DataOperationsCatalog.TrainTestSplit Method

Definition

Namespace:

Microsoft.ML

Assembly:

Microsoft.ML.Data.dll

Package:

Microsoft.ML v3.0.1

Split the dataset into the train set and test set according to the given fraction. Respects the samplingKeyColumnName if provided.

public Microsoft.ML.DataOperationsCatalog.TrainTestData TrainTestSplit (Microsoft.ML.IDataView data, double testFraction = 0.1, string samplingKeyColumnName = default, int? seed = default);

member this.TrainTestSplit : Microsoft.ML.IDataView * double * string * Nullable<int> -> Microsoft.ML.DataOperationsCatalog.TrainTestData

Public Function TrainTestSplit (data As IDataView, Optional testFraction As Double = 0.1, Optional samplingKeyColumnName As String = Nothing, Optional seed As Nullable(Of Integer) = Nothing) As DataOperationsCatalog.TrainTestData

Parameters

data

IDataView

The dataset to split.

testFraction

Double

The fraction of data to go into the test set.

samplingKeyColumnName

String

Name of a column to use for grouping rows. If two examples share the same value of the samplingKeyColumnName, they are guaranteed to appear in the same subset (train or test). This can be used to ensure no label leakage from the train to the test set. Note that when performing a Ranking Experiment, the samplingKeyColumnName must be the GroupId column. If null no row grouping will be performed.

seed

Nullable<Int32>

Seed for the random number generator used to select rows for the train-test split.

Returns

DataOperationsCatalog.TrainTestData

Examples

using System;
using System.Collections.Generic;
using Microsoft.ML;

namespace Samples.Dynamic
{
    /// <summary>
    /// Sample class showing how to use TrainTestSplit.
    /// </summary>
    public static class TrainTestSplit
    {
        public static void Example()
        {
            // Creating the ML.Net IHostEnvironment object, needed for the pipeline.
            var mlContext = new MLContext();

            // Generate some data points.
            var examples = GenerateRandomDataPoints(10);

            // Convert the examples list to an IDataView object, which is consumable
            // by ML.NET API.
            var dataview = mlContext.Data.LoadFromEnumerable(examples);

            // Leave out 10% of the dataset for testing.For some types of problems,
            // for example for ranking or anomaly detection, we must ensure that the
            // split leaves the rows with the same value in a particular column, in
            // one of the splits. So below, we specify Group column as the column
            // containing the sampling keys. Notice how keeping the rows with the
            // same value in the Group column overrides the testFraction definition. 
            var split = mlContext.Data
                .TrainTestSplit(dataview, testFraction: 0.1,
                samplingKeyColumnName: "Group");

            var trainSet = mlContext.Data
                .CreateEnumerable<DataPoint>(split.TrainSet, reuseRowObject: false);

            var testSet = mlContext.Data
                .CreateEnumerable<DataPoint>(split.TestSet, reuseRowObject: false);

            PrintPreviewRows(trainSet, testSet);

            //  The data in the Train split.
            //  [Group, 1], [Features, 0.8173254]
            //  [Group, 1], [Features, 0.5581612]
            //  [Group, 1], [Features, 0.5588848]
            //  [Group, 1], [Features, 0.4421779]
            //  [Group, 1], [Features, 0.2737045]

            //  The data in the Test split.
            //  [Group, 0], [Features, 0.7262433]
            //  [Group, 0], [Features, 0.7680227]
            //  [Group, 0], [Features, 0.2060332]
            //  [Group, 0], [Features, 0.9060271]
            //  [Group, 0], [Features, 0.9775497]

            // Example of a split without specifying a sampling key column.
            split = mlContext.Data.TrainTestSplit(dataview, testFraction: 0.2);
            trainSet = mlContext.Data
                .CreateEnumerable<DataPoint>(split.TrainSet, reuseRowObject: false);

            testSet = mlContext.Data
                .CreateEnumerable<DataPoint>(split.TestSet, reuseRowObject: false);

            PrintPreviewRows(trainSet, testSet);

            // The data in the Train split.
            // [Group, 0], [Features, 0.7262433]
            // [Group, 1], [Features, 0.8173254]
            // [Group, 0], [Features, 0.7680227]
            // [Group, 1], [Features, 0.5581612]
            // [Group, 0], [Features, 0.2060332]
            // [Group, 1], [Features, 0.4421779]
            // [Group, 0], [Features, 0.9775497]
            // [Group, 1], [Features, 0.2737045]

            // The data in the Test split.
            // [Group, 1], [Features, 0.5588848]
            // [Group, 0], [Features, 0.9060271]

        }

        private static IEnumerable<DataPoint> GenerateRandomDataPoints(int count,
            int seed = 0)

        {
            var random = new Random(seed);
            for (int i = 0; i < count; i++)
            {
                yield return new DataPoint
                {
                    Group = i % 2,

                    // Create random features that are correlated with label.
                    Features = (float)random.NextDouble()
                };
            }
        }

        // Example with label and group column. A data set is a collection of such
        // examples.
        private class DataPoint
        {
            public float Group { get; set; }

            public float Features { get; set; }
        }

        // print helper
        private static void PrintPreviewRows(IEnumerable<DataPoint> trainSet,
            IEnumerable<DataPoint> testSet)

        {

            Console.WriteLine($"The data in the Train split.");
            foreach (var row in trainSet)
                Console.WriteLine($"{row.Group}, {row.Features}");

            Console.WriteLine($"\nThe data in the Test split.");
            foreach (var row in testSet)
                Console.WriteLine($"{row.Group}, {row.Features}");
        }
    }
}