private static void subtestLogisticsMap <T>(double[][] x, double[][] y, double[][] tx, double[][] ty, ITraining bprop) where T : IRule, new()
{
KMEANSExtractorIO extractor = new KMEANSExtractorIO(10);
var timer = Stopwatch.StartNew();
ANFIS fis = ANFISBuilder <T> .Build(x, y, extractor, bprop, 1000);
timer.Stop();
double err = bprop.Error(tx, ty, fis.RuleBase);
Trace.WriteLine(string.Format("[{1} - {4}]\tLogistic map Error {0}\tElapsed {2}\tRuleBase {3}", err, bprop.GetType().Name, timer.Elapsed, fis.RuleBase.Length, typeof(T).Name), "training");
Assert.IsFalse(err > 1e-2);
}
public void ANFIS_OutputFromDataSet()
{
var sampleSize = RobotArmDataSet.Input.Length - 1;
//StochasticBatch sprop = new StochasticBatch(sampleSize, 1e-5);
//sprop.UnknownCaseFaced += AddRule<GaussianRule2>;
//var sprop = new Backprop(1e-1);
var sprop = new StochasticQprop(sampleSize);
var extractor = new KMEANSExtractorIO(25);
//ANFIS fis = ANFISBuilder<GaussianRule2>.Build(RobotArmDataSet.Input, RobotArmDataSet.OutputTheta1, extractor, sprop, 150);
ANFIS fis = ANFISBuilder <GaussianRule> .Build(RobotArmDataSet.Input, RobotArmDataSet.OutputTheta1, extractor, sprop, 150);
var output1 = fis.Inference(new[] { 1.10413546487088, 2.81104319371924 }).FirstOrDefault(); // 1.1
var output2 = fis.Inference(new[] { 2.31665592712393, 1.9375717475909 }).FirstOrDefault(); // 0.6
var output3 = fis.Inference(new[] { 2.88944142930409, 16.7526454098038 }).FirstOrDefault(); // 1.4
}
public void TestRulesetGeneration()
{
int trainingSamples = 100;
double[][] x = new double[trainingSamples][];
double[][] y = new double[trainingSamples][];
Random rnd = new Random();
for (int i = 0; i < trainingSamples; i++)
{
bool isRigth = i % 2 == 0;
double valx = (isRigth ? 1 : -1) + (0.5 - rnd.NextDouble());
x[i] = new double[] { valx, valx };
y[i] = new double[] { isRigth ? 1 : 0, isRigth ? 0 : 1 };
}
KMEANSExtractorIO extractor = new KMEANSExtractorIO(2);
List <GaussianRule> ruleBase = RuleSetFactory <GaussianRule> .Build(x, y, extractor);
if (ruleBase[0].Z[0] > 0.5)
{
Assert.AreEqual(ruleBase[0].Z[0], 1, 1e-2);
Assert.AreEqual(ruleBase[0].Z[1], 0, 1e-2);
Assert.AreEqual(ruleBase[1].Z[1], 1, 1e-2);
Assert.AreEqual(ruleBase[1].Z[0], 0, 1e-2);
Assert.AreEqual(ruleBase[0].Parameters[0], 1, 1e-1);
Assert.AreEqual(ruleBase[0].Parameters[1], 1, 1e-1);
Assert.AreEqual(ruleBase[1].Parameters[0], -1, 1e-1);
Assert.AreEqual(ruleBase[1].Parameters[1], -1, 1e-1);
}
else
{
Assert.AreEqual(ruleBase[0].Z[1], 1, 1e-2);
Assert.AreEqual(ruleBase[0].Z[0], 0, 1e-2);
Assert.AreEqual(ruleBase[1].Z[0], 1, 1e-2);
Assert.AreEqual(ruleBase[1].Z[1], 0, 1e-2);
Assert.AreEqual(ruleBase[1].Parameters[0], 1, 1e-1);
Assert.AreEqual(ruleBase[1].Parameters[1], 1, 1e-1);
Assert.AreEqual(ruleBase[0].Parameters[0], -1, 1e-1);
Assert.AreEqual(ruleBase[0].Parameters[1], -1, 1e-1);
}
}
void Awake()
{
//Check if there is already an instance of SoundManager
if (instance == null)
{
//if not, set it to this.
instance = this;
}
//If instance already exists:
else if (instance != this)
{
//Destroy this, this enforces our singleton pattern so there can only be one instance of SoundManager.
Destroy(gameObject);
}
//Set SoundManager to DontDestroyOnLoad so that it won't be destroyed when reloading our scene.
DontDestroyOnLoad(gameObject);
Backprop bprop = new Backprop(1e-2);
KMEANSExtractorIO extractor = new KMEANSExtractorIO(5);
fis = ANFISBuilder <GaussianRule> .Build(gameParams, musicParams, extractor, bprop, 5);
}
/// <summary>
///
/// </summary>
/// <param name="ruleNumber"></param>
/// <param name="maxIterations"></param>
/// <returns></returns>
public Task <bool> TrainANFIS(int ruleNumber, int maxIterations, bool useAnalicitalOutcomeForTraining = false)
{
return(Task.Run(() => {
if (!IsDataSetCalculated)
{
throw new ApplicationException("DataSet is not calculated or provided.");
}
var sampleSize = Positions.Count() - 1;
var dynamicObj = useAnalicitalOutcomeForTraining ? Positions.Select(x => new { Point = x, KinematicOutCome = CalculateArmJoint(x).GetAwaiter().GetResult().FirstOrDefault() }) : null;
var input = useAnalicitalOutcomeForTraining
? dynamicObj.Select(x => x.Point).ConvertToANFISParameter()
: Positions.ConvertToANFISParameter();
var theta1ANFIS = Task.Run(() => {
var sPropTheta1 = new StochasticQprop(sampleSize);
var extractorForTheta1 = new KMEANSExtractorIO(ruleNumber);
var expectedOutcome = useAnalicitalOutcomeForTraining
? dynamicObj.Select(x => new[] { x.KinematicOutCome.Theta1.ConvertRadiansToDegrees() }).ToArray()
: AnglesGrid.First().ConvertToANFISParameter();
Theta1ANFIS = ANFISBuilder <GaussianRule> .Build(input,
expectedOutcome, extractorForTheta1, sPropTheta1, maxIterations);
});
var theta2ANFIS = Task.Run(() => {
var sPropTheta2 = new StochasticQprop(sampleSize);
var extractorForTheta2 = new KMEANSExtractorIO(ruleNumber);
var expectedOutcome2 = useAnalicitalOutcomeForTraining
? dynamicObj.Select(x => new[] { x.KinematicOutCome.Theta2.ConvertRadiansToDegrees() }).ToArray()
: AnglesGrid.Last().ConvertToANFISParameter();
Theta2ANFIS = ANFISBuilder <GaussianRule> .Build(input,
expectedOutcome2, extractorForTheta2, sPropTheta2, maxIterations);
});
Task.WaitAll(theta1ANFIS, theta2ANFIS);
IsANFISTrained = true;
return true;
}));
}
