public static double[] NormalizeZScore(this double[] data)
{
var d = new DenseVector(data);
var result = new DenseVector(d.Count);
d.CopyTo(result);
result = (DenseVector) ((result - d.Mean())/(d.StandardDeviation()));
return result.ToArray();
}
public void VarianceRatioTest(double[] data, int lag = 1, TimeSeriesType cor = TimeSeriesType.HOM)
{
DenseVector x = new DenseVector(data);
//Mean
double mu = x.Mean();
//Variance for 1st Order Difference
double s1 = (x.SubVector(lag, x.Count - lag) - x.SubVector(0, x.Count - 2)).Variance();
double dLag = lag;
double varvrt = double.NaN;
switch (cor)
{
case TimeSeriesType.HOM:
{
varvrt = 2*(2*dLag - 1)*(dLag - 1)/(3*dLag*x.Count);
break;
}
case TimeSeriesType.HET:
{
varvrt = 0;
double sum2 = 0;
for (int j = 0; j < lag; j++)
{
double sum1a = 0; //(x(j+2:n)-x(j+1:n-1)-mu).^2
double sum1b = 0; //(x(2:n-j)-x(1:n-j-1)-mu).^2;
double sum1 = sum1a*sum1b;
double delta = sum1/(Math.Pow(sum2, 2));
varvrt += 0; //(2*(q(i)-j)/q(i))^2)*delta;
}
break;
}
}
ZScore = (VRatio - 1)/Math.Sqrt(varvrt);
PValue = NormCDF(ZScore);
}
public static double Covariance(DenseVector x, DenseVector y)
{
var cov = new DenseVector(x.Count);
double xMean = x.Mean();
double yMean = y.Mean();
if (xMean.Equals(double.NaN) || yMean.Equals(double.NaN))
return double.NaN;
for (int i = 0; i < x.Count; i++)
{
cov[i] = (x[i] - xMean)*(y[i] - yMean);
}
return ((double) 1/(x.Count - 1))*cov.Sum();
}
public static double Kurtosis(this double[] data)
{
var d = new DenseVector(data);
var num = new DenseVector(d.Count);
var denom = new DenseVector(d.Count);
d.CopyTo(num);
d.CopyTo(denom);
for (int i = 0; i < num.Count; i++)
{
num[i] = Math.Pow(d[i] - d.Mean(), 4);
denom[i] = Math.Pow(d[i] - d.Mean(), 2);
}
return (num.Sum()*num.Count)/(Math.Pow(denom.Sum(), 2));
}
public static double AverageRawReturn(this DenseVector d)
{
double[] result = RawRateOfReturn(d.ToArray());
var e = new DenseVector(result);
return e.Mean();
}
