Add algorithm ReLU_function (#569)

Author: ThanhDungCR7

Algorithms.Tests/Numeric/ReluTest.cs

@@ -0,0 +1,88 @@
+using Algorithms.Numeric;
+using NUnit.Framework;
+using System;
+
+namespace Algorithms.Tests.Numeric;
+
+[TestFixture]
+public static class ReluTests
+{
+    // Tolerance for floating-point comparisons
+    private const double Tolerance = 1e-9;
+
+    // --- SCALAR TESTS (Relu.Compute(double)) ---
+
+    [TestCase(0.0, 0.0)]
+    [TestCase(1.0, 1.0)]
+    [TestCase(-1.0, 0.0)]
+    [TestCase(5.0, 5.0)]
+    [TestCase(-5.0, 0.0)]
+    public static void ReluFunction_Scalar_ReturnsCorrectValue(double input, double expected)
+    {
+        var result = Relu.Compute(input);
+        Assert.That(result, Is.EqualTo(expected).Within(Tolerance));
+    }
+
+    [Test]
+    public static void ReluFunction_Scalar_HandlesLimitsAndNaN()
+    {
+        // Positive infinity stays +Infinity, negative infinity becomes 0, NaN propagates
+        Assert.That(RelUComputePositiveInfinity(), Is.EqualTo(double.PositiveInfinity));
+        Assert.That(RelUComputeNegativeInfinity(), Is.EqualTo(0.0).Within(Tolerance));
+        Assert.That(RelUComputeNaN(), Is.NaN);
+
+        static double RelUComputePositiveInfinity() => Relu.Compute(double.PositiveInfinity);
+        static double RelUComputeNegativeInfinity() => Relu.Compute(double.NegativeInfinity);
+        static double RelUComputeNaN() => Relu.Compute(double.NaN);
+    }
+
+    [TestCase(100.0)]
+    [TestCase(0.0001)]
+    [TestCase(-100.0)]
+    public static void ReluFunction_Scalar_ResultIsNonNegative(double input)
+    {
+        var result = Relu.Compute(input);
+        Assert.That(result, Is.GreaterThanOrEqualTo(0.0));
+    }
+
+    // --- VECTOR TESTS (Relu.Compute(double[])) ---
+
+    [Test]
+    public static void ReluFunction_Vector_ReturnsCorrectValues()
+    {
+        var input = new[] { 0.0, 1.0, -2.0 };
+        var expected = new[] { 0.0, 1.0, 0.0 };
+
+        var result = Relu.Compute(input);
+
+        Assert.That(result, Is.EqualTo(expected).Within(Tolerance));
+    }
+
+    [Test]
+    public static void ReluFunction_Vector_HandlesLimitsAndNaN()
+    {
+        var input = new[] { double.PositiveInfinity, 0.0, double.NaN };
+        var result = Relu.Compute(input);
+
+        Assert.That(result.Length, Is.EqualTo(input.Length));
+        Assert.That(result[0], Is.EqualTo(double.PositiveInfinity));
+        Assert.That(result[1], Is.EqualTo(0.0).Within(Tolerance));
+        Assert.That(result[2], Is.NaN);
+    }
+
+    // --- EXCEPTION TESTS ---
+
+    [Test]
+    public static void ReluFunction_Vector_ThrowsOnNullInput()
+    {
+        double[]? input = null;
+        Assert.Throws<ArgumentNullException>(() => Relu.Compute(input!));
+    }
+
+    [Test]
+    public static void ReluFunction_Vector_ThrowsOnEmptyInput()
+    {
+        var input = Array.Empty<double>();
+        Assert.Throws<ArgumentException>(() => Relu.Compute(input));
+    }
+}

Algorithms/Numeric/Relu.cs

@@ -0,0 +1,46 @@
+namespace Algorithms.Numeric;
+
+/// <summary>
+///     Implementation of the Rectified Linear Unit (ReLU) function.
+///     ReLU is defined as: ReLU(x) = max(0, x).
+///     It is commonly used as an activation function in neural networks.
+/// </summary>
+public static class Relu
+{
+    /// <summary>
+    ///     Compute the Rectified Linear Unit (ReLU) for a single value.
+    /// </summary>
+    /// <param name="input">The input real number.</param>
+    /// <returns>The output real number (>= 0).</returns>
+    public static double Compute(double input)
+    {
+        return Math.Max(0.0, input);
+    }
+
+    /// <summary>
+    ///     Compute the Rectified Linear Unit (ReLU) element-wise for a vector.
+    /// </summary>
+    /// <param name="input">The input vector of real numbers.</param>
+    /// <returns>The output vector where each element is max(0, input[i]).</returns>
+    public static double[] Compute(double[] input)
+    {
+        if (input is null)
+        {
+            throw new ArgumentNullException(nameof(input));
+        }
+
+        if (input.Length == 0)
+        {
+            throw new ArgumentException("Array is empty.");
+        }
+
+        var output = new double[input.Length];
+
+        for (var i = 0; i < input.Length; i++)
+        {
+            output[i] = Math.Max(0.0, input[i]);
+        }
+
+        return output;
+    }
+}