Create a new GPMechanism.

Usage

GPMechanism$new(num_results = 1000L, n_features = 300L, jitter = 1e-06)

Arguments

num_results: Integer - number of posterior draws to keep.
n_features: Integer - number of random Fourier features.
jitter: Numeric - diagonal jitter added to the kernel.

`GPMechanism$fit()`

Fit the node (exact GP if it has parents, marginal Normal if not).

Usage

GPMechanism$fit(node, parents, data, ...)

Arguments

node: Character scalar - the node name (response).
parents: Character vector of parent node names.
data: A data.frame with columns for the node and its parents.
...: Backend-specific arguments.

`GPMechanism$predict_mean()`

Conditional mean draws (see Mechanism).

Usage

GPMechanism$predict_mean(node, parents, parent_values, fit)

Arguments

node: Character scalar - the node name.
parents: Character vector of parent node names.
parent_values: Named list of [n_chains, n_draws] matrices, one per parent, giving the parent values for each posterior draw.
fit: A fit object from $fit().

`GPMechanism$log_marglik()`

Exact GP log marginal likelihood (see Mechanism).

Usage

GPMechanism$log_marglik(node, parents, data)

Arguments

node: Character scalar - the node name.
parents: Character vector of parent node names.
data: A data.frame.

`GPMechanism$posterior_shape()`

Posterior shape (single chain).

Usage

GPMechanism$posterior_shape(fit)

Arguments

fit: A fit object from $fit().

`GPMechanism$node_terms()`

Per-node summary rows: each parent's direct local effect (function gradient at the parent means) plus the residual noise sd.

Usage

GPMechanism$node_terms(node, parents, data, fit)

Arguments

node: Character scalar - the node name.
parents: Character vector of parent node names.
data: A data.frame.
fit: A fit object from $fit().

`GPMechanism$clone()`

The objects of this class are cloneable with this method.

Usage

GPMechanism$clone(deep = FALSE)

Arguments

deep: Whether to make a deep clone.

Examples

mech <- GPMechanism$new()