metacausal.ComponentCateEstimate

class metacausal.ComponentCateEstimate(cate, ci_lower=None, ci_upper=None, details=None)

Bases: object

CATE result from a single component method.

Variables:

• cate (numpy.ndarray) – Per-observation treatment effect estimates, shape (n,).

• ci_lower (numpy.ndarray | None) – Lower bounds of confidence/credible intervals, shape (n,). From the method’s native inference (e.g., EconML analytical CIs, stochtree posterior quantiles). None if unavailable.

• ci_upper (numpy.ndarray | None) – Upper bounds, shape (n,). None if unavailable.

• details (dict | None) – Method-specific additional information.

Parameters:

• cate (ndarray)

• ci_lower (ndarray | None)

• ci_upper (ndarray | None)

• details (dict | None)

Methods

__init__

Attributes

ci_lower
ci_upper
details
cate

Details

cate: ndarray

ci_lower: ndarray | None = None

ci_upper: ndarray | None = None

details: dict | None = None