public Bandit(
int kArm = 4,
double epsilon = 0.0,
double initial = 0.0,
double stepSize = 0.1,
bool sampleAverages = false,
int?ucbParam = null,
bool gradient = false,
bool gradientBaseline = false,
double trueReward = 0.0)
{
KArm = kArm;
Epsilon = epsilon;
Initial = initial;
StepSize = stepSize;
SampleAverages = sampleAverages;
UcbParam = ucbParam;
Gradient = gradient;
GradientBaseline = gradientBaseline;
TrueReward = trueReward;
Time = 0;
Indices = Enumerable.Range(0, KArm).ToList();
AverageReward = 0.0;
Ran = new Random(DateTime.Now.Second);
ProbabilityRandom = new ProbabilityRandom <int>();
}
public void Test1()
{
var dicsOfProbs = new Dictionary <string, double>
{
{ "A", 0.1 },
{ "B", 0.2 },
{ "C", 0.4 },
{ "D", 0.2 },
{ "E", 0.1 },
};
var probabilityRandom = new ProbabilityRandom <string>();
foreach (var item in dicsOfProbs)
{
probabilityRandom.SetProb(item.Key, item.Value);
}
probabilityRandom.SetProb("D", 0.1);
probabilityRandom.SetProb("F", 0.1);
var rand = new Random();
var actualProbs = new Dictionary <string, double>
{
{ "A", 0 },
{ "B", 0 },
{ "C", 0 },
{ "D", 0 },
{ "E", 0 },
{ "F", 0 },
};
int maxRand = 50000;
for (int i = 0; i < maxRand; i++)
{
var num = probabilityRandom.Next();
actualProbs[num] = actualProbs[num] + 1;
}
double sum = 0;
foreach (var item in actualProbs)
{
console.WriteLine($"P({item.Key}) = {item.Value / maxRand} ({probabilityRandom.DicOfProbs[item.Key]})");
sum += item.Value / maxRand;
}
console.WriteLine($"P = {sum}");
}