@danielsimonjr/mathts-autograd 0.3.1 → 0.3.3

package/dist/index.d.ts

@@ -58,6 +58,14 @@ declare class DualTensor {
      */
     divide(other: DualTensor): DualTensor;
     private _unaryElementwise;
+    /**
+     * Apply a shared elementary rule from core's `DUAL_UNARY_RULES` element-wise.
+     * Sourcing `primal`/`deriv` from the canonical table (rather than inline
+     * `Math.*`) keeps this AD path and the scalar `Dual` in lock-step and lets a
+     * future accelerated `primal` in the standard layer propagate here for free —
+     * the loop still calls plain scalar functions, so it stays allocation-free.
+     */
+    private _rule;
     /** Exponential. f'(x) = eˣ = y. */
     exp(): DualTensor;
     /** Natural logarithm. f'(x) = 1/x. */

package/dist/index.js

@@ -1,5 +1,6 @@
 // src/dual-tensor.ts
 import { Tensor } from "@danielsimonjr/mathts-tensor";
+import { DUAL_UNARY_RULES } from "@danielsimonjr/mathts-core";
 var DualTensor = class _DualTensor {
   shape;
   primal;
@@ -138,47 +139,51 @@ var DualTensor = class _DualTensor {
     }
     return new _DualTensor(this.shape, p, t);
   }
+  /**
+   * Apply a shared elementary rule from core's `DUAL_UNARY_RULES` element-wise.
+   * Sourcing `primal`/`deriv` from the canonical table (rather than inline
+   * `Math.*`) keeps this AD path and the scalar `Dual` in lock-step and lets a
+   * future accelerated `primal` in the standard layer propagate here for free —
+   * the loop still calls plain scalar functions, so it stays allocation-free.
+   */
+  _rule(rule) {
+    return this._unaryElementwise(rule.primal, rule.deriv);
+  }
   /** Exponential. f'(x) = eˣ = y. */
   exp() {
-    return this._unaryElementwise(Math.exp, (_x, y) => y);
+    return this._rule(DUAL_UNARY_RULES.exp);
   }
   /** Natural logarithm. f'(x) = 1/x. */
   log() {
-    return this._unaryElementwise(Math.log, (x) => 1 / x);
+    return this._rule(DUAL_UNARY_RULES.log);
   }
   /** Sine. f'(x) = cos(x). */
   sin() {
-    return this._unaryElementwise(Math.sin, (x) => Math.cos(x));
+    return this._rule(DUAL_UNARY_RULES.sin);
   }
   /** Cosine. f'(x) = −sin(x). */
   cos() {
-    return this._unaryElementwise(Math.cos, (x) => -Math.sin(x));
+    return this._rule(DUAL_UNARY_RULES.cos);
   }
   /** Tangent. f'(x) = 1/cos²(x). */
   tan() {
-    return this._unaryElementwise(Math.tan, (x) => 1 / (Math.cos(x) * Math.cos(x)));
+    return this._rule(DUAL_UNARY_RULES.tan);
   }
   /** Square root. f'(x) = 1/(2·y) where y = √x. */
   sqrt() {
-    return this._unaryElementwise(Math.sqrt, (_x, y) => 1 / (2 * y));
+    return this._rule(DUAL_UNARY_RULES.sqrt);
   }
   /** Square (x²). f'(x) = 2x. */
   square() {
-    return this._unaryElementwise(
-      (x) => x * x,
-      (x) => 2 * x
-    );
+    return this._rule(DUAL_UNARY_RULES.square);
   }
   /** Reciprocal (1/x). f'(x) = −1/x². */
   reciprocal() {
-    return this._unaryElementwise(
-      (x) => 1 / x,
-      (x) => -1 / (x * x)
-    );
+    return this._rule(DUAL_UNARY_RULES.reciprocal);
   }
   /** Absolute value. f'(x) = sign(x) (subgradient 0 at exact zero). */
   abs() {
-    return this._unaryElementwise(Math.abs, (x) => Math.sign(x));
+    return this._rule(DUAL_UNARY_RULES.abs);
   }
   /** Fixed-exponent power x^k. f'(x) = k·x^(k−1). */
   pow(k) {
@@ -189,63 +194,63 @@ var DualTensor = class _DualTensor {
   }
   /** Hyperbolic sine. f'(x) = cosh(x). */
   sinh() {
-    return this._unaryElementwise(Math.sinh, (x) => Math.cosh(x));
+    return this._rule(DUAL_UNARY_RULES.sinh);
   }
   /** Hyperbolic cosine. f'(x) = sinh(x). */
   cosh() {
-    return this._unaryElementwise(Math.cosh, (x) => Math.sinh(x));
+    return this._rule(DUAL_UNARY_RULES.cosh);
   }
   /** Hyperbolic tangent. f'(x) = 1 − tanh²(x) = 1 − y². */
   tanh() {
-    return this._unaryElementwise(Math.tanh, (_x, y) => 1 - y * y);
+    return this._rule(DUAL_UNARY_RULES.tanh);
   }
   /** Arcsine. f'(x) = 1/√(1 − x²). */
   asin() {
-    return this._unaryElementwise(Math.asin, (x) => 1 / Math.sqrt(1 - x * x));
+    return this._rule(DUAL_UNARY_RULES.asin);
   }
   /** Arccosine. f'(x) = −1/√(1 − x²). */
   acos() {
-    return this._unaryElementwise(Math.acos, (x) => -1 / Math.sqrt(1 - x * x));
+    return this._rule(DUAL_UNARY_RULES.acos);
   }
   /** Arctangent. f'(x) = 1/(1 + x²). */
   atan() {
-    return this._unaryElementwise(Math.atan, (x) => 1 / (1 + x * x));
+    return this._rule(DUAL_UNARY_RULES.atan);
   }
   /** Inverse hyperbolic sine. f'(x) = 1/√(x² + 1). */
   asinh() {
-    return this._unaryElementwise(Math.asinh, (x) => 1 / Math.sqrt(x * x + 1));
+    return this._rule(DUAL_UNARY_RULES.asinh);
   }
   /** Inverse hyperbolic cosine (x ≥ 1). f'(x) = 1/√(x² − 1). */
   acosh() {
-    return this._unaryElementwise(Math.acosh, (x) => 1 / Math.sqrt(x * x - 1));
+    return this._rule(DUAL_UNARY_RULES.acosh);
   }
   /** Inverse hyperbolic tangent (|x| < 1). f'(x) = 1/(1 − x²). */
   atanh() {
-    return this._unaryElementwise(Math.atanh, (x) => 1 / (1 - x * x));
+    return this._rule(DUAL_UNARY_RULES.atanh);
   }
   /** Base-2 logarithm. f'(x) = 1/(x·ln2). */
   log2() {
-    return this._unaryElementwise(Math.log2, (x) => 1 / (x * Math.LN2));
+    return this._rule(DUAL_UNARY_RULES.log2);
   }
   /** Base-10 logarithm. f'(x) = 1/(x·ln10). */
   log10() {
-    return this._unaryElementwise(Math.log10, (x) => 1 / (x * Math.LN10));
+    return this._rule(DUAL_UNARY_RULES.log10);
   }
   /** log(1 + x). f'(x) = 1/(1 + x). */
   log1p() {
-    return this._unaryElementwise(Math.log1p, (x) => 1 / (1 + x));
+    return this._rule(DUAL_UNARY_RULES.log1p);
   }
   /** exp(x) − 1. f'(x) = eˣ = y + 1. */
   expm1() {
-    return this._unaryElementwise(Math.expm1, (_x, y) => y + 1);
+    return this._rule(DUAL_UNARY_RULES.expm1);
   }
   /** Cube root. f'(x) = 1/(3·y²) where y = x^(1/3) (→∞ at x = 0). */
   cbrt() {
-    return this._unaryElementwise(Math.cbrt, (_x, y) => 1 / (3 * y * y));
+    return this._rule(DUAL_UNARY_RULES.cbrt);
   }
   /** Sign (−1/0/+1). f'(x) = 0 almost everywhere. */
   sign() {
-    return this._unaryElementwise(Math.sign, () => 0);
+    return this._rule(DUAL_UNARY_RULES.sign);
   }
   /**
    * Two-argument arctangent: atan2(this, other), this = y, other = x.

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@danielsimonjr/mathts-autograd",
-  "version": "0.3.1",
+  "version": "0.3.3",
   "description": "Forward + reverse-mode automatic differentiation for MathTS rank-N Tensor",
   "author": "Daniel Simon Jr.",
   "license": "MIT",
@@ -31,8 +31,8 @@
     "build:prod": "tsup src/index.ts --format esm --dts --clean --minify --treeshake"
   },
   "dependencies": {
-    "@danielsimonjr/mathts-core": "^0.2.0",
-    "@danielsimonjr/mathts-tensor": "^0.2.1"
+    "@danielsimonjr/mathts-core": "^0.4.0",
+    "@danielsimonjr/mathts-tensor": "^0.2.4"
   },
   "devDependencies": {
     "@types/node": "^25.5.2",