Interface: SklearnStep

Defined in: ml/model.ts:418

A scikit-learn estimator as a tree of steps. Leaves are estimators; steps makes a composite — a Pipeline (mode: "sequential") or a FeatureUnion / ColumnTransformer (mode: "parallel").

Properties

columns?

optional columns?: string[];

Defined in: ml/model.ts:432

Columns a ColumnTransformer branch operates on (shown in the label).

---

mode?

optional mode?: "sequential" | "parallel";

Defined in: ml/model.ts:426

How the children combine. Default "sequential".

---

name

name: string;

Defined in: ml/model.ts:420

Step name ("scaler", "clf", …).

---

params?

optional params?: number;

Defined in: ml/model.ts:430

Fitted parameter count, if known.

---

shape?

optional shape?: number[];

Defined in: ml/model.ts:428

Output feature count / shape, if known.

---

steps?

optional steps?: SklearnStep[];

Defined in: ml/model.ts:424

Child steps of a composite estimator.

---

type

type: string;

Defined in: ml/model.ts:422

Estimator class: "StandardScaler", "RandomForestClassifier", …