TheAlgorithms/C-Sharp · error · ArgumentNullException

input

Error message

input

What it means

Tanh.Compute applies the hyperbolic tangent element-wise and requires a non-null input array. A null reference cannot be iterated, so the method throws ArgumentNullException naming the 'input' parameter instead of failing with a NullReferenceException.

Solutions

1. Initialize the array before calling (e.g. Array.Empty<double>() instead of null)
2. Add a null check at the call site and skip/return a default result when null
3. Fix the producer that returns null to return an empty array instead

Example fix

// before
double[] activations = null;
var out = Tanh.Compute(activations); // throws
// after
var out = Tanh.Compute(activations ?? Array.Empty<double>());

if (input is null) throw new ArgumentNullException(nameof(input));

static bool IsNotNull<T>(T[]? a) => a is not null;

try { return Tanh.Compute(input); }
catch (ArgumentNullException ex) when (ex.ParamName == "input") { /* initialize or skip */ }

When it happens

Trigger: Calling Tanh.Compute(null), typically when an array field/property was never initialized or a method returned null instead of an empty array.

Common situations: Uninitialized class fields for activations; a lookup that returned null in older API versions; optional configuration vectors left unset.

Related errors

• Error value is not on interval (0.0; 1.0).
• Array is empty.
• Input must be a non-negative integer.
• Array is empty.
• Invalid parameter settings for Ascon Hash

AI-assisted analysis of TheAlgorithms/C-Sharp@96e2905cab (2026-09-13). Data as JSON: /api/errors/3447e62864194f91. Report an issue: GitHub.

Appendix: source

Thrown at Algorithms/Numeric/Tanh.cs:33

    /// </summary>
    /// <param name="input">The input real number.</param>
    /// <returns>The output real number in the range [-1, 1].</returns>
    public static double Compute(double input)
    {
        // For a single double, we can directly use the optimized Math.Tanh method.
        return Math.Tanh(input);
    }

    /// <summary>
    ///     Compute the Hyperbolic Tangent (Tanh) function element-wise for a vector.
    /// </summary>
    /// <param name="input">The input vector of real numbers.</param>
    /// <returns>The output vector of real numbers, where each element is in the range [-1, 1].</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 outputVector = new double[input.Length];

        for (var index = 0; index < input.Length; index++)
        {
            // Apply Tanh to each element using the optimized Math.Tanh method.
            outputVector[index] = Math.Tanh(input[index]);
        }

        return outputVector;
    }
}

View on GitHub (pinned to 96e2905cab)