void TestSelectKeepWithOrder()
{
var data = new[] { new TestClass()
{
A = 1, B = 2, C = 3,
}, new TestClass()
{
A = 4, B = 5, C = 6
} };
var dataView = ComponentCreation.CreateDataView(Env, data);
// Expected output will be CA
var est = ColumnSelectingEstimator.KeepColumns(Env, "C", "A");
var transformer = est.Fit(dataView);
var result = transformer.Transform(dataView);
var foundColumnA = result.Schema.TryGetColumnIndex("A", out int aIdx);
var foundColumnB = result.Schema.TryGetColumnIndex("B", out int bIdx);
var foundColumnC = result.Schema.TryGetColumnIndex("C", out int cIdx);
Assert.True(foundColumnA);
Assert.Equal(1, aIdx);
Assert.False(foundColumnB);
Assert.True(foundColumnC);
Assert.Equal(0, cIdx);
}
void TestSelectColumnsWithMissing()
{
var data = new[] { new TestClass()
{
A = 1, B = 2, C = 3,
}, new TestClass()
{
A = 4, B = 5, C = 6
} };
var dataView = ComponentCreation.CreateDataView(Env, data);
var est = ColumnSelectingEstimator.KeepColumns(Env, "D", "G");
Assert.Throws <ArgumentOutOfRangeException>(() => est.Fit(dataView));
}
public void TestSelectColumnsWithMissing()
{
var data = new[] { new TestClass()
{
A = 1, B = 2, C = 3,
}, new TestClass()
{
A = 4, B = 5, C = 6
} };
var dataView = ML.Data.LoadFromEnumerable(data);
var est = ColumnSelectingEstimator.KeepColumns(Env, "D", "G");
Assert.Throws <ArgumentOutOfRangeException>(() => est.Fit(dataView));
}
void TestSelectKeep()
{
var data = new[] { new TestClass()
{
A = 1, B = 2, C = 3,
}, new TestClass()
{
A = 4, B = 5, C = 6
} };
var dataView = ML.Data.LoadFromEnumerable(data);
var est = ColumnSelectingEstimator.KeepColumns(Env, "A", "C");
var transformer = est.Fit(dataView);
var result = transformer.Transform(dataView);
var foundColumnA = result.Schema.TryGetColumnIndex("A", out int aIdx);
var foundColumnB = result.Schema.TryGetColumnIndex("B", out int bIdx);
var foundColumnC = result.Schema.TryGetColumnIndex("C", out int cIdx);
Assert.True(foundColumnA);
Assert.Equal(0, aIdx);
Assert.False(foundColumnB);
Assert.True(foundColumnC);
Assert.Equal(1, cIdx);
}
/// <summary> /// KeepColumns is used to select a list of columns that the user wants to keep on a given an input. Any column not specified /// will be dropped from the output output schema. /// </summary> /// <param name="catalog">The transform's catalog.</param> /// <param name="columnsToKeep">The array of column names to keep.</param> public static ColumnSelectingEstimator KeepColumns(this TransformsCatalog catalog, params string[] columnsToKeep) => ColumnSelectingEstimator.KeepColumns(CatalogUtils.GetEnvironment(catalog), columnsToKeep);