/// <summary>
/// Generates a random vector of observations from the current distribution.
/// </summary>
///
/// <returns>A random vector of observations drawn from this distribution.</returns>
///
public override double Generate()
{
var g = new Accord.Math.Random.GaussianGenerator(
(float)mean, (float)stdDev, Accord.Math.Random.Generator.Random.Next());
return(g.Next());
}
/// <summary>
/// Generates a random vector of observations from the
/// Inverse Gaussian distribution with the given parameters.
/// </summary>
///
/// <param name="mean">The mean parameter mu.</param>
/// <param name="shape">The shape parameter lambda.</param>
/// <param name="samples">The number of samples to generate.</param>
///
/// <returns>An array of double values sampled from the specified Uniform distribution.</returns>
///
public static double[] Random(double mean, double shape, int samples)
{
var u = Accord.Math.Random.Generator.Random;
var g = new Accord.Math.Random.GaussianGenerator(0, 1, u.Next());
double[] r = new double[samples];
for (int i = 0; i < r.Length; i++)
{
double v = g.Next();
double y = v * v;
double x = mean + (mean * mean * y) / (2 * shape) - (mean / (2 * shape)) * Math.Sqrt(4 * mean * shape * y + mean * mean * y * y);
double t = u.NextDouble();
if (t <= (mean) / (mean + x))
{
r[i] = x;
}
else
{
r[i] = (mean * mean) / x;
}
}
return(r);
}
public static Matrix Random(int rows, int cols, double initParamsStdDev, Random rng)
{
Accord.Math.Random.GaussianGenerator gen = new Accord.Math.Random.GaussianGenerator(0, (float)initParamsStdDev, rng.Next());
Matrix result = new Matrix(rows, cols);
for (int i = 0; i < result.W.Length; i++)
{
result.W[i] = gen.Next();
}
return(result);
}
/// <summary>
/// Generates a random vector of observations from the current distribution.
/// </summary>
///
/// <param name="samples">The number of samples to generate.</param>
///
/// <returns>A random vector of observations drawn from this distribution.</returns>
///
public override double[] Generate(int samples)
{
double[] r = new double[samples];
var g = new Accord.Math.Random.GaussianGenerator(
(float)mean, (float)stdDev, Accord.Math.Random.Generator.Random.Next());
for (int i = 0; i < r.Length; i++)
{
r[i] = g.Next();
}
return(r);
}
/// <summary>
/// Generates a random observation from the
/// Inverse Gaussian distribution with the given parameters.
/// </summary>
///
/// <param name="mean">The mean parameter mu.</param>
/// <param name="shape">The shape parameter lambda.</param>
///
/// <returns>A random double value sampled from the specified Uniform distribution.</returns>
///
public static double Random(double mean, double shape)
{
var u = Accord.Math.Random.Generator.Random;
var g = new Accord.Math.Random.GaussianGenerator(0, 1, u.Next());
double v = g.Next();
double y = v * v;
double x = mean + (mean * mean * y) / (2 * shape) - (mean / (2 * shape)) * Math.Sqrt(4 * mean * shape * y + mean * mean * y * y);
double t = u.NextDouble();
if (t <= (mean) / (mean + x))
{
return(x);
}
return((mean * mean) / x);
}
private void FillNaNs(double[][] data)
{
string prefix = "Filling NaNs...";
Utils.DrawMessage(prefix, Utils.CreateProgressBar(Utils.ProgressBarLength, 0), ConsoleColor.Gray);
int drawEvery = Utils.PercentIntervalByLength(data[0].Length);
Random currentRandom = new Random(DataManager.Config.Random.Next());
// Loop through all columns
int row = 0;
for (int col = 0; col < data[0].Length; ++col)
{
// Najdeme si statisticke vlastnosti
double[] d = data.Select(x => x[col]).ToArray();
double[] diff = new double[d.Length - 1];
for (int i = 1; i < d.Length; ++i)
{
diff[i - 1] = d[i] - d[i - 1];
}
diff = diff.Where(x => !double.IsNaN(x)).ToArray();
double mean = diff.Mean();
double stddev = diff.StandardDeviation(mean);
if (diff.Length <= 1)
{
mean = 0;
stddev = 1;
}
// Nahodny generator
Accord.Math.Random.GaussianGenerator gen = new Accord.Math.Random.GaussianGenerator((float)mean, (float)stddev, currentRandom.Next());
// Doplnime zacatek
for (row = 0; row < data.Length; ++row)
{
if (!double.IsNaN(data[row][col]))
{
for (row = row - 1; row >= 0; --row)
{
data[row][col] = data[row + 1][col] + gen.Next();
}
break;
}
}
// Doplnime konec
for (row = data.Length - 1; row >= 0; --row)
{
if (!double.IsNaN(data[row][col]))
{
for (row = row + 1; row < data.Length; ++row)
{
data[row][col] = data[row - 1][col] + gen.Next();
}
break;
}
}
// Doplnime mezery
for (row = 0; row < data.Length; ++row)
{
if (double.IsNaN(data[row][col]))
{
int start = row - 1;
for (row = row + 1; row < data.Length; ++row)
{
if (!double.IsNaN(data[row][col]))
{
int end = row;
// Doplnime od start do end
for (int i = start + 1; i < end; ++i)
{
// Najdeme hodnotu pri aproximaci
double aprox = data[start][col] + (data[end][col] - data[start][col]) / (end - start) * (i - start);
data[i][col] = aprox + gen.Next();
}
break;
}
}
}
}
// Pro jistotu rozkopirovani
for (row = 1; row < data.Length; ++row)
{
if (double.IsNaN(data[row][col]))
{
data[row][col] = data[row - 1][col];
}
}
for (row = data.Length - 2; row >= 0; --row)
{
if (double.IsNaN(data[row][col]))
{
data[row][col] = data[row + 1][col];
}
}
if (col % drawEvery == 0)
{
Utils.DrawMessage(prefix, Utils.CreateProgressBar(Utils.ProgressBarLength, (double)col / data[0].Length * 100.0), ConsoleColor.Gray);
}
}
Utils.DrawMessage(prefix, Utils.CreateProgressBar(Utils.ProgressBarLength, 100), ConsoleColor.Green);
Console.WriteLine();
}
