public void TestLabelBinarizerSetLabelEncoding()
{
var lb = new LabelBinarizer<int>(negLabel: -2, posLabel: 2);
// two-class case
var inp = new[] {0, 1, 1, 0};
var expected = new double[,] {{-2, 2, 2, -2}}.ToDenseMatrix().Transpose();
var got = lb.Fit(inp).Transform(inp);
Assert.AreEqual(expected, got);
Assert.IsTrue(lb.InverseTransform(got).SequenceEqual(inp));
// multi-class case
inp = new[] {3, 2, 1, 2, 0};
expected = new double[,]
{
{-2, -2, -2, +2},
{-2, -2, +2, -2},
{-2, +2, -2, -2},
{-2, -2, +2, -2},
{+2, -2, -2, -2}
}.ToDenseMatrix();
got = lb.Fit(inp).Transform(inp);
Assert.AreEqual(expected, got);
Assert.IsTrue(lb.InverseTransform(got).SequenceEqual(inp));
}
public void TestLabelBinarizer()
{
var lb = new LabelBinarizer<string>();
// two-class case
var inp = new[] {"neg", "pos", "pos", "neg"};
var expected = new double[,] {{0, 1, 1, 0}}.ToDenseMatrix().Transpose();
var got = lb.Fit(inp).Transform(inp);
Assert.AreEqual(expected, got);
Assert.IsTrue(lb.InverseTransform(got).SequenceEqual(inp));
// multi-class case
inp = new[] {"spam", "ham", "eggs", "ham", "0"};
expected = new double[,]
{
{0, 0, 0, 1},
{0, 0, 1, 0},
{0, 1, 0, 0},
{0, 0, 1, 0},
{1, 0, 0, 0}
}.ToDenseMatrix();
got = lb.Fit(inp).Transform(inp);
Assert.AreEqual(expected, got);
Assert.IsTrue(lb.InverseTransform(got).SequenceEqual(inp));
}
public void TestLabelBinarizerSetLabelEncoding()
{
var lb = new LabelBinarizer <int>(negLabel: -2, posLabel: 2);
// two-class case
var inp = new[] { 0, 1, 1, 0 };
var expected = new double[, ] {
{ -2, 2, 2, -2 }
}.ToDenseMatrix().Transpose();
var got = lb.Fit(inp).Transform(inp);
Assert.AreEqual(expected, got);
Assert.IsTrue(lb.InverseTransform(got).SequenceEqual(inp));
// multi-class case
inp = new[] { 3, 2, 1, 2, 0 };
expected = new double[, ]
{
{ -2, -2, -2, +2 },
{ -2, -2, +2, -2 },
{ -2, +2, -2, -2 },
{ -2, -2, +2, -2 },
{ +2, -2, -2, -2 }
}.ToDenseMatrix();
got = lb.Fit(inp).Transform(inp);
Assert.AreEqual(expected, got);
Assert.IsTrue(lb.InverseTransform(got).SequenceEqual(inp));
}
public void TestLabelBinarizer()
{
var lb = new LabelBinarizer <string>();
// two-class case
var inp = new[] { "neg", "pos", "pos", "neg" };
var expected = new double[, ] {
{ 0, 1, 1, 0 }
}.ToDenseMatrix().Transpose();
var got = lb.Fit(inp).Transform(inp);
Assert.AreEqual(expected, got);
Assert.IsTrue(lb.InverseTransform(got).SequenceEqual(inp));
// multi-class case
inp = new[] { "spam", "ham", "eggs", "ham", "0" };
expected = new double[, ]
{
{ 0, 0, 0, 1 },
{ 0, 0, 1, 0 },
{ 0, 1, 0, 0 },
{ 0, 0, 1, 0 },
{ 1, 0, 0, 0 }
}.ToDenseMatrix();
got = lb.Fit(inp).Transform(inp);
Assert.AreEqual(expected, got);
Assert.IsTrue(lb.InverseTransform(got).SequenceEqual(inp));
}
/// <summary>
/// Fit Ridge regression model.
/// </summary>
/// <param name="x">[n_samples,n_features]. Training data</param>
/// <param name="y">Target values.</param>
/// <param name="sampleWeight">Sample weights.</param>
/// <returns>Instance of self.</returns>
public void Fit(Matrix <double> x, TLabel[] y, Vector <double> sampleWeight = null)
{
if (sampleWeight != null)
{
throw new ArgumentException("Sample weights are not supported by the classifier");
}
this.labelBinarizer = new LabelBinarizer <TLabel>(posLabel: 1, negLabel: -1);
Matrix <double> Y = this.labelBinarizer.Fit(y).Transform(y);
// second parameter is used only for ClassWeight.Auto, which we don't support here.
// So fake it.
Vector cw = this.ClassWeightEstimator.ComputeWeights(this.Classes, new int[0]);
//# get the class weight corresponding to each sample
Vector sampleWeightClasses = y.Select(v => cw[Array.BinarySearch(this.Classes, v)]).ToArray().ToDenseVector();
base.Fit(x, Y, sampleWeight: sampleWeightClasses);
}
public void TestLabelBinarizerErrors()
{
var oneClass = new[] {0, 0, 0, 0};
var lb = new LabelBinarizer<int>().Fit(oneClass);
lb = new LabelBinarizer<int>();
try
{
lb.Transform(new int[0]);
Assert.Fail();
}
catch (InvalidOperationException)
{
}
try
{
lb.InverseTransform(new DenseMatrix(0, 0));
Assert.Fail();
}
catch (ArgumentException)
{
}
try
{
new LabelBinarizer<int>(negLabel: 2, posLabel: 1);
Assert.Fail();
}
catch (ArgumentException)
{
}
try
{
new LabelBinarizer<int>(negLabel: 2, posLabel: 2);
Assert.Fail();
}
catch (ArgumentException)
{
}
}
public void TestLabelBinarizerErrors()
{
var oneClass = new[] { 0, 0, 0, 0 };
var lb = new LabelBinarizer <int>().Fit(oneClass);
lb = new LabelBinarizer <int>();
try
{
lb.Transform(new int[0]);
Assert.Fail();
}
catch (InvalidOperationException)
{
}
try
{
lb.InverseTransform(new DenseMatrix(0, 0));
Assert.Fail();
}
catch (ArgumentException)
{
}
try
{
new LabelBinarizer <int>(negLabel: 2, posLabel: 1);
Assert.Fail();
}
catch (ArgumentException)
{
}
try
{
new LabelBinarizer <int>(negLabel: 2, posLabel: 2);
Assert.Fail();
}
catch (ArgumentException)
{
}
}
