metacausal.CateEstimate

class metacausal.CateEstimate(cate, component_estimates, aggregation, ensemble_weights=None)

Bases: object

Ensemble CATE result. Pure data, no model references.

Variables:

• cate (numpy.ndarray) – Aggregated per-observation treatment effects, shape (n,).

• component_estimates (dict[str, metacausal.estimators.ComponentCateEstimate]) – Per-method CATE results, keyed by method name. Only includes methods with supports_cate=True.

• aggregation (str) – Aggregation strategy class name.

• ensemble_weights (metacausal.aggregation.weights.EnsembleWeights | None) – Weight vector from the aggregation strategy. None for pointwise strategies (Median, Mean).

Parameters:

• cate (ndarray)

• component_estimates (dict[str, ComponentCateEstimate])

• aggregation (str)

• ensemble_weights (EnsembleWeights | None)

Methods

__init__
cate_profile	Ensemble CATE along one covariate (no CI band; use cate_profile() for a bootstrap band).
summary	Return a formatted, multi-line CATE summary.

Attributes

component_cates	CATE arrays from each component method.
ensemble_weights
n_methods	Number of methods that produced CATE estimates.
cate
component_estimates
aggregation

Details

cate_profile(x, *, xlabel, ax=None, show_components=True, ylim=None, ensemble_label='Ensemble CATE')

Ensemble CATE along one covariate (no CI band; use cate_profile() for a bootstrap band).

Thin wrapper around metacausal.plots.cate_profile(); see there for the full parameter reference. Requires the plots extra (pip install 'metacausal[plots]').

Returns:

matplotlib Axes the plot was drawn on.

Parameters:

• x (np.ndarray)

• xlabel (str)

• ax (Axes | None)

• show_components (bool)

• ylim (tuple[float, float] | None)

• ensemble_label (str)

Return type:

Axes

summary(*, digits=None, signed=False)

Return a formatted, multi-line CATE summary.

Parameters:

• digits (int | None)

• signed (bool)

Return type:

str

aggregation: str

cate: ndarray

property component_cates: dict[str, ndarray]

CATE arrays from each component method.

component_estimates: dict[str, ComponentCateEstimate]

ensemble_weights: EnsembleWeights | None = None

property n_methods: int

Number of methods that produced CATE estimates.