void TestSelectWorkout()
{
var data = new[] { new TestClass()
{
A = 1, B = 2, C = 3,
}, new TestClass()
{
A = 4, B = 5, C = 6
} };
var invalidData = new [] { new TestClass2 {
D = 3, E = 5
} };
var dataView = ComponentCreation.CreateDataView(Env, data);
var invalidDataView = ComponentCreation.CreateDataView(Env, invalidData);
// Workout on keep columns
var est = new ColumnSelectingEstimator(Env, new[] { "A", "B" }, null, true, false);
TestEstimatorCore(est, validFitInput: dataView, invalidInput: invalidDataView);
// Workout on drop columns
est = new ColumnSelectingEstimator(Env, null, new[] { "A", "B" }, true, false);
TestEstimatorCore(est, validFitInput: dataView, invalidInput: invalidDataView);
// Workout on keep columns with ignore mismatch -- using invalid data set
est = new ColumnSelectingEstimator(Env, new[] { "A", "B" }, null, true, true);
TestEstimatorCore(est, validFitInput: invalidDataView);
}
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 TestSelectDrop()
{
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.DropColumns(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.False(foundColumnA);
Assert.True(foundColumnB);
Assert.Equal(0, bIdx);
Assert.False(foundColumnC);
}
void TestSelectDrop()
{
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 = new ColumnSelectingEstimator(Env, null, new string[] { "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.False(foundColumnA);
Assert.Equal(0, aIdx);
Assert.True(foundColumnB);
Assert.Equal(0, bIdx);
Assert.False(foundColumnC);
Assert.Equal(0, cIdx);
}
/// <summary> /// DropColumns is used to select a list of columns that user wants to drop from a given input. Any column not specified will /// be maintained in the output schema. /// </summary> /// <param name="catalog">The transform's catalog.</param> /// <param name="columnsToDrop">The array of column names to drop.</param> public static ColumnSelectingEstimator DropColumns(this TransformsCatalog catalog, params string[] columnsToDrop) => ColumnSelectingEstimator.DropColumns(CatalogUtils.GetEnvironment(catalog), columnsToDrop);
