TheAlgorithms/C-Sharp · error · ArgumentException

Input features cannot be empty.

Error message

Input features cannot be empty.

What it means

LogisticRegression.Fit validates its training input and throws ArgumentException when the feature matrix x is null-length or has zero columns. Training on an empty matrix would produce meaningless weights (nFeatures = x[0].Length would also fail), so the method rejects it up front.

Solutions

1. Ensure the training set contains at least one sample with at least one feature before calling Fit.
2. Guard the caller: skip training or surface a domain-specific message when the dataset is empty.
3. Fix the upstream data-loading/filtering step that produced an empty matrix.

Example fix

// before
model.Fit(new double[0][], new int[0]); // throws

// after
if (x.Length > 0 && x[0].Length > 0)
{
    model.Fit(x, y);
}

if (x == null || x.Length == 0 || x[0].Length == 0)
{
    throw new ArgumentException("Training features must contain at least one sample with one feature.");
}
model.Fit(x, y);

static bool HasSamples(double[][] x) => x != null && x.Length > 0 && x[0] != null && x[0].Length > 0;

try
{
    model.Fit(x, y);
}
catch (ArgumentException ex) when (ex.Message == "Input features cannot be empty.")
{
    logger.LogWarning("Skipping training: empty dataset.");
}

When it happens

Trigger: Calling Fit(double[][] x, int[] y, ...) with an empty array (x.Length == 0) or with rows of zero length (x[0].Length == 0).

Common situations: Upstream data pipeline returned no rows (empty CSV, filtered-out dataset); a reshape/split step produced an empty feature matrix; a deserialization bug yielding an empty array.

Understand the failure class

Background: "must not be empty", "cannot be empty" — required-field validation errors across open-source libraries — this error's family across 41 libraries.

Related errors

• k must be at least 1.
• Number of samples and labels must match.
• Feature count mismatch.
• message
• key must be non-empty string

AI-assisted analysis of TheAlgorithms/C-Sharp@96e2905cab (2026-09-13). Data as JSON: /api/errors/dd782173da7e8225. Report an issue: GitHub.

Appendix: source

Thrown at Algorithms/MachineLearning/LogisticRegression.cs:27

public class LogisticRegression
{
    private double[] weights = [];
    private double bias;

    public int FeatureCount => weights.Length;

    /// <summary>
    /// Fit the model using gradient descent.
    /// </summary>
    /// <param name="x">2D array of features (samples x features).</param>
    /// <param name="y">Array of labels (0 or 1).</param>
    /// <param name="epochs">Number of iterations.</param>
    /// <param name="learningRate">Step size.</param>
    public void Fit(double[][] x, int[] y, int epochs = 1000, double learningRate = 0.01)
    {
        if (x.Length == 0 || x[0].Length == 0)
        {
            throw new ArgumentException("Input features cannot be empty.");
        }

        if (x.Length != y.Length)
        {
            throw new ArgumentException("Number of samples and labels must match.");
        }

        int nSamples = x.Length;
        int nFeatures = x[0].Length;
        weights = new double[nFeatures];
        bias = 0;

        for (int epoch = 0; epoch < epochs; epoch++)
        {
            double[] dw = new double[nFeatures];
            double db = 0;
            for (int i = 0; i < nSamples; i++)
            {

View on GitHub (pinned to 96e2905cab)