DataSplitResult<T>

Models & Types · AiDotNet.Preprocessing.DataPreparation

Contains the results of a data split operation for Matrix/Vector data.

For Beginners

After splitting your data, this class holds all the pieces:

  • Training data (XTrain, yTrain): What your model learns from
  • Test data (XTest, yTest): What you evaluate your model on
  • Validation data (XValidation, yValidation): Optional, for hyperparameter tuning
  • Indices: Which rows from the original data ended up in each set

How It Works

Why Track Indices? Knowing which original samples are in each set is useful for:

  • Debugging: “Why did my model get sample #42 wrong?”
  • Analysis: Comparing predictions to original data
  • Reproducibility: Recording exactly how data was split

Constructors

Constructor Summary
DataSplitResult()

Properties

Property Summary
Matrix<T> XTest { get; set; } Gets the feature matrix for testing.
Matrix<T> XTrain { get; set; } Gets the feature matrix for training.
Matrix<T> XValidation { get; set; } Gets the feature matrix for validation (optional three-way split).
Vector<T> yTest { get; set; } Gets the target vector for testing (null for unsupervised learning).
Vector<T> yTrain { get; set; } Gets the target vector for training (null for unsupervised learning).
Vector<T> yValidation { get; set; } Gets the target vector for validation (optional three-way split).
bool HasValidationSet { get; } Gets whether this result includes a validation set.
int? FoldIndex { get; set; } Gets the fold index for cross-validation methods (0-based).
int? RepeatIndex { get; set; } Gets the repeat index for repeated cross-validation methods (0-based).
int? TotalFolds { get; set; } Gets the total number of folds for cross-validation methods.
int? TotalRepeats { get; set; } Gets the total number of repeats for repeated cross-validation methods.
int[] TestIndices { get; set; } Gets the indices of rows from the original data that are in the test set.
int[] TrainIndices { get; set; } Gets the indices of rows from the original data that are in the training set.
int[] ValidationIndices { get; set; } Gets the indices of rows from the original data that are in the validation set (optional).

Methods

Method Summary
string GetSummary() Gets a human-readable summary of this split result.