`xgboost` – XGBoost gradient-boosted trees (optional dependency).#

Back to family axis | Back to L4 | Browse all options

Operational op under axis family, sub-layer L4_A_model_selection, layer l4. Standalone callable: mf.functions.xgboost_fit.

Function signature#

mf.functions.xgboost_fit(
    X: np.ndarray | pd.DataFrame,
    y: np.ndarray | pd.Series,
) -> XGBoostFitResult

Parameters#

name	type	default	constraint	description
`X`	`np.ndarray	pd.DataFrame`	—	—
`y`	`np.ndarray	pd.Series`	—	—

Returns#

XGBoostFitResult — frozen dataclass with fit results.

Attribute	Type	Description
`.feature_importances_`	`np.ndarray`	Feature importances (gain-based) from XGBoost, shape (n_features,).
`.n_estimators_used`	`int`	Number of boosting rounds (= n_estimators parameter).
`.predict(X)`	`np.ndarray`	Predictions for new data X, shape (n_samples,).
`.summary()`	`str`	Human-readable table of fit results including top-3 feature importances.

Behavior#

Requires pip install macroforecast[xgboost]. Histogram-based tree construction; native quantile loss; GPU support. Standard production-grade boosting library.

When to use

Production sweeps where xgboost’s speed matters; quantile forecasting (xgb 2.0+).

When NOT to use

Lightweight installs (no extra installed) – raises ImportError.

In recipe context#

Set params.family = "xgboost" in the relevant layer to activate this op within a recipe:

# Layer L4 recipe fragment
params:
  family: xgboost

References#

macroforecast design Part 2, L4: ‘forecasting model is the layer where every authoring iteration ends – pick family, tune, repeat.’
Chen & Guestrin (2016) ‘XGBoost: A Scalable Tree Boosting System’, KDD.

xgboost – XGBoost gradient-boosted trees (optional dependency).#