public void VStackNDArrays()
{
//1D
var np = new NumPyGeneric <double>();
var n1 = np.array(new double[] { 1, 2, 3 });
var n2 = np.array(new double[] { 2, 3, 4 });
var n = np.vstack(n1, n2);
Assert.IsTrue(n.Size == (n1.Size + n2.Size));
Assert.IsTrue(n[0, 0] == 1);
Assert.IsTrue(n[1, 0] == 2);
Assert.IsTrue(n[1, 2] == 4);
//2D
n1 = np.array(new double[][] { new double[] { 1 }, new double[] { 2 }, new double[] { 3 } });
n2 = np.array(new double[][] { new double[] { 4 }, new double[] { 5 }, new double[] { 6 } });
n = np.vstack(n1, n2);
Assert.IsTrue(n.Size == (n1.Size + n2.Size));
Assert.IsTrue(n[0, 0] == 1);
Assert.IsTrue(n[1, 0] == 2);
Assert.IsTrue(n[2, 0] == 3);
Assert.IsTrue(n[3, 0] == 4);
Assert.IsTrue(n[4, 0] == 5);
Assert.IsTrue(n[5, 0] == 6);
}
public void LinearRegression()
{
var np = new NumPyGeneric <double>();
// the time array
var time = np.linspace(0, 10, 1000);
// the values over t - linear dependence with noice
var values = time * 3 + 5 + np.random.randn(time.Data.Length);
var A = np.vstack(np.ones(new Shape(1000)), time);
var A_T = A.transpose();
var A_T_A = A_T.dot(A);
var A_PseudoInv = A.transpose().dot(A).inv();
var param = A_PseudoInv.dot(A.transpose());
//param = param.dot(param);
}
