Statistics.NormalFit Method

Overload List

#	Signature	Description
1	procedure NormalFit(const X: TVec; out mu: Double; out sigma: Double);	Calculate parameters for normally distributed values.
2	procedure NormalFit(const X: TVec; out mu: Double; out sigma: Double; var PCIMu: TDoubleArray; var PCISigma: TDoubleArray; Alpha: Double);	Calculate parameters for normally distributed values.

Overload 1: procedure NormalFit(const X: TVec; out mu: Double; out sigma: Double);

Calculate parameters for normally distributed values.

#	Name	Type
1	X	TVec
2	mu	Double
3	sigma	Double

Result: stored in self (calling object)

Examples

var vec1: Vector;
resMu, resSigma : double;
CIMu,CISigma: TTwoElmReal;
begin
    // first, generate 1000 normaly distributed
    // numbers with Mu a=0.0 and Sigma =1.0
    vec1.Size(1000);
    RandomNormal(0.0,1.0,vec1);
    // Now extract the Mu,Sigma and their 95% confidence intervals.
    // Use at max 400 iterations and tolerance 0.0001
    NormalFit(vec1,resMu,resSigma,CIMu,CISigma);
end;

Overload 2: procedure NormalFit(const X: TVec; out mu: Double; out sigma: Double; var PCIMu: TDoubleArray; var PCISigma: TDoubleArray; Alpha: Double);

Calculate parameters for normally distributed values.

#	Name	Description
1	X	Stores data which is assumed to be normaly distributed.
2	mu	Return normal distribution parameter estimator Mu.
3	sigma	Return normal distribution parameter estimator Sigma.
4	PCIMu	Mu (1-Alpha)*100 percent confidence interval.
5	PCISigma	Sigma (1-Alpha)*100 percent confidence interval.
6	Alpha	Confidence interval percentage.

Result: stored in self (calling object)

See Also: StatRandom.RandomNormal, Probabilities.NormalStat