@seanhogg/builderforce-memory-engine 2026.9.33 → 2026.10.1

package/dist/ml/classical/linear.d.ts

@@ -0,0 +1,29 @@
+/**
+ * ml/classical/linear.ts — ridge linear regression, closed form.
+ *
+ * Solves (XᵀWX + λI)w = XᵀWy with an unpenalised intercept by Gaussian
+ * elimination with partial pivoting. Small dims only (the catalog's numeric
+ * tasks), which is exactly where a closed form beats iterating.
+ */
+import type { MlPrediction, MlPredictor, MlSample } from "../types.js";
+export interface LinearOptions {
+    /** L2 penalty on weights (not the intercept). Default 1e-6. */
+    l2?: number;
+}
+export declare class LinearModel implements MlPredictor {
+    readonly weights: Float32Array;
+    intercept: number;
+    /** Coefficient of determination on the fitted data, clamped to [0, 1]. */
+    readonly r2: number;
+    constructor(weights: Float32Array, intercept: number, 
+    /** Coefficient of determination on the fitted data, clamped to [0, 1]. */
+    r2: number);
+    value(x: ArrayLike<number>): number;
+    predict(x: Float32Array): MlPrediction;
+    exportParams(): Float32Array;
+    importParams(params: ArrayLike<number>): void;
+}
+/** Solve A·x = b in place (A is n×n row-major). Throws when singular. */
+export declare function solveLinearSystem(A: Float64Array, b: Float64Array, n: number): Float64Array;
+export declare function fitLinear(samples: readonly MlSample[], opts?: LinearOptions): LinearModel;
+//# sourceMappingURL=linear.d.ts.map

package/dist/ml/classical/linear.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"linear.d.ts","sourceRoot":"","sources":["../../../src/ml/classical/linear.ts"],"names":[],"mappings":"AAAA;;;;;;GAMG;AACH,OAAO,KAAK,EAAE,YAAY,EAAE,WAAW,EAAE,QAAQ,EAAE,MAAM,aAAa,CAAC;AAGvE,MAAM,WAAW,aAAa;IAC5B,+DAA+D;IAC/D,EAAE,CAAC,EAAE,MAAM,CAAC;CACb;AAED,qBAAa,WAAY,YAAW,WAAW;IAE3C,QAAQ,CAAC,OAAO,EAAE,YAAY;IACvB,SAAS,EAAE,MAAM;IACxB,0EAA0E;IAC1E,QAAQ,CAAC,EAAE,EAAE,MAAM;gBAHV,OAAO,EAAE,YAAY,EACvB,SAAS,EAAE,MAAM;IACxB,0EAA0E;IACjE,EAAE,EAAE,MAAM;IAGrB,KAAK,CAAC,CAAC,EAAE,SAAS,CAAC,MAAM,CAAC,GAAG,MAAM;IAMnC,OAAO,CAAC,CAAC,EAAE,YAAY,GAAG,YAAY;IAItC,YAAY,IAAI,YAAY;IAO5B,YAAY,CAAC,MAAM,EAAE,SAAS,CAAC,MAAM,CAAC,GAAG,IAAI;CAO9C;AAED,yEAAyE;AACzE,wBAAgB,iBAAiB,CAAC,CAAC,EAAE,YAAY,EAAE,CAAC,EAAE,YAAY,EAAE,CAAC,EAAE,MAAM,GAAG,YAAY,CA6B3F;AAED,wBAAgB,SAAS,CAAC,OAAO,EAAE,SAAS,QAAQ,EAAE,EAAE,IAAI,GAAE,aAAkB,GAAG,WAAW,CAqC7F"}

package/dist/ml/classical/linear.js

@@ -0,0 +1,114 @@
+import { featureDim } from "./math.js";
+export class LinearModel {
+    weights;
+    intercept;
+    r2;
+    constructor(weights, intercept, 
+    /** Coefficient of determination on the fitted data, clamped to [0, 1]. */
+    r2) {
+        this.weights = weights;
+        this.intercept = intercept;
+        this.r2 = r2;
+    }
+    value(x) {
+        let s = this.intercept;
+        for (let i = 0; i < this.weights.length; i++)
+            s += this.weights[i] * (x[i] ?? 0);
+        return s;
+    }
+    predict(x) {
+        return { scores: [this.value(x)], confidence: this.r2 };
+    }
+    exportParams() {
+        const out = new Float32Array(this.weights.length + 1);
+        out.set(this.weights, 0);
+        out[this.weights.length] = this.intercept;
+        return out;
+    }
+    importParams(params) {
+        if (params.length !== this.weights.length + 1) {
+            throw new Error(`LinearModel.importParams: expected ${this.weights.length + 1} values, got ${params.length}`);
+        }
+        for (let i = 0; i < this.weights.length; i++)
+            this.weights[i] = params[i];
+        this.intercept = params[this.weights.length];
+    }
+}
+/** Solve A·x = b in place (A is n×n row-major). Throws when singular. */
+export function solveLinearSystem(A, b, n) {
+    for (let col = 0; col < n; col++) {
+        let piv = col;
+        for (let r = col + 1; r < n; r++)
+            if (Math.abs(A[r * n + col]) > Math.abs(A[piv * n + col]))
+                piv = r;
+        if (Math.abs(A[piv * n + col]) < 1e-12)
+            throw new Error("solveLinearSystem: singular matrix");
+        if (piv !== col) {
+            for (let c = 0; c < n; c++) {
+                const t = A[col * n + c];
+                A[col * n + c] = A[piv * n + c];
+                A[piv * n + c] = t;
+            }
+            const t = b[col];
+            b[col] = b[piv];
+            b[piv] = t;
+        }
+        for (let r = col + 1; r < n; r++) {
+            const f = A[r * n + col] / A[col * n + col];
+            if (f === 0)
+                continue;
+            for (let c = col; c < n; c++)
+                A[r * n + c] = A[r * n + c] - f * A[col * n + c];
+            b[r] = b[r] - f * b[col];
+        }
+    }
+    const x = new Float64Array(n);
+    for (let r = n - 1; r >= 0; r--) {
+        let s = b[r];
+        for (let c = r + 1; c < n; c++)
+            s -= A[r * n + c] * x[c];
+        x[r] = s / A[r * n + r];
+    }
+    return x;
+}
+export function fitLinear(samples, opts = {}) {
+    const d = featureDim(samples.map((s) => s.x));
+    const n = d + 1; // last column = intercept
+    const A = new Float64Array(n * n);
+    const b = new Float64Array(n);
+    const row = new Float64Array(n);
+    for (const s of samples) {
+        const w = s.weight ?? 1;
+        for (let i = 0; i < d; i++)
+            row[i] = s.x[i];
+        row[d] = 1;
+        for (let i = 0; i < n; i++) {
+            b[i] = b[i] + w * row[i] * s.y;
+            for (let j = 0; j < n; j++)
+                A[i * n + j] = A[i * n + j] + w * row[i] * row[j];
+        }
+    }
+    const l2 = opts.l2 ?? 1e-6;
+    for (let i = 0; i < d; i++)
+        A[i * n + i] = A[i * n + i] + l2;
+    const sol = solveLinearSystem(A, b, n);
+    const model = new LinearModel(Float32Array.from(sol.subarray(0, d)), sol[d], 0);
+    // R² on the fitted data — the confidence the runner reports.
+    let wSum = 0;
+    let mean = 0;
+    for (const s of samples) {
+        wSum += s.weight ?? 1;
+        mean += (s.weight ?? 1) * s.y;
+    }
+    mean /= wSum;
+    let ssRes = 0;
+    let ssTot = 0;
+    for (const s of samples) {
+        const w = s.weight ?? 1;
+        ssRes += w * (s.y - model.value(s.x)) ** 2;
+        ssTot += w * (s.y - mean) ** 2;
+    }
+    const r2 = ssTot > 0 ? Math.min(1, Math.max(0, 1 - ssRes / ssTot)) : ssRes < 1e-12 ? 1 : 0;
+    return new LinearModel(model.weights, model.intercept, r2);
+}
+//# sourceMappingURL=linear.js.map

package/dist/ml/classical/linear.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"linear.js","sourceRoot":"","sources":["../../../src/ml/classical/linear.ts"],"names":[],"mappings":"AAQA,OAAO,EAAE,UAAU,EAAE,MAAM,WAAW,CAAC;AAOvC,MAAM,OAAO,WAAW;IAEX;IACF;IAEE;IAJX,YACW,OAAqB,EACvB,SAAiB;IACxB,0EAA0E;IACjE,EAAU;QAHV,YAAO,GAAP,OAAO,CAAc;QACvB,cAAS,GAAT,SAAS,CAAQ;QAEf,OAAE,GAAF,EAAE,CAAQ;IAClB,CAAC;IAEJ,KAAK,CAAC,CAAoB;QACxB,IAAI,CAAC,GAAG,IAAI,CAAC,SAAS,CAAC;QACvB,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,IAAI,CAAC,OAAO,CAAC,MAAM,EAAE,CAAC,EAAE;YAAE,CAAC,IAAI,IAAI,CAAC,OAAO,CAAC,CAAC,CAAE,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC;QAClF,OAAO,CAAC,CAAC;IACX,CAAC;IAED,OAAO,CAAC,CAAe;QACrB,OAAO,EAAE,MAAM,EAAE,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,EAAE,UAAU,EAAE,IAAI,CAAC,EAAE,EAAE,CAAC;IAC1D,CAAC;IAED,YAAY;QACV,MAAM,GAAG,GAAG,IAAI,YAAY,CAAC,IAAI,CAAC,OAAO,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC;QACtD,GAAG,CAAC,GAAG,CAAC,IAAI,CAAC,OAAO,EAAE,CAAC,CAAC,CAAC;QACzB,GAAG,CAAC,IAAI,CAAC,OAAO,CAAC,MAAM,CAAC,GAAG,IAAI,CAAC,SAAS,CAAC;QAC1C,OAAO,GAAG,CAAC;IACb,CAAC;IAED,YAAY,CAAC,MAAyB;QACpC,IAAI,MAAM,CAAC,MAAM,KAAK,IAAI,CAAC,OAAO,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;YAC9C,MAAM,IAAI,KAAK,CAAC,sCAAsC,IAAI,CAAC,OAAO,CAAC,MAAM,GAAG,CAAC,gBAAgB,MAAM,CAAC,MAAM,EAAE,CAAC,CAAC;QAChH,CAAC;QACD,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,IAAI,CAAC,OAAO,CAAC,MAAM,EAAE,CAAC,EAAE;YAAE,IAAI,CAAC,OAAO,CAAC,CAAC,CAAC,GAAG,MAAM,CAAC,CAAC,CAAE,CAAC;QAC3E,IAAI,CAAC,SAAS,GAAG,MAAM,CAAC,IAAI,CAAC,OAAO,CAAC,MAAM,CAAE,CAAC;IAChD,CAAC;CACF;AAED,yEAAyE;AACzE,MAAM,UAAU,iBAAiB,CAAC,CAAe,EAAE,CAAe,EAAE,CAAS;IAC3E,KAAK,IAAI,GAAG,GAAG,CAAC,EAAE,GAAG,GAAG,CAAC,EAAE,GAAG,EAAE,EAAE,CAAC;QACjC,IAAI,GAAG,GAAG,GAAG,CAAC;QACd,KAAK,IAAI,CAAC,GAAG,GAAG,GAAG,CAAC,EAAE,CAAC,GAAG,CAAC,EAAE,CAAC,EAAE;YAAE,IAAI,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,GAAG,GAAG,CAAE,CAAC,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,GAAG,GAAG,CAAC,GAAG,GAAG,CAAE,CAAC;gBAAE,GAAG,GAAG,CAAC,CAAC;QACvG,IAAI,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,GAAG,GAAG,CAAC,GAAG,GAAG,CAAE,CAAC,GAAG,KAAK;YAAE,MAAM,IAAI,KAAK,CAAC,oCAAoC,CAAC,CAAC;QAC/F,IAAI,GAAG,KAAK,GAAG,EAAE,CAAC;YAChB,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC;gBAC3B,MAAM,CAAC,GAAG,CAAC,CAAC,GAAG,GAAG,CAAC,GAAG,CAAC,CAAE,CAAC;gBAC1B,CAAC,CAAC,GAAG,GAAG,CAAC,GAAG,CAAC,CAAC,GAAG,CAAC,CAAC,GAAG,GAAG,CAAC,GAAG,CAAC,CAAE,CAAC;gBACjC,CAAC,CAAC,GAAG,GAAG,CAAC,GAAG,CAAC,CAAC,GAAG,CAAC,CAAC;YACrB,CAAC;YACD,MAAM,CAAC,GAAG,CAAC,CAAC,GAAG,CAAE,CAAC;YAClB,CAAC,CAAC,GAAG,CAAC,GAAG,CAAC,CAAC,GAAG,CAAE,CAAC;YACjB,CAAC,CAAC,GAAG,CAAC,GAAG,CAAC,CAAC;QACb,CAAC;QACD,KAAK,IAAI,CAAC,GAAG,GAAG,GAAG,CAAC,EAAE,CAAC,GAAG,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC;YACjC,MAAM,CAAC,GAAG,CAAC,CAAC,CAAC,GAAG,CAAC,GAAG,GAAG,CAAE,GAAG,CAAC,CAAC,GAAG,GAAG,CAAC,GAAG,GAAG,CAAE,CAAC;YAC9C,IAAI,CAAC,KAAK,CAAC;gBAAE,SAAS;YACtB,KAAK,IAAI,CAAC,GAAG,GAAG,EAAE,CAAC,GAAG,CAAC,EAAE,CAAC,EAAE;gBAAE,CAAC,CAAC,CAAC,GAAG,CAAC,GAAG,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,GAAG,CAAC,GAAG,CAAC,CAAE,GAAG,CAAC,GAAG,CAAC,CAAC,GAAG,GAAG,CAAC,GAAG,CAAC,CAAE,CAAC;YACjF,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,CAAE,GAAG,CAAC,GAAG,CAAC,CAAC,GAAG,CAAE,CAAC;QAC7B,CAAC;IACH,CAAC;IACD,MAAM,CAAC,GAAG,IAAI,YAAY,CAAC,CAAC,CAAC,CAAC;IAC9B,KAAK,IAAI,CAAC,GAAG,CAAC,GAAG,CAAC,EAAE,CAAC,IAAI,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC;QAChC,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,CAAE,CAAC;QACd,KAAK,IAAI,CAAC,GAAG,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,CAAC,EAAE,CAAC,EAAE;YAAE,CAAC,IAAI,CAAC,CAAC,CAAC,GAAG,CAAC,GAAG,CAAC,CAAE,GAAG,CAAC,CAAC,CAAC,CAAE,CAAC;QAC3D,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,GAAG,CAAC,CAAC,CAAC,GAAG,CAAC,GAAG,CAAC,CAAE,CAAC;IAC3B,CAAC;IACD,OAAO,CAAC,CAAC;AACX,CAAC;AAED,MAAM,UAAU,SAAS,CAAC,OAA4B,EAAE,OAAsB,EAAE;IAC9E,MAAM,CAAC,GAAG,UAAU,CAAC,OAAO,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;IAC9C,MAAM,CAAC,GAAG,CAAC,GAAG,CAAC,CAAC,CAAC,0BAA0B;IAC3C,MAAM,CAAC,GAAG,IAAI,YAAY,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC;IAClC,MAAM,CAAC,GAAG,IAAI,YAAY,CAAC,CAAC,CAAC,CAAC;IAC9B,MAAM,GAAG,GAAG,IAAI,YAAY,CAAC,CAAC,CAAC,CAAC;IAChC,KAAK,MAAM,CAAC,IAAI,OAAO,EAAE,CAAC;QACxB,MAAM,CAAC,GAAG,CAAC,CAAC,MAAM,IAAI,CAAC,CAAC;QACxB,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,CAAC,EAAE,CAAC,EAAE;YAAE,GAAG,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC,CAAE,CAAC;QAC7C,GAAG,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC;QACX,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC;YAC3B,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,CAAE,GAAG,CAAC,GAAG,GAAG,CAAC,CAAC,CAAE,GAAG,CAAC,CAAC,CAAC,CAAC;YACjC,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,CAAC,EAAE,CAAC,EAAE;gBAAE,CAAC,CAAC,CAAC,GAAG,CAAC,GAAG,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,GAAG,CAAC,GAAG,CAAC,CAAE,GAAG,CAAC,GAAG,GAAG,CAAC,CAAC,CAAE,GAAG,GAAG,CAAC,CAAC,CAAE,CAAC;QACnF,CAAC;IACH,CAAC;IACD,MAAM,EAAE,GAAG,IAAI,CAAC,EAAE,IAAI,IAAI,CAAC;IAC3B,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,CAAC,EAAE,CAAC,EAAE;QAAE,CAAC,CAAC,CAAC,GAAG,CAAC,GAAG,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,GAAG,CAAC,GAAG,CAAC,CAAE,GAAG,EAAE,CAAC;IAC9D,MAAM,GAAG,GAAG,iBAAiB,CAAC,CAAC,EAAE,CAAC,EAAE,CAAC,CAAC,CAAC;IACvC,MAAM,KAAK,GAAG,IAAI,WAAW,CAAC,YAAY,CAAC,IAAI,CAAC,GAAG,CAAC,QAAQ,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,EAAE,GAAG,CAAC,CAAC,CAAE,EAAE,CAAC,CAAC,CAAC;IAEjF,6DAA6D;IAC7D,IAAI,IAAI,GAAG,CAAC,CAAC;IACb,IAAI,IAAI,GAAG,CAAC,CAAC;IACb,KAAK,MAAM,CAAC,IAAI,OAAO,EAAE,CAAC;QACxB,IAAI,IAAI,CAAC,CAAC,MAAM,IAAI,CAAC,CAAC;QACtB,IAAI,IAAI,CAAC,CAAC,CAAC,MAAM,IAAI,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC;IAChC,CAAC;IACD,IAAI,IAAI,IAAI,CAAC;IACb,IAAI,KAAK,GAAG,CAAC,CAAC;IACd,IAAI,KAAK,GAAG,CAAC,CAAC;IACd,KAAK,MAAM,CAAC,IAAI,OAAO,EAAE,CAAC;QACxB,MAAM,CAAC,GAAG,CAAC,CAAC,MAAM,IAAI,CAAC,CAAC;QACxB,KAAK,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,GAAG,KAAK,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC;QAC3C,KAAK,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,GAAG,IAAI,CAAC,IAAI,CAAC,CAAC;IACjC,CAAC;IACD,MAAM,EAAE,GAAG,KAAK,GAAG,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,CAAC,GAAG,KAAK,GAAG,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,KAAK,GAAG,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;IAC3F,OAAO,IAAI,WAAW,CAAC,KAAK,CAAC,OAAO,EAAE,KAAK,CAAC,SAAS,EAAE,EAAE,CAAC,CAAC;AAC7D,CAAC"}

package/dist/ml/classical/logistic.d.ts

@@ -0,0 +1,39 @@
+/**
+ * ml/classical/logistic.ts — (multinomial) logistic regression.
+ *
+ * Softmax regression with L2, fitted by full-batch gradient descent; two classes
+ * is ordinary logistic regression. No internal feature scaling: a task supplies
+ * features on a stable scale, so a weight means the same thing on every machine —
+ * which is what lets trained weights FEDERATE (`exportParams` / `importParams`).
+ */
+import type { MlPrediction, MlPredictor, MlSample } from "../types.js";
+export interface LogisticOptions {
+    classes?: number;
+    learningRate?: number;
+    epochs?: number;
+    /** L2 penalty on weights (not biases). Default 1e-3. */
+    l2?: number;
+}
+export declare class LogisticModel implements MlPredictor {
+    readonly dim: number;
+    readonly classes: number;
+    /** Row-major weights [classes × dim]. */
+    readonly weights: Float32Array;
+    readonly bias: Float32Array;
+    constructor(dim: number, classes: number);
+    logits(x: ArrayLike<number>): number[];
+    proba(x: ArrayLike<number>): number[];
+    predict(x: Float32Array): MlPrediction;
+    /** Weighted mean cross-entropy over `samples` (the eval-gate loss). */
+    loss(samples: readonly MlSample[]): number;
+    /** One full-batch gradient step; returns the pre-step mean loss. */
+    step(samples: readonly MlSample[], learningRate: number, l2: number): number;
+    exportParams(): Float32Array;
+    importParams(params: ArrayLike<number>): void;
+}
+/**
+ * Fit logistic regression. Pass `init` to continue from existing weights (a local
+ * round on top of a merged federated head).
+ */
+export declare function fitLogistic(samples: readonly MlSample[], opts?: LogisticOptions, init?: LogisticModel): LogisticModel;
+//# sourceMappingURL=logistic.d.ts.map

package/dist/ml/classical/logistic.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"logistic.d.ts","sourceRoot":"","sources":["../../../src/ml/classical/logistic.ts"],"names":[],"mappings":"AAAA;;;;;;;GAOG;AACH,OAAO,KAAK,EAAE,YAAY,EAAE,WAAW,EAAE,QAAQ,EAAE,MAAM,aAAa,CAAC;AAGvE,MAAM,WAAW,eAAe;IAC9B,OAAO,CAAC,EAAE,MAAM,CAAC;IACjB,YAAY,CAAC,EAAE,MAAM,CAAC;IACtB,MAAM,CAAC,EAAE,MAAM,CAAC;IAChB,wDAAwD;IACxD,EAAE,CAAC,EAAE,MAAM,CAAC;CACb;AAED,qBAAa,aAAc,YAAW,WAAW;IAM7C,QAAQ,CAAC,GAAG,EAAE,MAAM;IACpB,QAAQ,CAAC,OAAO,EAAE,MAAM;IAN1B,yCAAyC;IACzC,QAAQ,CAAC,OAAO,EAAE,YAAY,CAAC;IAC/B,QAAQ,CAAC,IAAI,EAAE,YAAY,CAAC;gBAGjB,GAAG,EAAE,MAAM,EACX,OAAO,EAAE,MAAM;IAM1B,MAAM,CAAC,CAAC,EAAE,SAAS,CAAC,MAAM,CAAC,GAAG,MAAM,EAAE;IAWtC,KAAK,CAAC,CAAC,EAAE,SAAS,CAAC,MAAM,CAAC,GAAG,MAAM,EAAE;IAIrC,OAAO,CAAC,CAAC,EAAE,YAAY,GAAG,YAAY;IAKtC,uEAAuE;IACvE,IAAI,CAAC,OAAO,EAAE,SAAS,QAAQ,EAAE,GAAG,MAAM;IAW1C,oEAAoE;IACpE,IAAI,CAAC,OAAO,EAAE,SAAS,QAAQ,EAAE,EAAE,YAAY,EAAE,MAAM,EAAE,EAAE,EAAE,MAAM,GAAG,MAAM;IAyB5E,YAAY,IAAI,YAAY;IAO5B,YAAY,CAAC,MAAM,EAAE,SAAS,CAAC,MAAM,CAAC,GAAG,IAAI;CAO9C;AAED;;;GAGG;AACH,wBAAgB,WAAW,CACzB,OAAO,EAAE,SAAS,QAAQ,EAAE,EAC5B,IAAI,GAAE,eAAoB,EAC1B,IAAI,CAAC,EAAE,aAAa,GACnB,aAAa,CAUf"}

package/dist/ml/classical/logistic.js

@@ -0,0 +1,103 @@
+import { classCount, featureDim, softmax, topProbability } from "./math.js";
+export class LogisticModel {
+    dim;
+    classes;
+    /** Row-major weights [classes × dim]. */
+    weights;
+    bias;
+    constructor(dim, classes) {
+        this.dim = dim;
+        this.classes = classes;
+        this.weights = new Float32Array(classes * dim);
+        this.bias = new Float32Array(classes);
+    }
+    logits(x) {
+        const z = [];
+        for (let c = 0; c < this.classes; c++) {
+            let s = this.bias[c];
+            const off = c * this.dim;
+            for (let i = 0; i < this.dim; i++)
+                s += this.weights[off + i] * (x[i] ?? 0);
+            z.push(s);
+        }
+        return z;
+    }
+    proba(x) {
+        return Array.from(softmax(this.logits(x)));
+    }
+    predict(x) {
+        const scores = this.proba(x);
+        return { scores, confidence: topProbability(scores) };
+    }
+    /** Weighted mean cross-entropy over `samples` (the eval-gate loss). */
+    loss(samples) {
+        let sum = 0;
+        let wSum = 0;
+        for (const s of samples) {
+            const w = s.weight ?? 1;
+            sum += -w * Math.log(Math.max(this.proba(s.x)[s.y] ?? 0, 1e-12));
+            wSum += w;
+        }
+        return wSum > 0 ? sum / wSum : 0;
+    }
+    /** One full-batch gradient step; returns the pre-step mean loss. */
+    step(samples, learningRate, l2) {
+        const gW = new Float64Array(this.weights.length);
+        const gB = new Float64Array(this.classes);
+        let loss = 0;
+        let wSum = 0;
+        for (const s of samples) {
+            const w = s.weight ?? 1;
+            const p = this.proba(s.x);
+            loss += -w * Math.log(Math.max(p[s.y], 1e-12));
+            wSum += w;
+            for (let c = 0; c < this.classes; c++) {
+                const g = w * (p[c] - (c === s.y ? 1 : 0));
+                gB[c] = gB[c] + g;
+                const off = c * this.dim;
+                for (let i = 0; i < this.dim; i++)
+                    gW[off + i] = gW[off + i] + g * s.x[i];
+            }
+        }
+        const inv = wSum > 0 ? 1 / wSum : 0;
+        for (let j = 0; j < this.weights.length; j++) {
+            this.weights[j] = this.weights[j] - learningRate * (gW[j] * inv + l2 * this.weights[j]);
+        }
+        for (let c = 0; c < this.classes; c++)
+            this.bias[c] = this.bias[c] - learningRate * gB[c] * inv;
+        return loss * inv;
+    }
+    exportParams() {
+        const out = new Float32Array(this.weights.length + this.bias.length);
+        out.set(this.weights, 0);
+        out.set(this.bias, this.weights.length);
+        return out;
+    }
+    importParams(params) {
+        if (params.length !== this.weights.length + this.bias.length) {
+            throw new Error(`LogisticModel.importParams: expected ${this.weights.length + this.bias.length} values, got ${params.length}`);
+        }
+        for (let j = 0; j < this.weights.length; j++)
+            this.weights[j] = params[j];
+        for (let c = 0; c < this.bias.length; c++)
+            this.bias[c] = params[this.weights.length + c];
+    }
+}
+/**
+ * Fit logistic regression. Pass `init` to continue from existing weights (a local
+ * round on top of a merged federated head).
+ */
+export function fitLogistic(samples, opts = {}, init) {
+    const d = featureDim(samples.map((s) => s.x));
+    const k = opts.classes ?? init?.classes ?? classCount(samples.map((s) => s.y));
+    const model = init ?? new LogisticModel(d, k);
+    if (model.dim !== d || model.classes !== k)
+        throw new Error(`fitLogistic: init is ${model.dim}×${model.classes}, data is ${d}×${k}`);
+    const lr = opts.learningRate ?? 0.5;
+    const l2 = opts.l2 ?? 1e-3;
+    const epochs = opts.epochs ?? 300;
+    for (let e = 0; e < epochs; e++)
+        model.step(samples, lr, l2);
+    return model;
+}
+//# sourceMappingURL=logistic.js.map

package/dist/ml/classical/logistic.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"logistic.js","sourceRoot":"","sources":["../../../src/ml/classical/logistic.ts"],"names":[],"mappings":"AASA,OAAO,EAAE,UAAU,EAAE,UAAU,EAAE,OAAO,EAAE,cAAc,EAAE,MAAM,WAAW,CAAC;AAU5E,MAAM,OAAO,aAAa;IAMb;IACA;IANX,yCAAyC;IAChC,OAAO,CAAe;IACtB,IAAI,CAAe;IAE5B,YACW,GAAW,EACX,OAAe;QADf,QAAG,GAAH,GAAG,CAAQ;QACX,YAAO,GAAP,OAAO,CAAQ;QAExB,IAAI,CAAC,OAAO,GAAG,IAAI,YAAY,CAAC,OAAO,GAAG,GAAG,CAAC,CAAC;QAC/C,IAAI,CAAC,IAAI,GAAG,IAAI,YAAY,CAAC,OAAO,CAAC,CAAC;IACxC,CAAC;IAED,MAAM,CAAC,CAAoB;QACzB,MAAM,CAAC,GAAa,EAAE,CAAC;QACvB,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,IAAI,CAAC,OAAO,EAAE,CAAC,EAAE,EAAE,CAAC;YACtC,IAAI,CAAC,GAAG,IAAI,CAAC,IAAI,CAAC,CAAC,CAAE,CAAC;YACtB,MAAM,GAAG,GAAG,CAAC,GAAG,IAAI,CAAC,GAAG,CAAC;YACzB,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,IAAI,CAAC,GAAG,EAAE,CAAC,EAAE;gBAAE,CAAC,IAAI,IAAI,CAAC,OAAO,CAAC,GAAG,GAAG,CAAC,CAAE,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC;YAC7E,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;QACZ,CAAC;QACD,OAAO,CAAC,CAAC;IACX,CAAC;IAED,KAAK,CAAC,CAAoB;QACxB,OAAO,KAAK,CAAC,IAAI,CAAC,OAAO,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;IAC7C,CAAC;IAED,OAAO,CAAC,CAAe;QACrB,MAAM,MAAM,GAAG,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC;QAC7B,OAAO,EAAE,MAAM,EAAE,UAAU,EAAE,cAAc,CAAC,MAAM,CAAC,EAAE,CAAC;IACxD,CAAC;IAED,uEAAuE;IACvE,IAAI,CAAC,OAA4B;QAC/B,IAAI,GAAG,GAAG,CAAC,CAAC;QACZ,IAAI,IAAI,GAAG,CAAC,CAAC;QACb,KAAK,MAAM,CAAC,IAAI,OAAO,EAAE,CAAC;YACxB,MAAM,CAAC,GAAG,CAAC,CAAC,MAAM,IAAI,CAAC,CAAC;YACxB,GAAG,IAAI,CAAC,CAAC,GAAG,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,EAAE,KAAK,CAAC,CAAC,CAAC;YACjE,IAAI,IAAI,CAAC,CAAC;QACZ,CAAC;QACD,OAAO,IAAI,GAAG,CAAC,CAAC,CAAC,CAAC,GAAG,GAAG,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC;IACnC,CAAC;IAED,oEAAoE;IACpE,IAAI,CAAC,OAA4B,EAAE,YAAoB,EAAE,EAAU;QACjE,MAAM,EAAE,GAAG,IAAI,YAAY,CAAC,IAAI,CAAC,OAAO,CAAC,MAAM,CAAC,CAAC;QACjD,MAAM,EAAE,GAAG,IAAI,YAAY,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC;QAC1C,IAAI,IAAI,GAAG,CAAC,CAAC;QACb,IAAI,IAAI,GAAG,CAAC,CAAC;QACb,KAAK,MAAM,CAAC,IAAI,OAAO,EAAE,CAAC;YACxB,MAAM,CAAC,GAAG,CAAC,CAAC,MAAM,IAAI,CAAC,CAAC;YACxB,MAAM,CAAC,GAAG,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;YAC1B,IAAI,IAAI,CAAC,CAAC,GAAG,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAE,EAAE,KAAK,CAAC,CAAC,CAAC;YAChD,IAAI,IAAI,CAAC,CAAC;YACV,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,IAAI,CAAC,OAAO,EAAE,CAAC,EAAE,EAAE,CAAC;gBACtC,MAAM,CAAC,GAAG,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAE,GAAG,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;gBAC5C,EAAE,CAAC,CAAC,CAAC,GAAG,EAAE,CAAC,CAAC,CAAE,GAAG,CAAC,CAAC;gBACnB,MAAM,GAAG,GAAG,CAAC,GAAG,IAAI,CAAC,GAAG,CAAC;gBACzB,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,IAAI,CAAC,GAAG,EAAE,CAAC,EAAE;oBAAE,EAAE,CAAC,GAAG,GAAG,CAAC,CAAC,GAAG,EAAE,CAAC,GAAG,GAAG,CAAC,CAAE,GAAG,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC,CAAE,CAAC;YAC9E,CAAC;QACH,CAAC;QACD,MAAM,GAAG,GAAG,IAAI,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC,GAAG,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC;QACpC,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,IAAI,CAAC,OAAO,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;YAC7C,IAAI,CAAC,OAAO,CAAC,CAAC,CAAC,GAAG,IAAI,CAAC,OAAO,CAAC,CAAC,CAAE,GAAG,YAAY,GAAG,CAAC,EAAE,CAAC,CAAC,CAAE,GAAG,GAAG,GAAG,EAAE,GAAG,IAAI,CAAC,OAAO,CAAC,CAAC,CAAE,CAAC,CAAC;QAC7F,CAAC;QACD,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,IAAI,CAAC,OAAO,EAAE,CAAC,EAAE;YAAE,IAAI,CAAC,IAAI,CAAC,CAAC,CAAC,GAAG,IAAI,CAAC,IAAI,CAAC,CAAC,CAAE,GAAG,YAAY,GAAG,EAAE,CAAC,CAAC,CAAE,GAAG,GAAG,CAAC;QAClG,OAAO,IAAI,GAAG,GAAG,CAAC;IACpB,CAAC;IAED,YAAY;QACV,MAAM,GAAG,GAAG,IAAI,YAAY,CAAC,IAAI,CAAC,OAAO,CAAC,MAAM,GAAG,IAAI,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;QACrE,GAAG,CAAC,GAAG,CAAC,IAAI,CAAC,OAAO,EAAE,CAAC,CAAC,CAAC;QACzB,GAAG,CAAC,GAAG,CAAC,IAAI,CAAC,IAAI,EAAE,IAAI,CAAC,OAAO,CAAC,MAAM,CAAC,CAAC;QACxC,OAAO,GAAG,CAAC;IACb,CAAC;IAED,YAAY,CAAC,MAAyB;QACpC,IAAI,MAAM,CAAC,MAAM,KAAK,IAAI,CAAC,OAAO,CAAC,MAAM,GAAG,IAAI,CAAC,IAAI,CAAC,MAAM,EAAE,CAAC;YAC7D,MAAM,IAAI,KAAK,CAAC,wCAAwC,IAAI,CAAC,OAAO,CAAC,MAAM,GAAG,IAAI,CAAC,IAAI,CAAC,MAAM,gBAAgB,MAAM,CAAC,MAAM,EAAE,CAAC,CAAC;QACjI,CAAC;QACD,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,IAAI,CAAC,OAAO,CAAC,MAAM,EAAE,CAAC,EAAE;YAAE,IAAI,CAAC,OAAO,CAAC,CAAC,CAAC,GAAG,MAAM,CAAC,CAAC,CAAE,CAAC;QAC3E,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,IAAI,CAAC,IAAI,CAAC,MAAM,EAAE,CAAC,EAAE;YAAE,IAAI,CAAC,IAAI,CAAC,CAAC,CAAC,GAAG,MAAM,CAAC,IAAI,CAAC,OAAO,CAAC,MAAM,GAAG,CAAC,CAAE,CAAC;IAC7F,CAAC;CACF;AAED;;;GAGG;AACH,MAAM,UAAU,WAAW,CACzB,OAA4B,EAC5B,OAAwB,EAAE,EAC1B,IAAoB;IAEpB,MAAM,CAAC,GAAG,UAAU,CAAC,OAAO,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;IAC9C,MAAM,CAAC,GAAG,IAAI,CAAC,OAAO,IAAI,IAAI,EAAE,OAAO,IAAI,UAAU,CAAC,OAAO,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;IAC/E,MAAM,KAAK,GAAG,IAAI,IAAI,IAAI,aAAa,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC;IAC9C,IAAI,KAAK,CAAC,GAAG,KAAK,CAAC,IAAI,KAAK,CAAC,OAAO,KAAK,CAAC;QAAE,MAAM,IAAI,KAAK,CAAC,wBAAwB,KAAK,CAAC,GAAG,IAAI,KAAK,CAAC,OAAO,aAAa,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IACrI,MAAM,EAAE,GAAG,IAAI,CAAC,YAAY,IAAI,GAAG,CAAC;IACpC,MAAM,EAAE,GAAG,IAAI,CAAC,EAAE,IAAI,IAAI,CAAC;IAC3B,MAAM,MAAM,GAAG,IAAI,CAAC,MAAM,IAAI,GAAG,CAAC;IAClC,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,MAAM,EAAE,CAAC,EAAE;QAAE,KAAK,CAAC,IAAI,CAAC,OAAO,EAAE,EAAE,EAAE,EAAE,CAAC,CAAC;IAC7D,OAAO,KAAK,CAAC;AACf,CAAC"}

package/dist/ml/classical/math.d.ts

@@ -0,0 +1,26 @@
+/** The engine's ONE sigmoid / softmax (`training/cpu_ops.ts`). */
+export { sigmoid, softmax } from "../../training/cpu_ops.js";
+export type Vec = ArrayLike<number>;
+export declare function dot(a: Vec, b: Vec): number;
+export declare function norm(a: Vec): number;
+/** Cosine similarity in [-1, 1]; 0 when either vector is zero. */
+export declare function cosine(a: Vec, b: Vec): number;
+export declare function squaredDistance(a: Vec, b: Vec): number;
+/** Median of a numeric list (copy-sorted). NaN for an empty list. */
+export declare function median(values: readonly number[]): number;
+/** Index of the largest value (first on ties). */
+export declare function argmax(values: ArrayLike<number>): number;
+/** Number of distinct classes implied by integer labels 0..k-1. */
+export declare function classCount(labels: readonly number[]): number;
+/** Feature dimension of a non-empty sample matrix; throws on ragged rows. */
+export declare function featureDim(X: readonly Vec[]): number;
+/** Confidence of a probability vector: its top probability. */
+export declare function topProbability(p: readonly number[]): number;
+/** A deterministic [0,1) source for algorithms that sample — the engine's ONE seeded RNG. */
+export declare function seededUniform(seed: number): () => number;
+/**
+ * Average path length of an unsuccessful BST search over n points — the
+ * isolation-forest normaliser c(n).
+ */
+export declare function averagePathLength(n: number): number;
+//# sourceMappingURL=math.d.ts.map

package/dist/ml/classical/math.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"math.d.ts","sourceRoot":"","sources":["../../../src/ml/classical/math.ts"],"names":[],"mappings":"AAMA,kEAAkE;AAClE,OAAO,EAAE,OAAO,EAAE,OAAO,EAAE,MAAM,2BAA2B,CAAC;AAE7D,MAAM,MAAM,GAAG,GAAG,SAAS,CAAC,MAAM,CAAC,CAAC;AAEpC,wBAAgB,GAAG,CAAC,CAAC,EAAE,GAAG,EAAE,CAAC,EAAE,GAAG,GAAG,MAAM,CAK1C;AAED,wBAAgB,IAAI,CAAC,CAAC,EAAE,GAAG,GAAG,MAAM,CAEnC;AAED,kEAAkE;AAClE,wBAAgB,MAAM,CAAC,CAAC,EAAE,GAAG,EAAE,CAAC,EAAE,GAAG,GAAG,MAAM,CAI7C;AAED,wBAAgB,eAAe,CAAC,CAAC,EAAE,GAAG,EAAE,CAAC,EAAE,GAAG,GAAG,MAAM,CAQtD;AAED,qEAAqE;AACrE,wBAAgB,MAAM,CAAC,MAAM,EAAE,SAAS,MAAM,EAAE,GAAG,MAAM,CAKxD;AAED,kDAAkD;AAClD,wBAAgB,MAAM,CAAC,MAAM,EAAE,SAAS,CAAC,MAAM,CAAC,GAAG,MAAM,CAIxD;AAED,mEAAmE;AACnE,wBAAgB,UAAU,CAAC,MAAM,EAAE,SAAS,MAAM,EAAE,GAAG,MAAM,CAO5D;AAED,6EAA6E;AAC7E,wBAAgB,UAAU,CAAC,CAAC,EAAE,SAAS,GAAG,EAAE,GAAG,MAAM,CAKpD;AAED,+DAA+D;AAC/D,wBAAgB,cAAc,CAAC,CAAC,EAAE,SAAS,MAAM,EAAE,GAAG,MAAM,CAE3D;AAED,6FAA6F;AAC7F,wBAAgB,aAAa,CAAC,IAAI,EAAE,MAAM,GAAG,MAAM,MAAM,CAGxD;AAED;;;GAGG;AACH,wBAAgB,iBAAiB,CAAC,CAAC,EAAE,MAAM,GAAG,MAAM,CAKnD"}

package/dist/ml/classical/math.js

@@ -0,0 +1,90 @@
+/**
+ * ml/classical/math.ts — the small vector helpers every classical algorithm shares.
+ * Pure, allocation-light, deterministic.
+ */
+import { SeededRng } from "../../utils/rng.js";
+/** The engine's ONE sigmoid / softmax (`training/cpu_ops.ts`). */
+export { sigmoid, softmax } from "../../training/cpu_ops.js";
+export function dot(a, b) {
+    const n = Math.min(a.length, b.length);
+    let s = 0;
+    for (let i = 0; i < n; i++)
+        s += a[i] * b[i];
+    return s;
+}
+export function norm(a) {
+    return Math.sqrt(dot(a, a));
+}
+/** Cosine similarity in [-1, 1]; 0 when either vector is zero. */
+export function cosine(a, b) {
+    const na = norm(a);
+    const nb = norm(b);
+    return na > 0 && nb > 0 ? dot(a, b) / (na * nb) : 0;
+}
+export function squaredDistance(a, b) {
+    const n = Math.min(a.length, b.length);
+    let s = 0;
+    for (let i = 0; i < n; i++) {
+        const d = a[i] - b[i];
+        s += d * d;
+    }
+    return s;
+}
+/** Median of a numeric list (copy-sorted). NaN for an empty list. */
+export function median(values) {
+    if (values.length === 0)
+        return Number.NaN;
+    const s = [...values].sort((a, b) => a - b);
+    const m = s.length >> 1;
+    return s.length % 2 ? s[m] : (s[m - 1] + s[m]) / 2;
+}
+/** Index of the largest value (first on ties). */
+export function argmax(values) {
+    let best = 0;
+    for (let i = 1; i < values.length; i++)
+        if (values[i] > values[best])
+            best = i;
+    return best;
+}
+/** Number of distinct classes implied by integer labels 0..k-1. */
+export function classCount(labels) {
+    let k = 0;
+    for (const y of labels) {
+        if (!Number.isInteger(y) || y < 0)
+            throw new Error(`class label must be a non-negative integer, got ${y}`);
+        k = Math.max(k, y + 1);
+    }
+    return Math.max(k, 2);
+}
+/** Feature dimension of a non-empty sample matrix; throws on ragged rows. */
+export function featureDim(X) {
+    if (X.length === 0)
+        throw new Error("no samples");
+    const d = X[0].length;
+    for (const x of X)
+        if (x.length !== d)
+            throw new Error(`ragged features: expected ${d}, got ${x.length}`);
+    return d;
+}
+/** Confidence of a probability vector: its top probability. */
+export function topProbability(p) {
+    return p.length ? Math.max(...p) : 0;
+}
+/** A deterministic [0,1) source for algorithms that sample — the engine's ONE seeded RNG. */
+export function seededUniform(seed) {
+    const rng = new SeededRng(seed);
+    return () => rng.next();
+}
+/**
+ * Average path length of an unsuccessful BST search over n points — the
+ * isolation-forest normaliser c(n).
+ */
+export function averagePathLength(n) {
+    if (n <= 1)
+        return 0;
+    if (n === 2)
+        return 1;
+    const harmonic = Math.log(n - 1) + 0.5772156649;
+    return 2 * harmonic - (2 * (n - 1)) / n;
+}
+//# sourceMappingURL=math.js.map

package/dist/ml/classical/math.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"math.js","sourceRoot":"","sources":["../../../src/ml/classical/math.ts"],"names":[],"mappings":"AAAA;;;GAGG;AACH,OAAO,EAAE,SAAS,EAAE,MAAM,oBAAoB,CAAC;AAE/C,kEAAkE;AAClE,OAAO,EAAE,OAAO,EAAE,OAAO,EAAE,MAAM,2BAA2B,CAAC;AAI7D,MAAM,UAAU,GAAG,CAAC,CAAM,EAAE,CAAM;IAChC,MAAM,CAAC,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,MAAM,EAAE,CAAC,CAAC,MAAM,CAAC,CAAC;IACvC,IAAI,CAAC,GAAG,CAAC,CAAC;IACV,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,CAAC,EAAE,CAAC,EAAE;QAAE,CAAC,IAAI,CAAC,CAAC,CAAC,CAAE,GAAG,CAAC,CAAC,CAAC,CAAE,CAAC;IAC/C,OAAO,CAAC,CAAC;AACX,CAAC;AAED,MAAM,UAAU,IAAI,CAAC,CAAM;IACzB,OAAO,IAAI,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC;AAC9B,CAAC;AAED,kEAAkE;AAClE,MAAM,UAAU,MAAM,CAAC,CAAM,EAAE,CAAM;IACnC,MAAM,EAAE,GAAG,IAAI,CAAC,CAAC,CAAC,CAAC;IACnB,MAAM,EAAE,GAAG,IAAI,CAAC,CAAC,CAAC,CAAC;IACnB,OAAO,EAAE,GAAG,CAAC,IAAI,EAAE,GAAG,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,EAAE,CAAC,CAAC,GAAG,CAAC,EAAE,GAAG,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;AACtD,CAAC;AAED,MAAM,UAAU,eAAe,CAAC,CAAM,EAAE,CAAM;IAC5C,MAAM,CAAC,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,MAAM,EAAE,CAAC,CAAC,MAAM,CAAC,CAAC;IACvC,IAAI,CAAC,GAAG,CAAC,CAAC;IACV,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC;QAC3B,MAAM,CAAC,GAAG,CAAC,CAAC,CAAC,CAAE,GAAG,CAAC,CAAC,CAAC,CAAE,CAAC;QACxB,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;IACb,CAAC;IACD,OAAO,CAAC,CAAC;AACX,CAAC;AAED,qEAAqE;AACrE,MAAM,UAAU,MAAM,CAAC,MAAyB;IAC9C,IAAI,MAAM,CAAC,MAAM,KAAK,CAAC;QAAE,OAAO,MAAM,CAAC,GAAG,CAAC;IAC3C,MAAM,CAAC,GAAG,CAAC,GAAG,MAAM,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC;IAC5C,MAAM,CAAC,GAAG,CAAC,CAAC,MAAM,IAAI,CAAC,CAAC;IACxB,OAAO,CAAC,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAE,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,CAAE,GAAG,CAAC,CAAC,CAAC,CAAE,CAAC,GAAG,CAAC,CAAC;AACxD,CAAC;AAED,kDAAkD;AAClD,MAAM,UAAU,MAAM,CAAC,MAAyB;IAC9C,IAAI,IAAI,GAAG,CAAC,CAAC;IACb,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,MAAM,CAAC,MAAM,EAAE,CAAC,EAAE;QAAE,IAAI,MAAM,CAAC,CAAC,CAAE,GAAG,MAAM,CAAC,IAAI,CAAE;YAAE,IAAI,GAAG,CAAC,CAAC;IACjF,OAAO,IAAI,CAAC;AACd,CAAC;AAED,mEAAmE;AACnE,MAAM,UAAU,UAAU,CAAC,MAAyB;IAClD,IAAI,CAAC,GAAG,CAAC,CAAC;IACV,KAAK,MAAM,CAAC,IAAI,MAAM,EAAE,CAAC;QACvB,IAAI,CAAC,MAAM,CAAC,SAAS,CAAC,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC;YAAE,MAAM,IAAI,KAAK,CAAC,mDAAmD,CAAC,EAAE,CAAC,CAAC;QAC3G,CAAC,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,CAAC,GAAG,CAAC,CAAC,CAAC;IACzB,CAAC;IACD,OAAO,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC;AACxB,CAAC;AAED,6EAA6E;AAC7E,MAAM,UAAU,UAAU,CAAC,CAAiB;IAC1C,IAAI,CAAC,CAAC,MAAM,KAAK,CAAC;QAAE,MAAM,IAAI,KAAK,CAAC,YAAY,CAAC,CAAC;IAClD,MAAM,CAAC,GAAG,CAAC,CAAC,CAAC,CAAE,CAAC,MAAM,CAAC;IACvB,KAAK,MAAM,CAAC,IAAI,CAAC;QAAE,IAAI,CAAC,CAAC,MAAM,KAAK,CAAC;YAAE,MAAM,IAAI,KAAK,CAAC,6BAA6B,CAAC,SAAS,CAAC,CAAC,MAAM,EAAE,CAAC,CAAC;IAC1G,OAAO,CAAC,CAAC;AACX,CAAC;AAED,+DAA+D;AAC/D,MAAM,UAAU,cAAc,CAAC,CAAoB;IACjD,OAAO,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;AACvC,CAAC;AAED,6FAA6F;AAC7F,MAAM,UAAU,aAAa,CAAC,IAAY;IACxC,MAAM,GAAG,GAAG,IAAI,SAAS,CAAC,IAAI,CAAC,CAAC;IAChC,OAAO,GAAG,EAAE,CAAC,GAAG,CAAC,IAAI,EAAE,CAAC;AAC1B,CAAC;AAED;;;GAGG;AACH,MAAM,UAAU,iBAAiB,CAAC,CAAS;IACzC,IAAI,CAAC,IAAI,CAAC;QAAE,OAAO,CAAC,CAAC;IACrB,IAAI,CAAC,KAAK,CAAC;QAAE,OAAO,CAAC,CAAC;IACtB,MAAM,QAAQ,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,GAAG,CAAC,CAAC,GAAG,YAAY,CAAC;IAChD,OAAO,CAAC,GAAG,QAAQ,GAAG,CAAC,CAAC,GAAG,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC;AAC1C,CAAC"}

package/dist/ml/classical/naive_bayes.d.ts

@@ -0,0 +1,24 @@
+/**
+ * ml/classical/naive_bayes.ts — Gaussian naive Bayes.
+ *
+ * Per class: a weighted prior, and an independent Gaussian per feature. Variance
+ * is smoothed by a fraction of the largest feature variance (as scikit-learn's
+ * `var_smoothing`), so a constant feature never divides by zero.
+ */
+import type { MlPrediction, MlPredictor, MlSample } from "../types.js";
+export interface NaiveBayesOptions {
+    classes?: number;
+    /** Fraction of the largest feature variance added to every variance. Default 1e-9. */
+    varSmoothing?: number;
+}
+export declare class NaiveBayesModel implements MlPredictor {
+    readonly logPrior: Float64Array;
+    readonly mean: Float64Array[];
+    readonly variance: Float64Array[];
+    constructor(logPrior: Float64Array, mean: Float64Array[], variance: Float64Array[]);
+    /** Per-class joint log-likelihood log P(c) + Σ log N(x_i | μ, σ²). */
+    jointLogLikelihood(x: ArrayLike<number>): number[];
+    predict(x: Float32Array): MlPrediction;
+}
+export declare function fitNaiveBayes(samples: readonly MlSample[], opts?: NaiveBayesOptions): NaiveBayesModel;
+//# sourceMappingURL=naive_bayes.d.ts.map

package/dist/ml/classical/naive_bayes.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"naive_bayes.d.ts","sourceRoot":"","sources":["../../../src/ml/classical/naive_bayes.ts"],"names":[],"mappings":"AAAA;;;;;;GAMG;AACH,OAAO,KAAK,EAAE,YAAY,EAAE,WAAW,EAAE,QAAQ,EAAE,MAAM,aAAa,CAAC;AAGvE,MAAM,WAAW,iBAAiB;IAChC,OAAO,CAAC,EAAE,MAAM,CAAC;IACjB,sFAAsF;IACtF,YAAY,CAAC,EAAE,MAAM,CAAC;CACvB;AAED,qBAAa,eAAgB,YAAW,WAAW;IAE/C,QAAQ,CAAC,QAAQ,EAAE,YAAY;IAC/B,QAAQ,CAAC,IAAI,EAAE,YAAY,EAAE;IAC7B,QAAQ,CAAC,QAAQ,EAAE,YAAY,EAAE;gBAFxB,QAAQ,EAAE,YAAY,EACtB,IAAI,EAAE,YAAY,EAAE,EACpB,QAAQ,EAAE,YAAY,EAAE;IAGnC,sEAAsE;IACtE,kBAAkB,CAAC,CAAC,EAAE,SAAS,CAAC,MAAM,CAAC,GAAG,MAAM,EAAE;IAYlD,OAAO,CAAC,CAAC,EAAE,YAAY,GAAG,YAAY;CAIvC;AAED,wBAAgB,aAAa,CAAC,OAAO,EAAE,SAAS,QAAQ,EAAE,EAAE,IAAI,GAAE,iBAAsB,GAAG,eAAe,CAkCzG"}

package/dist/ml/classical/naive_bayes.js

@@ -0,0 +1,67 @@
+import { classCount, featureDim, softmax, topProbability } from "./math.js";
+export class NaiveBayesModel {
+    logPrior;
+    mean;
+    variance;
+    constructor(logPrior, mean, variance) {
+        this.logPrior = logPrior;
+        this.mean = mean;
+        this.variance = variance;
+    }
+    /** Per-class joint log-likelihood log P(c) + Σ log N(x_i | μ, σ²). */
+    jointLogLikelihood(x) {
+        return this.mean.map((mu, c) => {
+            const v = this.variance[c];
+            let ll = this.logPrior[c];
+            for (let i = 0; i < mu.length; i++) {
+                const d = (x[i] ?? 0) - mu[i];
+                ll += -0.5 * Math.log(2 * Math.PI * v[i]) - (d * d) / (2 * v[i]);
+            }
+            return ll;
+        });
+    }
+    predict(x) {
+        const scores = Array.from(softmax(this.jointLogLikelihood(x)));
+        return { scores, confidence: topProbability(scores) };
+    }
+}
+export function fitNaiveBayes(samples, opts = {}) {
+    const d = featureDim(samples.map((s) => s.x));
+    const k = opts.classes ?? classCount(samples.map((s) => s.y));
+    const wSum = new Float64Array(k);
+    const mean = Array.from({ length: k }, () => new Float64Array(d));
+    const variance = Array.from({ length: k }, () => new Float64Array(d));
+    for (const s of samples) {
+        const w = s.weight ?? 1;
+        wSum[s.y] = wSum[s.y] + w;
+        for (let i = 0; i < d; i++)
+            mean[s.y][i] = mean[s.y][i] + w * s.x[i];
+    }
+    for (let c = 0; c < k; c++)
+        if (wSum[c] > 0)
+            for (let i = 0; i < d; i++)
+                mean[c][i] = mean[c][i] / wSum[c];
+    for (const s of samples) {
+        const w = s.weight ?? 1;
+        for (let i = 0; i < d; i++) {
+            const diff = s.x[i] - mean[s.y][i];
+            variance[s.y][i] = variance[s.y][i] + w * diff * diff;
+        }
+    }
+    let maxVar = 0;
+    for (let c = 0; c < k; c++) {
+        for (let i = 0; i < d; i++) {
+            variance[c][i] = wSum[c] > 0 ? variance[c][i] / wSum[c] : 0;
+            maxVar = Math.max(maxVar, variance[c][i]);
+        }
+    }
+    const eps = (opts.varSmoothing ?? 1e-9) * Math.max(maxVar, 1e-12) + 1e-12;
+    for (let c = 0; c < k; c++)
+        for (let i = 0; i < d; i++)
+            variance[c][i] = variance[c][i] + eps;
+    const total = wSum.reduce((a, b) => a + b, 0);
+    // An unseen class keeps a vanishing (not -∞) prior so softmax stays defined.
+    const logPrior = Float64Array.from(wSum, (w) => Math.log(Math.max(w, 1e-12) / total));
+    return new NaiveBayesModel(logPrior, mean, variance);
+}
+//# sourceMappingURL=naive_bayes.js.map

package/dist/ml/classical/naive_bayes.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"naive_bayes.js","sourceRoot":"","sources":["../../../src/ml/classical/naive_bayes.ts"],"names":[],"mappings":"AAQA,OAAO,EAAE,UAAU,EAAE,UAAU,EAAE,OAAO,EAAE,cAAc,EAAE,MAAM,WAAW,CAAC;AAQ5E,MAAM,OAAO,eAAe;IAEf;IACA;IACA;IAHX,YACW,QAAsB,EACtB,IAAoB,EACpB,QAAwB;QAFxB,aAAQ,GAAR,QAAQ,CAAc;QACtB,SAAI,GAAJ,IAAI,CAAgB;QACpB,aAAQ,GAAR,QAAQ,CAAgB;IAChC,CAAC;IAEJ,sEAAsE;IACtE,kBAAkB,CAAC,CAAoB;QACrC,OAAO,IAAI,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,EAAE,CAAC,EAAE,EAAE;YAC7B,MAAM,CAAC,GAAG,IAAI,CAAC,QAAQ,CAAC,CAAC,CAAE,CAAC;YAC5B,IAAI,EAAE,GAAG,IAAI,CAAC,QAAQ,CAAC,CAAC,CAAE,CAAC;YAC3B,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,EAAE,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;gBACnC,MAAM,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,GAAG,EAAE,CAAC,CAAC,CAAE,CAAC;gBAC/B,EAAE,IAAI,CAAC,GAAG,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,GAAG,IAAI,CAAC,EAAE,GAAG,CAAC,CAAC,CAAC,CAAE,CAAC,GAAG,CAAC,CAAC,GAAG,CAAC,CAAC,GAAG,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,CAAE,CAAC,CAAC;YACrE,CAAC;YACD,OAAO,EAAE,CAAC;QACZ,CAAC,CAAC,CAAC;IACL,CAAC;IAED,OAAO,CAAC,CAAe;QACrB,MAAM,MAAM,GAAG,KAAK,CAAC,IAAI,CAAC,OAAO,CAAC,IAAI,CAAC,kBAAkB,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;QAC/D,OAAO,EAAE,MAAM,EAAE,UAAU,EAAE,cAAc,CAAC,MAAM,CAAC,EAAE,CAAC;IACxD,CAAC;CACF;AAED,MAAM,UAAU,aAAa,CAAC,OAA4B,EAAE,OAA0B,EAAE;IACtF,MAAM,CAAC,GAAG,UAAU,CAAC,OAAO,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;IAC9C,MAAM,CAAC,GAAG,IAAI,CAAC,OAAO,IAAI,UAAU,CAAC,OAAO,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;IAC9D,MAAM,IAAI,GAAG,IAAI,YAAY,CAAC,CAAC,CAAC,CAAC;IACjC,MAAM,IAAI,GAAG,KAAK,CAAC,IAAI,CAAC,EAAE,MAAM,EAAE,CAAC,EAAE,EAAE,GAAG,EAAE,CAAC,IAAI,YAAY,CAAC,CAAC,CAAC,CAAC,CAAC;IAClE,MAAM,QAAQ,GAAG,KAAK,CAAC,IAAI,CAAC,EAAE,MAAM,EAAE,CAAC,EAAE,EAAE,GAAG,EAAE,CAAC,IAAI,YAAY,CAAC,CAAC,CAAC,CAAC,CAAC;IAEtE,KAAK,MAAM,CAAC,IAAI,OAAO,EAAE,CAAC;QACxB,MAAM,CAAC,GAAG,CAAC,CAAC,MAAM,IAAI,CAAC,CAAC;QACxB,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC,GAAG,IAAI,CAAC,CAAC,CAAC,CAAC,CAAE,GAAG,CAAC,CAAC;QAC3B,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,CAAC,EAAE,CAAC,EAAE;YAAE,IAAI,CAAC,CAAC,CAAC,CAAC,CAAE,CAAC,CAAC,CAAC,GAAG,IAAI,CAAC,CAAC,CAAC,CAAC,CAAE,CAAC,CAAC,CAAE,GAAG,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC,CAAE,CAAC;IAC3E,CAAC;IACD,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,CAAC,EAAE,CAAC,EAAE;QAAE,IAAI,IAAI,CAAC,CAAC,CAAE,GAAG,CAAC;YAAE,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,CAAC,EAAE,CAAC,EAAE;gBAAE,IAAI,CAAC,CAAC,CAAE,CAAC,CAAC,CAAC,GAAG,IAAI,CAAC,CAAC,CAAE,CAAC,CAAC,CAAE,GAAG,IAAI,CAAC,CAAC,CAAE,CAAC;IAChH,KAAK,MAAM,CAAC,IAAI,OAAO,EAAE,CAAC;QACxB,MAAM,CAAC,GAAG,CAAC,CAAC,MAAM,IAAI,CAAC,CAAC;QACxB,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC;YAC3B,MAAM,IAAI,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC,CAAE,GAAG,IAAI,CAAC,CAAC,CAAC,CAAC,CAAE,CAAC,CAAC,CAAE,CAAC;YACtC,QAAQ,CAAC,CAAC,CAAC,CAAC,CAAE,CAAC,CAAC,CAAC,GAAG,QAAQ,CAAC,CAAC,CAAC,CAAC,CAAE,CAAC,CAAC,CAAE,GAAG,CAAC,GAAG,IAAI,GAAG,IAAI,CAAC;QAC3D,CAAC;IACH,CAAC;IACD,IAAI,MAAM,GAAG,CAAC,CAAC;IACf,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC;QAC3B,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC;YAC3B,QAAQ,CAAC,CAAC,CAAE,CAAC,CAAC,CAAC,GAAG,IAAI,CAAC,CAAC,CAAE,GAAG,CAAC,CAAC,CAAC,CAAC,QAAQ,CAAC,CAAC,CAAE,CAAC,CAAC,CAAE,GAAG,IAAI,CAAC,CAAC,CAAE,CAAC,CAAC,CAAC,CAAC,CAAC;YACjE,MAAM,GAAG,IAAI,CAAC,GAAG,CAAC,MAAM,EAAE,QAAQ,CAAC,CAAC,CAAE,CAAC,CAAC,CAAE,CAAC,CAAC;QAC9C,CAAC;IACH,CAAC;IACD,MAAM,GAAG,GAAG,CAAC,IAAI,CAAC,YAAY,IAAI,IAAI,CAAC,GAAG,IAAI,CAAC,GAAG,CAAC,MAAM,EAAE,KAAK,CAAC,GAAG,KAAK,CAAC;IAC1E,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,CAAC,EAAE,CAAC,EAAE;QAAE,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,CAAC,EAAE,CAAC,EAAE;YAAE,QAAQ,CAAC,CAAC,CAAE,CAAC,CAAC,CAAC,GAAG,QAAQ,CAAC,CAAC,CAAE,CAAC,CAAC,CAAE,GAAG,GAAG,CAAC;IAEjG,MAAM,KAAK,GAAG,IAAI,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC,CAAC;IAC9C,6EAA6E;IAC7E,MAAM,QAAQ,GAAG,YAAY,CAAC,IAAI,CAAC,IAAI,EAAE,CAAC,CAAC,EAAE,EAAE,CAAC,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,KAAK,CAAC,GAAG,KAAK,CAAC,CAAC,CAAC;IACtF,OAAO,IAAI,eAAe,CAAC,QAAQ,EAAE,IAAI,EAAE,QAAQ,CAAC,CAAC;AACvD,CAAC"}

package/dist/ml/classical/robust_z.d.ts

@@ -0,0 +1,30 @@
+/**
+ * ml/classical/robust_z.ts — robust (median/MAD) z-scores for anomaly detection.
+ *
+ * Per feature: z = |x − median| / (1.4826 · MAD). Median and MAD are not dragged
+ * by the very outliers being looked for, unlike mean and standard deviation. When
+ * MAD is 0 (more than half the history identical) the mean absolute deviation
+ * stands in, then a floor, so a constant history still scores a change.
+ *
+ * Prediction: `scores[0] = max_i z_i / threshold` — ≥ 1 is anomalous.
+ */
+import type { MlPrediction, MlPredictor, MlSample } from "../types.js";
+export interface RobustZOptions {
+    /** z at which a feature counts as anomalous. Default 3.5 (Iglewicz & Hoaglin). */
+    threshold?: number;
+    /** Smallest scale per feature, as a fraction of |median| (plus 1e-9). Default 0.01. */
+    minScaleFraction?: number;
+}
+export declare class RobustZModel implements MlPredictor {
+    readonly center: Float64Array;
+    readonly scale: Float64Array;
+    readonly threshold: number;
+    constructor(center: Float64Array, scale: Float64Array, threshold: number);
+    /** Per-feature robust z-scores (absolute). */
+    zScores(x: ArrayLike<number>): number[];
+    predict(x: Float32Array): MlPrediction;
+}
+/** Confidence of an anomaly score normalised so 1 is the threshold: distance from it, capped. */
+export declare function anomalyConfidence(score: number): number;
+export declare function fitRobustZ(samples: readonly MlSample[], opts?: RobustZOptions): RobustZModel;
+//# sourceMappingURL=robust_z.d.ts.map

package/dist/ml/classical/robust_z.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"robust_z.d.ts","sourceRoot":"","sources":["../../../src/ml/classical/robust_z.ts"],"names":[],"mappings":"AAAA;;;;;;;;;GASG;AACH,OAAO,KAAK,EAAE,YAAY,EAAE,WAAW,EAAE,QAAQ,EAAE,MAAM,aAAa,CAAC;AAQvE,MAAM,WAAW,cAAc;IAC7B,kFAAkF;IAClF,SAAS,CAAC,EAAE,MAAM,CAAC;IACnB,uFAAuF;IACvF,gBAAgB,CAAC,EAAE,MAAM,CAAC;CAC3B;AAED,qBAAa,YAAa,YAAW,WAAW;IAE5C,QAAQ,CAAC,MAAM,EAAE,YAAY;IAC7B,QAAQ,CAAC,KAAK,EAAE,YAAY;IAC5B,QAAQ,CAAC,SAAS,EAAE,MAAM;gBAFjB,MAAM,EAAE,YAAY,EACpB,KAAK,EAAE,YAAY,EACnB,SAAS,EAAE,MAAM;IAG5B,8CAA8C;IAC9C,OAAO,CAAC,CAAC,EAAE,SAAS,CAAC,MAAM,CAAC,GAAG,MAAM,EAAE;IAIvC,OAAO,CAAC,CAAC,EAAE,YAAY,GAAG,YAAY;CAIvC;AAED,iGAAiG;AACjG,wBAAgB,iBAAiB,CAAC,KAAK,EAAE,MAAM,GAAG,MAAM,CAEvD;AAED,wBAAgB,UAAU,CAAC,OAAO,EAAE,SAAS,QAAQ,EAAE,EAAE,IAAI,GAAE,cAAmB,GAAG,YAAY,CAgBhG"}

package/dist/ml/classical/robust_z.js

@@ -0,0 +1,46 @@
+import { featureDim, median } from "./math.js";
+/** MAD → standard deviation for normally-distributed data. */
+const MAD_TO_SD = 1.4826;
+/** Mean absolute deviation → standard deviation for normally-distributed data. */
+const MEANAD_TO_SD = 1.2533;
+export class RobustZModel {
+    center;
+    scale;
+    threshold;
+    constructor(center, scale, threshold) {
+        this.center = center;
+        this.scale = scale;
+        this.threshold = threshold;
+    }
+    /** Per-feature robust z-scores (absolute). */
+    zScores(x) {
+        return Array.from(this.center, (c, i) => Math.abs((x[i] ?? 0) - c) / this.scale[i]);
+    }
+    predict(x) {
+        const score = Math.max(0, ...this.zScores(x)) / this.threshold;
+        return { scores: [score], confidence: anomalyConfidence(score) };
+    }
+}
+/** Confidence of an anomaly score normalised so 1 is the threshold: distance from it, capped. */
+export function anomalyConfidence(score) {
+    return Math.max(0, Math.min(1, Math.abs(score - 1)));
+}
+export function fitRobustZ(samples, opts = {}) {
+    const X = samples.map((s) => s.x);
+    const d = featureDim(X);
+    const minFrac = opts.minScaleFraction ?? 0.01;
+    const center = new Float64Array(d);
+    const scale = new Float64Array(d);
+    for (let i = 0; i < d; i++) {
+        const col = X.map((x) => x[i]);
+        const med = median(col);
+        const dev = col.map((v) => Math.abs(v - med));
+        let s = MAD_TO_SD * median(dev);
+        if (!(s > 0))
+            s = MEANAD_TO_SD * (dev.reduce((a, b) => a + b, 0) / dev.length);
+        center[i] = med;
+        scale[i] = Math.max(s, minFrac * Math.abs(med) + 1e-9);
+    }
+    return new RobustZModel(center, scale, opts.threshold ?? 3.5);
+}
+//# sourceMappingURL=robust_z.js.map

package/dist/ml/classical/robust_z.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"robust_z.js","sourceRoot":"","sources":["../../../src/ml/classical/robust_z.ts"],"names":[],"mappings":"AAWA,OAAO,EAAE,UAAU,EAAE,MAAM,EAAE,MAAM,WAAW,CAAC;AAE/C,8DAA8D;AAC9D,MAAM,SAAS,GAAG,MAAM,CAAC;AACzB,kFAAkF;AAClF,MAAM,YAAY,GAAG,MAAM,CAAC;AAS5B,MAAM,OAAO,YAAY;IAEZ;IACA;IACA;IAHX,YACW,MAAoB,EACpB,KAAmB,EACnB,SAAiB;QAFjB,WAAM,GAAN,MAAM,CAAc;QACpB,UAAK,GAAL,KAAK,CAAc;QACnB,cAAS,GAAT,SAAS,CAAQ;IACzB,CAAC;IAEJ,8CAA8C;IAC9C,OAAO,CAAC,CAAoB;QAC1B,OAAO,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,MAAM,EAAE,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,GAAG,CAAC,CAAC,GAAG,IAAI,CAAC,KAAK,CAAC,CAAC,CAAE,CAAC,CAAC;IACvF,CAAC;IAED,OAAO,CAAC,CAAe;QACrB,MAAM,KAAK,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,GAAG,IAAI,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC,GAAG,IAAI,CAAC,SAAS,CAAC;QAC/D,OAAO,EAAE,MAAM,EAAE,CAAC,KAAK,CAAC,EAAE,UAAU,EAAE,iBAAiB,CAAC,KAAK,CAAC,EAAE,CAAC;IACnE,CAAC;CACF;AAED,iGAAiG;AACjG,MAAM,UAAU,iBAAiB,CAAC,KAAa;IAC7C,OAAO,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,IAAI,CAAC,GAAG,CAAC,KAAK,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC;AACvD,CAAC;AAED,MAAM,UAAU,UAAU,CAAC,OAA4B,EAAE,OAAuB,EAAE;IAChF,MAAM,CAAC,GAAG,OAAO,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;IAClC,MAAM,CAAC,GAAG,UAAU,CAAC,CAAC,CAAC,CAAC;IACxB,MAAM,OAAO,GAAG,IAAI,CAAC,gBAAgB,IAAI,IAAI,CAAC;IAC9C,MAAM,MAAM,GAAG,IAAI,YAAY,CAAC,CAAC,CAAC,CAAC;IACnC,MAAM,KAAK,GAAG,IAAI,YAAY,CAAC,CAAC,CAAC,CAAC;IAClC,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC;QAC3B,MAAM,GAAG,GAAG,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,CAAC,CAAE,CAAC,CAAC;QAChC,MAAM,GAAG,GAAG,MAAM,CAAC,GAAG,CAAC,CAAC;QACxB,MAAM,GAAG,GAAG,GAAG,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,GAAG,GAAG,CAAC,CAAC,CAAC;QAC9C,IAAI,CAAC,GAAG,SAAS,GAAG,MAAM,CAAC,GAAG,CAAC,CAAC;QAChC,IAAI,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC;YAAE,CAAC,GAAG,YAAY,GAAG,CAAC,GAAG,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC,GAAG,GAAG,CAAC,MAAM,CAAC,CAAC;QAC/E,MAAM,CAAC,CAAC,CAAC,GAAG,GAAG,CAAC;QAChB,KAAK,CAAC,CAAC,CAAC,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,OAAO,GAAG,IAAI,CAAC,GAAG,CAAC,GAAG,CAAC,GAAG,IAAI,CAAC,CAAC;IACzD,CAAC;IACD,OAAO,IAAI,YAAY,CAAC,MAAM,EAAE,KAAK,EAAE,IAAI,CAAC,SAAS,IAAI,GAAG,CAAC,CAAC;AAChE,CAAC"}