`accumulated_local_effect` – Apley & Zhu (2020) accumulated local effects – PDP alternative robust to correlation.#

Back to op axis | Back to L7 | Browse all options

Operational op under axis op, sub-layer L7_A_importance_dag_body, layer l7. Standalone callable: mf.functions.ale_importance.

Function signature#

mf.functions.ale_importance(
    result: FitResultBase,
    X: np.ndarray | pd.DataFrame,
) -> ALEImportanceResult

Parameters#

name	type	default	constraint	description
`result`	`FitResultBase`	—	—	Fitted result object exposing ._model (the raw sklearn estimator). Returned by any L4 standalone callable such as mf.functions.ridge_fit, mf.functions.random_forest_fit, etc.
`X`	`np.ndarray	pd.DataFrame`	—	—

Returns#

ALEImportanceResult — frozen dataclass with fit results.

Attribute	Type	Description
`.importances_`	`np.ndarray`	Mean absolute centred ALE (L1 norm) per feature.
`.feature_names_`	`list[str]`	Feature names.
`.ale_values_`	`dict[str, np.ndarray]`	Centred cumulative ALE values per feature.
`.summary(top_n=10)`	`str`	Human-readable text table.

Behavior#

For feature j, computes the cumulative local change Σ_{k≤K} E_{X_{-j} | x_j ∈ bin_k}[∂f/∂x_j]·Δx_j. The binning + conditioning eliminates the ‘extrapolation into low-density regions’ bias of plain PDPs.

When to use

Correlated feature panels (FRED-MD / -QD) where PDPs are misleading.

In recipe context#

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

# Layer L7 recipe fragment
params:
  op: accumulated_local_effect

References#

macroforecast design Part 3, L7: ‘every importance op produces (table, figure) pairs; the L7.B sub-layer governs export shape.’
Apley & Zhu (2020) ‘Visualizing the Effects of Predictor Variables in Black Box Supervised Learning Models’, JRSS Series B 82(4): 1059-1086.

accumulated_local_effect – Apley & Zhu (2020) accumulated local effects – PDP alternative robust to correlation.#