private void EvaluateExercice(string wavPath, CorrelationModel correlation)
{
var score = 0;
var random = new Random();
var count = 0;
if (this.CurrentActivity.F0_exactEvaluated)
{
count++;
this.CurrentActivity.F0_exact = correlation.PCC;
score += (int)(correlation.PCC * 100d);
}
if (this.CurrentActivity.F0_stableEvaluated)
{
count++;
this.CurrentActivity.F0_exact = correlation.CCC;
score += (int)(correlation.CCC * 100d);
}
if (this.CurrentActivity.Intensite_stableEvaluated)
{
count++;
this.CurrentActivity.F0_exact = correlation.CCCin;
score += (int)(correlation.CCCin * 100d);
}
if (this.CurrentActivity.Courbe_f0_exacteEvaluated)
{
count++;
this.CurrentActivity.F0_exact = correlation.PCC;
score += (int)(correlation.PCC * 100d);
}
if (this.CurrentActivity.Duree_exacteEvaluated)
{
count++;
var value = CalculateTimeLength(wavPath);
this.CurrentActivity.Duree_exacte = value;
score += ScoreProvider.EvaluateTimeLength(this.CurrentActivity.Duree_expected, value);
}
if (this.CurrentActivity.JitterEvaluated)
{
count++;
var value = CalculateJitter(wavPath);
this.CurrentActivity.Jitter = value;
score += ScoreProvider.EvaluateJitter(value);
}
if (count == 0)
{
return;
}
score = score / count;
string imagePath = ScoreProvider.ImageResult(score);
Task.Factory.StartNew(() =>
{
//Update Text on the UI thread
Application.Current.Dispatcher.BeginInvoke(DispatcherPriority.Input,
new Action(() => {
var splash = new SplashScreen(imagePath);
splash.Show(false);
splash.Close(new TimeSpan(0, 0, 5));
}));
});
}
public CorrelationModel CalculateCorrelation(ICollection <DataLineItem> expected, ICollection <DataLineItem> exercise)
{
foreach (var lineItem in expected)
{
x.Add(lineItem.Pitch);
xin.Add(lineItem.Intensity);
}
foreach (var lineItem in exercise)
{
y.Add(lineItem.Pitch);
yin.Add(lineItem.Intensity);
}
double[] Xsample = new double[x.Count];
double[] Ysample = new double[y.Count];
double[] Xinsample = new double[xin.Count];
double[] Yinsample = new double[yin.Count];
int lenght = 0;
if (x.Count > y.Count)
{
lenght = y.Count;
}
else
{
lenght = x.Count;
}
for (int i = 0; i < lenght; i++)
{
Xsample[i] = Convert.ToDouble(x[i]);
Ysample[i] = Convert.ToDouble(y[i]);
Xinsample[i] = Convert.ToDouble(xin[i]);
Yinsample[i] = Convert.ToDouble(yin[i]);
}
// calcul de la moyenne
X_moy = CalculerMoyenne(x);
Xin_moy = CalculerMoyenne(xin);
y_moy = CalculerMoyenne(y);
yin_moy = CalculerMoyenne(yin);
var corr = ComputeCoeff(Xsample.ToArray(), Ysample.ToArray());
double cov = ComputeCovariance(Xsample, Ysample);
Console.Out.WriteLine("correlation est: " + corr + "\n");
Console.Out.WriteLine("covariance est: " + cov + "\n");
for (int i = 0; i < x.Count; i++)
{
moyX_Xbar += Math.Pow((Convert.ToDouble(x[i]) - X_moy), 2);
moyXin_Xinbar += Math.Pow((Convert.ToDouble(xin[i]) - Xin_moy), 2);
}
// ********calcul de X_var
x_var = 1.0 / ((x.Count - 1)) * moyX_Xbar;
xin_var = 1.0 / ((xin.Count - 1)) * moyXin_Xinbar;
Console.Out.WriteLine(" xin-var: " + xin_var + "\n");
// calcul de la somme de (y-ymoy)*2
for (int i = 0; i < y.Count; i++)
{
moyY_Ybar += Math.Pow((Convert.ToDouble(y[i]) - y_moy), 2);
moyYin_Yinbar += Math.Pow((Convert.ToDouble(yin[i]) - yin_moy), 2);
}
// calcul de y_var
y_var = 1.0 / ((y.Count - 1)) * moyY_Ybar;
yin_var = 1.0 / ((yin.Count - 1)) * moyYin_Yinbar;
for (int i = 0; i < lenght; i++)
{
vect.Add((Convert.ToDouble(x[i]) - X_moy) * (Convert.ToDouble(y[i]) - y_moy));
vectin.Add(((Convert.ToDouble(xin[i]) - Xin_moy) * ((Convert.ToDouble(yin[i]) - yin_moy))));
}
for (int i = 0; i < lenght; i++)
{
covariance += ((Convert.ToDouble(x[i]) - X_moy) * (Convert.ToDouble(y[i]) - y_moy));
covariancein += ((Convert.ToDouble(xin[i]) - Xin_moy) * (Convert.ToDouble(yin[i]) - yin_moy));
}
covariance = covariance / (y.Count - 1);
covariancein = covariancein / (yin.Count - 1);
//double moy=calculerMoyenne(vect);
//covariance = ;
Console.Out.WriteLine("covariance: " + covariance + "\n");
Console.Out.WriteLine("covariancein: " + covariancein + "\n");
CCC = (2 * covariance) / (x_var + y_var + (Math.Pow((X_moy - y_moy), 2)));
PCC = covariance / ((Math.Sqrt(x_var)) * (Math.Sqrt(y_var)));
CCCin = (2 * covariancein) / (xin_var + yin_var + (Math.Pow((Xin_moy - yin_moy), 2)));
PCCin = covariancein / ((Math.Sqrt(xin_var)) * (Math.Sqrt(yin_var)));
var results = new CorrelationModel
{
CCC = this.CCC,
PCC = this.PCC,
CCCin = this.CCCin,
PCCin = this.PCCin
};
Console.Out.WriteLine("CCC = " + CCC + " PCC = " + PCC + " CCCin = " + CCCin + " PCCin = " + PCCin);
return(results);
}
