public double[] getSamples(int num) // 获取指定个数的样本
{
double[] ret = new double[num];
int[] ret_int = new int[num];
switch (DistributionName)
{
case "Normal":
normalDis.Samples(ret);
break;
case "ContinuousUniform":
continuousUniformDis.Samples(ret);
break;
case "Triangular":
triangularDis.Samples(ret);
break;
case "StudentT":
studentTDis.Samples(ret);
break;
case "DiscreteUniform":
discreteUniform.Samples(ret_int);
for (int i = 0; i < num; i++)
{
ret[i] = ret_int[i];
}
break;
}
return(ret);
}
/// <summary>
/// Given a random model
/// set the weights : an array fill of random float ranging [-1; 1]
/// representing the detailed Belief of an agent
/// </summary>
/// <param name="model"></param>
/// <param name="length"></param>
/// <param name="beliefWeightLevel"></param>
/// <returns></returns>
public void InitializeWeights(RandomGenerator model, byte length, BeliefWeightLevel beliefWeightLevel)
{
float[] beliefBits;
switch (beliefWeightLevel)
{
case BeliefWeightLevel.NoWeight:
beliefBits = DiscreteUniform.Samples(length, 0, 0);
break;
case BeliefWeightLevel.RandomWeight:
beliefBits = model == RandomGenerator.RandomUniform
? ContinuousUniform.Samples(length, 0, RangeMax)
: DiscreteUniform.Samples(length, 0, RangeMax);
break;
case BeliefWeightLevel.FullWeight:
beliefBits = DiscreteUniform.Samples(length, 1, 1);
break;
default:
throw new ArgumentOutOfRangeException(nameof(beliefWeightLevel), beliefWeightLevel, null);
}
Weights = new Bits(beliefBits, 0);
}
public override IList <Point> Generate(int dimensions, double d, int k, int total)
{
var result = new List <Point>(total);
var generated = 0;
var uniformDistribution = new ContinuousUniform(dimensions - dimensions * d, dimensions + 2 * dimensions * d, RandomSource);
while (generated++ < total)
{
var samples = new double[dimensions];
var isPositive = true;
uniformDistribution.Samples(samples);
for (var i = 1; i <= dimensions; i++)
{
if (samples[i - 1] < i || samples[i - 1] > i + i * d)
{
isPositive = false;
break;
}
}
result.Add(new Point(samples)
{
Label = isPositive
});
}
return(result);
}
public void CanSampleSequence()
{
var n = new ContinuousUniform();
var ied = n.Samples();
ied.Take(5).ToArray();
}
public static double[] continuousUniform(double v1, double v2, int num)
{
var n = new ContinuousUniform(v1, v2);
double[] ret = new double[num];
n.Samples(ret);
return(ret);
}
private Task playMatchInit()
{
return(Task.Run(() =>
{
Random rand1 = new Random((int)DateTime.Now.Ticks & 0x0000FFFF);
Random rand2 = new Random((int)DateTime.Now.Ticks & 0x0000FFFF);
ContinuousUniform.Samples(rand1, r1, 0.0, 1.0);
if (d1 == 0)
{
//ContinuousUniform.Samples(r1, 0.0, 0.1);
}
else if (d1 == 1)
{
r1 = r1.Select(z => Math.Sqrt(z)).ToArray();
}
else if (d1 == 2)
{
r1 = r1.Select(z => z * z).ToArray();
}
ContinuousUniform.Samples(rand2, r2, 0.0, 1.0);
if (d2 == 0)
{
//ContinuousUniform.Samples(r1, 0.0, 0.1);
}
else if (d2 == 1)
{
r2 = r2.Select(z => Math.Sqrt(z)).ToArray();
}
else if (d2 == 2)
{
r2 = r2.Select(z => z * z).ToArray();
}
foreach (string n in Team1[2].Split(':')[1].Split(','))
{
in2.bowlers[n] = new int[3];
}
foreach (string n in in2.bowlers.Keys)
{
for (int i = 0; i < 3; i++)
{
in2.bowlers[n][i] = 0;
}
}
foreach (string n in Team2[2].Split(':')[1].Split(','))
{
in1.bowlers[n] = new int[3];
}
foreach (string n in in1.bowlers.Keys)
{
for (int i = 0; i < 3; i++)
{
in1.bowlers[n][i] = 0;
}
}
}));
}
public void Cudnn_UseCase_ForwardConvolution_Double()
{
CudnnContext.DefaultType = CudnnType.Double;
CudnnContext.DefaultTensorFormat = CudnnTensorFormat.MajorRow;
using (var context = CudnnContext.Create())
{
// Set some options and tensor dimensions
int nInput = 100;
int filtersIn = 10;
int filtersOut = 8;
int heightIn = 20;
int widthIn = 20;
int heightFilter = 5;
int widthFilter = 5;
int paddingHeight = 4;
int paddingWeight = 4;
int verticalStride = 1;
int horizontalStride = 1;
int upscalex = 1;
int upscaley = 1;
// Input Tensor Data
double[] xData = new double[nInput * filtersIn * heightIn * widthIn];
ContinuousUniform.Samples(xData, 0, 1);
// Filter Tensor Data
double[] filterData = new double[filtersOut * filtersIn * heightFilter * widthFilter];
ContinuousUniform.Samples(filterData, 0, 1);
// Descriptor for input
var xTensor = CudnnContext.CreateTensor(new CudnnTensorDescriptorParameters(nInput, filtersIn, heightFilter, widthFilter));
// Filter descriptor
var filter = CudnnContext.CreateFilter(new CudnnFilterDescriptorParameters(filtersOut, filtersIn, heightFilter, widthFilter));
// Convolution descriptor
var convolution = CudnnContext.CreateConvolution(new CudnnConvolutionDescriptorParameters(CudnnConvolutionMode.CrossCorrelation, xTensor, filter, paddingHeight, paddingWeight, verticalStride, horizontalStride, upscalex, upscaley));
var output = convolution.GetOutputTensor(CudnnConvolutionPath.Forward);
// Output tensor
var yTensor = CudnnContext.CreateTensor(new CudnnTensorDescriptorParameters(nInput, filtersOut, output.Height, output.Width));
double[] yData = new double[nInput * filtersOut * output.Height * output.Width];
// Perform convolution
context.Forward(xTensor, xData, filter, filterData, convolution, yTensor, yData, CudnnAccumulateResult.DoNotAccumulate);
// Clean up
xTensor.Dispose();
yTensor.Dispose();
filter.Dispose();
convolution.Dispose();
}
}
public override IList <Point> Generate(int dimensions, double d, int k, int total)
{
var result = new List <Point>(total);
var generated = 0;
var uniformDistribution = new ContinuousUniform(-1, 2 * k + d, RandomSource);
while (generated++ < total)
{
var samples = new double[dimensions];
var isPositive = false;
uniformDistribution.Samples(samples);
for (var j = 1; j <= k; j++)
{
if (samples.Sum() > d * j)
{
continue;
}
isPositive = true;
for (var i = 1; i <= dimensions; i++)
{
for (var l = i + 1; l <= dimensions; l++)
{
if (samples[i - 1] / Math.Tan(Math.PI / 12) - samples[l - 1] * Math.Tan(Math.PI / 12) >= 2 * j - 2 &&
samples[l - 1] / Math.Tan(Math.PI / 12) - samples[i - 1] * Math.Tan(Math.PI / 12) >= 2 * j - 2)
{
continue;
}
isPositive = false;
break;
}
}
if (isPositive)
{
break;
}
}
result.Add(new Point(samples)
{
Label = isPositive
});
}
return(result);
}
public override IList <Point> Generate(int dimensions, double d, int k, int positives, int negatives)
{
var result = new List <Point>(positives + negatives);
var generatedPositives = 0;
var generatedNegatives = 0;
var uniformDistributionForPositives = new ContinuousUniform(1 - d, d + dimensions, RandomSource);
var uniformDistributionForNegatives = new ContinuousUniform(1 - 2 * d, dimensions + 2 * d, RandomSource);
while (generatedPositives < positives)
{
var samples = new double[dimensions];
uniformDistributionForPositives.Samples(samples);
if (Distance.Euclidean(Enumerable.Range(1, dimensions).Select(Convert.ToDouble).ToArray(), samples) > d)
{
continue;
}
generatedPositives++;
result.Add(new Point(samples)
{
Label = true
});
}
while (generatedNegatives < negatives)
{
var samples = new double[dimensions];
uniformDistributionForNegatives.Samples(samples);
if (Distance.Euclidean(Enumerable.Range(1, dimensions).Select(Convert.ToDouble).ToArray(), samples) <= d)
{
continue;
}
generatedNegatives++;
result.Add(new Point(samples)
{
Label = false
});
}
return(result);
}
/// <summary>
/// Initialize the forgetting knowledge process with a random Selecting mode
/// </summary>
/// <param name="actorKnowledge"></param>
/// <param name="nextForgettingRate"></param>
/// <returns></returns>
public float[] InitializeForgettingKnowledgeRandom(ActorKnowledge actorKnowledge, float nextForgettingRate)
{
if (actorKnowledge is null)
{
throw new ArgumentNullException(nameof(actorKnowledge));
}
var forgettingKnowledgeBits = ContinuousUniform.Samples(actorKnowledge.Length, 0, 1);
var threshold = NextMean();
for (byte i = 0; i < actorKnowledge.Length; i++)
{
forgettingKnowledgeBits[i] = forgettingKnowledgeBits[i] < threshold ? nextForgettingRate : 0;
}
return(forgettingKnowledgeBits);
}
/// <summary>
/// Given a random model and a BeliefLevel
/// return the beliefBits for the agent: an array fill of random binaries
/// representing the detailed belief of an agent
/// </summary>
/// <param name="model"></param>
/// <param name="beliefLevel"></param>
/// <returns></returns>
public float[] InitializeBits(RandomGenerator model, BeliefLevel beliefLevel)
{
float[] beliefBits;
switch (model)
{
case RandomGenerator.RandomUniform:
{
float min;
float max;
if (beliefLevel == BeliefLevel.Random)
{
min = RangeMin;
max = RangeMax;
}
else
{
min = GetMinFromBeliefLevel(beliefLevel);
max = GetMaxFromBeliefLevel(beliefLevel);
}
beliefBits = ContinuousUniform.Samples(Length, min, max);
break;
}
case RandomGenerator.RandomBinary:
{
if (beliefLevel == BeliefLevel.Random)
{
beliefBits = ContinuousUniform.FilteredSamples(Length, RangeMin, RangeMax);
}
else
{
var mean = 1 - GetValueFromBeliefLevel(beliefLevel);
beliefBits = ContinuousUniform.FilteredSamples(Length, mean);
}
break;
}
default:
throw new ArgumentOutOfRangeException(nameof(model), model, null);
}
return(beliefBits);
}
public override IList <Point> Generate(int dimensions, double d, int k, int total)
{
var result = new List <Point>(total);
var generated = 0;
var uniformDistribution = new ContinuousUniform(1 - 2 * d, dimensions + 2 * d + (2 * Math.Sqrt(6) * (k - 1) * d) / Math.PI, RandomSource);
while (generated++ < total)
{
var samples = new double[dimensions];
var isPositive = true;
uniformDistribution.Samples(samples);
for (var j = 1; j <= k; j++)
{
isPositive = true;
for (var i = 1; i <= dimensions; i++)
{
if (Distance.Euclidean(Enumerable.Range(1, dimensions).Select(c => Convert.ToDouble(c) + _edgeCoefficient * d * (j - 1) / c).ToArray(), samples) <= d)
{
continue;
}
isPositive = false;
break;
}
if (isPositive)
{
break;
}
}
result.Add(new Point(samples)
{
Label = isPositive
});
}
return(result);
}
public override IList <Point> Generate(int dimensions, double d, int k, int total)
{
var result = new List <Point>(total);
var generated = 0;
var uniformDistribution = new ContinuousUniform(-1, 2 + d, RandomSource);
while (generated++ < total)
{
var samples = new double[dimensions];
var isPositive = true;
uniformDistribution.Samples(samples);
if (samples.Sum() > d)
{
isPositive = false;
}
else
{
for (var i = 1; i <= dimensions; i++)
{
for (var j = i + 1; j < dimensions; j++)
{
if (samples[i - 1] / Math.Tan(Math.PI / 12) - samples[j - 1] * Math.Tan(Math.PI / 12) < 0 ||
samples[j - 1] / Math.Tan(Math.PI / 12) - samples[i - 1] * Math.Tan(Math.PI / 12) < 0)
{
isPositive = false;
}
}
}
}
result.Add(new Point(samples)
{
Label = isPositive
});
}
return(result);
}
public override IList <Point> Generate(int dimensions, double d, int k, int total)
{
var result = new List <Point>(total);
var generated = 0;
var uniformDistribution = new ContinuousUniform(1 - 2 * d, dimensions + 2 * d, RandomSource);
while (generated++ < total)
{
var samples = new double[dimensions];
uniformDistribution.Samples(samples);
var label = Distance.Euclidean(Enumerable.Range(1, dimensions).Select(Convert.ToDouble).ToArray(), samples) <= d;
result.Add(new Point(samples)
{
Label = label
});
}
return(result);
}
public void FailSampleSequenceStatic()
{
Assert.Throws <ArgumentOutOfRangeException>(() => ContinuousUniform.Samples(new Random(), 0.0, -1.0).First());
}
public void CanSampleSequenceStatic()
{
var ied = ContinuousUniform.Samples(new Random(), 0.0, 1.0);
ied.Take(5).ToArray();
}
public void FailSampleSequenceStatic()
{
Assert.That(() => ContinuousUniform.Samples(new Random(0), 0.0, -1.0).First(), Throws.ArgumentException);
}
/// <summary>
/// Given a KnowledgeModel and a KnowledgeLevel
/// return the knowledgeBits for the agent: an array fill of random binaries
/// representing the detailed knowledge of an agent
/// </summary>
/// <param name="length"></param>
/// <param name="model"></param>
/// <param name="knowledgeLevel"></param>
/// <param name="step"></param>
/// <returns></returns>
public void InitializeKnowledge(byte length, RandomGenerator model, KnowledgeLevel knowledgeLevel, ushort step)
{
float[] knowledgeBits;
switch (model)
{
case RandomGenerator.RandomUniform:
{
float min;
float max;
switch (knowledgeLevel)
{
case KnowledgeLevel.Random:
min = 0;
max = 1;
break;
default:
min = Knowledge.GetMinFromKnowledgeLevel(knowledgeLevel);
max = Knowledge.GetMaxFromKnowledgeLevel(knowledgeLevel);
break;
}
knowledgeBits = ContinuousUniform.Samples(length, min, max);
if (Math.Abs(min - max) < Constants.Tolerance)
{
SetKnowledgeBits(knowledgeBits, step);
return;
}
for (byte i = 0; i < knowledgeBits.Length; i++)
{
if (knowledgeBits[i] < min * (1 + 0.05))
{
// In randomUniform, there is quasi no bit == 0. But in reality, there are knowledgeBit we ignore.
// We force the lowest (Min +5%) knowledgeBit to 0
knowledgeBits[i] = 0;
}
}
break;
}
case RandomGenerator.RandomBinary:
{
switch (knowledgeLevel)
{
case KnowledgeLevel.Random:
knowledgeBits = ContinuousUniform.FilteredSamples(length, 0, 1);
break;
default:
var mean = 1 - Knowledge.GetValueFromKnowledgeLevel(knowledgeLevel);
knowledgeBits = ContinuousUniform.FilteredSamples(length, mean);
break;
}
break;
}
default:
throw new ArgumentOutOfRangeException(nameof(model), model, null);
}
SetKnowledgeBits(knowledgeBits, step);
}
public override IList <Point> Generate(int dimensions, double d, int k, int positives, int negatives)
{
var result = new List <Point>(positives + negatives);
var generatedPositives = 0;
var generatedNegatives = 0;
var uniformDistribution = new ContinuousUniform(-1, 2 + d, RandomSource);
while (generatedPositives < positives)
{
var samples = new double[dimensions];
var isPositive = true;
uniformDistribution.Samples(samples);
if (samples.Sum() > d)
{
continue;
}
for (var i = 1; i <= dimensions; i++)
{
for (var j = i + 1; j < dimensions; j++)
{
if (samples[i - 1] / Math.Tan(Math.PI / 12) - samples[j - 1] * Math.Tan(Math.PI / 12) < 0 ||
samples[j - 1] / Math.Tan(Math.PI / 12) - samples[i - 1] * Math.Tan(Math.PI / 12) < 0)
{
isPositive = false;
break;
}
}
}
if (!isPositive)
{
continue;
}
generatedPositives++;
result.Add(new Point(samples)
{
Label = true
});
}
while (generatedNegatives < negatives)
{
var samples = new double[dimensions];
uniformDistribution.Samples(samples);
if (samples.Sum() < d)
{
continue;
}
generatedNegatives++;
result.Add(new Point(samples)
{
Label = false
});
}
return(result);
}
public override IList <Point> Generate(int dimensions, double d, int k, int positives, int negatives)
{
var result = new List <Point>(positives + negatives);
var generatedPositives = 0;
var generatedNegatives = 0;
var uniformDistributionForPositives = new ContinuousUniform(1, dimensions + dimensions * d, RandomSource);
var uniformDistributionForNegatives = new ContinuousUniform(dimensions - dimensions * d, dimensions + 2 * dimensions * d, RandomSource);
while (generatedPositives < positives)
{
var samples = new double[dimensions];
var isPositive = true;
uniformDistributionForPositives.Samples(samples);
for (var i = 1; i <= dimensions; i++)
{
if (samples[i - 1] < i || samples[i - 1] > i + i * d)
{
isPositive = false;
break;
}
}
if (!isPositive)
{
continue;
}
generatedPositives++;
result.Add(new Point(samples)
{
Label = true
});
}
while (generatedNegatives < negatives)
{
var samples = new double[dimensions];
var isNegative = false;
uniformDistributionForNegatives.Samples(samples);
for (var i = 1; i <= dimensions; i++)
{
if (samples[i - 1] < i || samples[i - 1] > i + i * d)
{
isNegative = true;
break;
}
}
if (!isNegative)
{
continue;
}
generatedNegatives++;
result.Add(new Point(samples)
{
Label = false
});
}
return(result);
}
public void CanSampleSequence()
{
var n = new ContinuousUniform();
var ied = n.Samples();
var e = ied.Take(5).ToArray();
}
/// <summary>
/// Run example
/// </summary>
/// <a href="http://en.wikipedia.org/wiki/Uniform_distribution_%28continuous%29">ContinuousUniform distribution</a>
public void Run()
{
// 1. Initialize the new instance of the ContinuousUniform distribution class with default parameters.
var continuousUniform = new ContinuousUniform();
Console.WriteLine(@"1. Initialize the new instance of the ContinuousUniform distribution class with parameters Lower = {0}, Upper = {1}", continuousUniform.LowerBound, continuousUniform.UpperBound);
Console.WriteLine();
// 2. Distributuion properties:
Console.WriteLine(@"2. {0} distributuion properties:", continuousUniform);
// Cumulative distribution function
Console.WriteLine(@"{0} - Сumulative distribution at location '0.3'", continuousUniform.CumulativeDistribution(0.3).ToString(" #0.00000;-#0.00000"));
// Probability density
Console.WriteLine(@"{0} - Probability density at location '0.3'", continuousUniform.Density(0.3).ToString(" #0.00000;-#0.00000"));
// Log probability density
Console.WriteLine(@"{0} - Log probability density at location '0.3'", continuousUniform.DensityLn(0.3).ToString(" #0.00000;-#0.00000"));
// Entropy
Console.WriteLine(@"{0} - Entropy", continuousUniform.Entropy.ToString(" #0.00000;-#0.00000"));
// Largest element in the domain
Console.WriteLine(@"{0} - Largest element in the domain", continuousUniform.Maximum.ToString(" #0.00000;-#0.00000"));
// Smallest element in the domain
Console.WriteLine(@"{0} - Smallest element in the domain", continuousUniform.Minimum.ToString(" #0.00000;-#0.00000"));
// Mean
Console.WriteLine(@"{0} - Mean", continuousUniform.Mean.ToString(" #0.00000;-#0.00000"));
// Median
Console.WriteLine(@"{0} - Median", continuousUniform.Median.ToString(" #0.00000;-#0.00000"));
// Mode
Console.WriteLine(@"{0} - Mode", continuousUniform.Mode.ToString(" #0.00000;-#0.00000"));
// Variance
Console.WriteLine(@"{0} - Variance", continuousUniform.Variance.ToString(" #0.00000;-#0.00000"));
// Standard deviation
Console.WriteLine(@"{0} - Standard deviation", continuousUniform.StdDev.ToString(" #0.00000;-#0.00000"));
// Skewness
Console.WriteLine(@"{0} - Skewness", continuousUniform.Skewness.ToString(" #0.00000;-#0.00000"));
Console.WriteLine();
// 3. Generate 10 samples of the ContinuousUniform distribution
Console.WriteLine(@"3. Generate 10 samples of the ContinuousUniform distribution");
for (var i = 0; i < 10; i++)
{
Console.Write(continuousUniform.Sample().ToString("N05") + @" ");
}
Console.WriteLine();
Console.WriteLine();
// 4. Generate 100000 samples of the ContinuousUniform(0, 1) distribution and display histogram
Console.WriteLine(@"4. Generate 100000 samples of the ContinuousUniform(0, 1) distribution and display histogram");
var data = new double[100000];
ContinuousUniform.Samples(data, 0.0, 1.0);
ConsoleHelper.DisplayHistogram(data);
Console.WriteLine();
// 5. Generate 100000 samples of the ContinuousUniform(2, 10) distribution and display histogram
Console.WriteLine(@"5. Generate 100000 samples of the ContinuousUniform(2, 10) distribution and display histogram");
ContinuousUniform.Samples(data, 2, 10);
ConsoleHelper.DisplayHistogram(data);
}
public void FailSampleSequenceStatic()
{
var ied = ContinuousUniform.Samples(new Random(), 0.0, -1.0).First();
}
public static Vec MakeRandomVector(int size, double min, double max)
{
return(new DenseVector(ContinuousUniform.Samples(Random, min, max).Take(size).ToArray()));
}